Journal of Governance and Regulation / Volume 1, Issue 2, 2012

JOURNAL OF GOVERNANCE AND REGULATION

Postal Address:

Postal Box 36 Sumy 40014 Ukraine

Tel: +380-542-698125 Fax: +380-542-698125 e-mail: [email protected] www.virtusinterpress.org

Journal of Governance and Regulation is published four times a year, in September- November, December-February, March-May and June-August, by Publishing House “Virtus Interpress”, Kirova Str. 146/1, office 20, Sumy, 40021, Ukraine.

Information for subscribers: New orders requests should be addressed to the Editor by e- mail. See the section "Subscription details".

Back issues: Single issues are available from the Editor. Details, including prices, are available upon request.

Advertising: For details, please, contact the Editor of the journal.

Journal of Governance and Regulation

ISSN 2220-9352 (printed version)

3 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

EDITORIAL

Dear readers!

The recent issue of the journal is devoted to several governance and regulation issues.

Eiman A. Algharaballi and Michelle Goyen review Kuwaiti accounting and auditing requirements and document the regulation of equities trading in Kuwait over the period from 1983 to 2011.

Brendan Lambe analyses the returns on stock prices of target companies surrounding the first publicized dates of completed takeovers in the UK between 2001 and 2010. He investigates whether observable factors create this price run-up or if it is the result of disclosed insider trading.

Alessandro Carboni and Andrea Carboni study the cash-CDS basis and its implication for market strategies and price discovery, together with the role of credit risk common factors.

Scott A. Beaulier, Peter J. Boettke and Leonid A. Krasnozhon compare the Czech Republic’s economic, political, and social performance to these benchmarks in all other post- socialist countries since they began their transitions.

Marcela Giraldo evaluates the incentives that banks have to herd. A complete literature review that focuses on papers from the last fifteen years, and a model of several banks and infinite time periods is presented.

José Vaz Ferreira investigates the pre and post privatization financial, social and operational performance of forty two Portuguese companies in most of sectors of economic activity that experience full or partial privatization through public share offering, direct sale or public contest, for the period from 1989 to 2009.

Kevin A. Diehl seeks to update and finally determine for the Fortune 500 whether the market values the inventory valuation choice of last-in, first-out (LIFO) over first-in, first-out (FIFO) as some signal of reporting and management quality.

We hope that you will enjoy reading the journal and in the future we will receive new papers, outlining the most important issues in the field of governance and regulation.

4 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

JOURNAL OF GOVERNANCE AND REGULATION

VOLUME 1, ISSUE 2, 2012

CONTENTS

Editorial 4

CORPORATE REPORTING, SECURITY REGULATION AND TRADING ON THE KUWAITI STOCK EXCHANGE (KSE) – INSTITUTIONAL IMPLICATIONS FOR RESEARCH 7

Eiman A. Algharaballi, Michelle Goyen

The regulation of equities trading in Kuwait over the period from 1983 to 2011 is documented in this paper. An eclectic approach has resulted in overlapping responsibilities for the three main regulatory and supervisory bodies. Regulation appears to be responsive to market crises. As a result, regulations have tended to change with market conditions. Kuwaiti accounting and auditing requirements are also reviewed. The institutional setting in Kuwait has a number of implications for capital market based research. Informational inefficiency precludes research that relies on the assumption that security price reflects firm value.

IS INSIDER TRADING REGULATION EFFECTIVE? EVIDENCE FROM UK TAKEOVER ACTIVITY 24

Brendan Lambe

The returns on stock prices of target companies surrounding the first publicized dates of completed takeovers in the UK between 2001 and 2010 are analyzed in this study. Two samples are created of 209 and 197 firms for announcement and rumored dates respectively. Both demonstrate statistically significant cumulative abnormal returns (CARs) prior to the release of information about the impending bid. This paper investigates whether observable factors create this price run-up or if it is the result of disclosed insider trading. Cross sectional analysis of CARs indicates that trading on material nonpublic information goes undisclosed.

THE CASH-CDS BASIS FOR SOVEREIGN COUNTRIES: MARKET STRATEGY, PRICE DISCOVERY AND DETERMINANTS 49

Alessandro Carboni, Andrea Carboni

The cash-CDS basis and its implication for market strategies and price discovery, together with the role of credit risk common factors are analysed. A positive net income is derived with a negative basis, once funding costs are considered. There exists an arbitrage opportunity for Greece in 2010, with a negative basis of more than 100 bp. Our comparison with three different basis shows that while converging markets seem adopt the same strategy, in particular for Portugal, Ireland and Greece. Results for price discovery show that the CDS market moves ahead the bond market. Finally, our empirical analysis shows that the global risk factor contributes to increase the basis, while the banking sector vulnerability proxy offers a negative contribution.

5 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

IN DEFENSE OF SHOCK THERAPY: POST-SOCIALIST TRANSITION OF THE CZECH REPUBLIC 72

Scott A. Beaulier, Peter J. Boettke, Leonid A. Krasnozhon

The authors propose that a meaningful benchmark is the experience of the Czech Republic, Russia, and other transition economies which share similar approach to the market reforms, but have solved political economy problems of credibility and commitment differently. Researchers compare the Czech Republic’s economic, political, and social performance to these benchmarks in all other post-socialist countries since they began their transitions. Results show that the Czech transition is a consistent success because the Havel shock therapy has solved the political economy problems of reform’s credibility and state’s commitment to reform.

BANKS’ INCENTIVES TO OVER-HERD 86

Marcela Giraldo

This paper evaluates the incentives that banks have to herd. It includes a complete literature review that focuses on papers from the last fifteen years, and a model of several banks and infinite time periods. The literature review looks at recent academic papers that have examined the different causes of bank herding. The model is discussed theoretically and then a numerical example explores the significance of its coefficients. The model section concludes that any policy that reduces the costs of overinvestment increases the incentives of banks to herd.

THE PERFORMANCE OF NEWLY PRIVATIZED FIRMS: THE CASE OF PORTUGAL 92

José Vaz Ferreira

The aim of this study is to investigate the pre and post privatization financial, social and operational performance of forty two Portuguese companies in most of sectors of economic activity that experience full or partial privatization through public share offering, direct sale or public contest, for the period from 1989 to 2009. That is, this work investigates, whether or not, the privatization of state-owned enterprises (SOE’s) had caused improvements on the economic and financial health of those privatized companies, as it is suggested by the literature of property rights, public choice and agency theory. Firstly, significant improvements on profitability, operating efficiency, capital investment, real output, dividend payout, treasury applications, activity levels and capital structure are documented. Secondly, significant decreases in employment after privatization are experienced. Thirdly, it is observed that, following privatization, the financial equilibrium (short and long) of firms was negatively affected. Lastly, results are generally robust surviving the partition of the dataset into various sub-samples.

LOWER TAXES OR HIGHER EXECUTIVE BONUSES: HOW INVENTORY VALUATION CHOICES BEST EXHIBIT US CORPORATE GOVERNANCE FAILINGS 126

Kevin A. Diehl

This research seeks to update and finally determine for the Fortune 500 whether the market values the inventory valuation choice of last-in, first-out (LIFO) over first-in, first-out (FIFO) as some signal of reporting and management quality. The market can adjust LIFO earnings to FIFO earnings. Thus, the only issue then is that companies choosing FIFO pay higher taxes, which shareowners should disfavor. Indeed, only 20 percent of the Fortune 500 utilizes LIFO to value any inventory. However, after Spearman correlations and logistic regression, the research statistically significantly shows that investors are willing to give premiums on the price of stock for the choice of LIFO. Thus, companies should choose LIFO to reduce taxes and increase their stock prices.

SUBSCRIPTION DETAILS 130

6 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

CORPORATE REPORTING, SECURITY REGULATION AND TRADING ON THE KUWAITI STOCK EXCHANGE (KSE) – INSTITUTIONAL IMPLICATIONS FOR RESEARCH

Eiman A. Algharaballi*, Michelle Goyen**

Abstract

Key-words: Corporate Reporting, Security Regulation, Trading, Kuwaiti Stock Exchange

* USQ School of Business and Law, University of Southern Queensland, Australia. College of Business Studies, Kuwait ** UNE School of Business, University of New England, Armidale, Australia

1 Introduction 2 Kuwait Stock Exchange and Major Events The focus of this paper is the regulation of equities trading in Kuwait. Equities are currently traded on The Kuwait Stock Exchange (KSE) ranks first in the the official market – the Kuwait Stock Exchange Arab world based on turnover ratio, second in stock- (KSE) or on the ‘parallel market’ which facilitates value traded, and third in market capitalization the trading of companies that cannot meet the listing (Aldaihani and Aldeehani, 2008). The public requirements of the main board of the KSE. corporation has a relatively short history in Kuwait Increasingly, capital markets researchers are finding compared with western economies. Al-Yaqout (2006) that institutional features of markets impact on the identifies the discovery of oil, the subsequent influx applicability of theories developed in different of revenue and the Kuwaiti government’s recognition markets provide both challenges and opportunities of the benefits of corporate structures developing the for research. Therefore, this paper contributes to the economy as the critical factors in the emergence of developing literature based in Kuwaiti financial the corporation in Kuwait. The National Bank of markets by documenting changing regulations Kuwait, established in 1952, became the country’s relevant to equities trading over the last three first public company (Al-Sultan, 1989, Bley and decades. Chen, 2006). The major events relevant to regulatory The history of regulation of equities trading in changes discussed are the 1976-1977 crisis, the Al- Kuwait is largely characterised as ‘regulation in Manakh crisis, the Gulf War, the boom of 2003-2005 response to crises’. Therefore, the first section of this and the global financial crisis. paper outlines significant economic events to contextualise the discussion of regulation in section 2.1 The inception of corporate regulation 4. A brief overview of the types of equities that trade and the 1976-77 crisis on the KSE is presented in section 3. Institutional arrangements for the accounting and auditing The Commercial Companies Law (No.15), the first profession are discussed in section 5. The paper legislation to organise and regulate companies in concludes with some implications of Kuwaiti Kuwait, was promulgated in 1960. However, the first institutional for accounting and finance research. significant step toward regulating domestic securities trading did not occur until the introduction of Law No.32 in 1970 (Al-Yaqout, 2006; Alanezi, 2006). Law No.32 created a consultation committee to supervise trading, designed the stock market

7 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

framework, attempted to protect the economy from Investors once again failed to cover their post-dated stock price volatility, and provided for the assessment cheques and the OTC market collapsed (Oxford of foreign companies seeking to register their shares Business Group, 2006b). in the market. Stock trading on the exchange Analysts agree that the failure of the OTC commenced in 1977 (Gombers et al., 2008; Masih et market, known as the ‘Al-Manakh crisis’ was caused al., 2010). by inadequate government regulation, insufficient The Kuwaiti stock market was characterized by financial disclosure, frivolous speculation, the use of a rise in speculative activities in the 1970s (Hassan et post-dated cheques, and lack of government control al., 2003) as large numbers of inexperienced over Gulf shareholding companies (Al-Mutairi, investors enthusiastically entered the market (Al- 2004). The crisis precipitated the Ameri Decree Yaqout, 2006; Alanezi, 2006). This rise in which reorganised the Kuwaiti stock market as an speculative trading was fuelled by the “forward independent financial institution guided by an method”, which allowed traders to use post-dated executive administration and a Market Committee to cheques to settle payments (Al-Qenae, 2000; Al- protect investors, issuers, and brokers (Oxford Yaqout, 2006; Alanezi, 2006). Effectively, shares Business Group, 2006a). The KSE was then were purchased at multiples of the current price with established and trading commenced in 1984 (KSE, deferred settlement (Al-Sultan, 1989) and lack of 2010a). Regulators tightened market controls by funds to invest was no impediment to trading. stiffening listing requirements (Alsalman, 2002). As Combined with a lack of sufficient regulation, these a result, the numbers of new listings, shares issued, features provided scope for key investors to and stockbrokers were limited. Numerous regulations manipulate the market for short-term gains during were also promulgated to bolster investors’ this period (Al-Qenae, 2000). confidence in the market (Al-Qenae, 2000) and the By late 1976, the rapid rise in trading activity government spent many years trying to control debt and huge inflation of share prices contributed to a and scheduling settlements for outstanding post-dated calamitous market collapse. Demand from investors cheques after the Al-Manakh crisis. fell as stock prices rose and the widespread use of the “forward method” of settlement created significant 2.3 The Gulf War, 1990-1991 levels of debt for investors (Al-Yaqout, 2006), the clearing system was inadequate and the organization The invasion of Kuwait was a major shock to the of the stock market remained poor (Al-Qenae, 2000; nation and to its economic system. The massive Al-Yaqout, 2006; Alanezi, 2006). By the end of destruction that followed changed the development 1977, stocks prices had fallen sharply and trading path of the Kuwaiti economy (International Monetary volume dropped 66% compared with the previous Fund, 2005). An unfortunate implication for research year (Al-Sultan, 1989; Al-Yaqout, 2006). In response is that much historical data were lost, constraining to the collapse, the government placed a moratorium the sampling timeframe for Kuwaiti market research. on the establishment of any new local shareholding The post-war period in Kuwait was a companies or new equity raisings by existing challenging one. Economic performance suffered companies from 1977 to 1979 (Al-Sultan, 1989). greatly in the 1990s, and the KSE was closed from August 1990 to September 1992 (Annual Economic 2.2 The Al-Manakh crisis, 1982 Report of the KSE, 1990/1991). After reopening, the KSE struggled to rebuild investor confidence and As the investment opportunity set for Kuwaiti stock market activity remained sluggish. The Kuwait investors was constrained by the ban on new Automated Trading System (KATS), designed to shareholding companies and new equity issues by improve market competition, liquidity, and existing companies, ‘Gulf shareholding companies’ transparency, was introduced in 1996 (Annual provided popular investment vehicles. These Gulf Economic Report of the KSE, 1996). The KATS companies were effectively owned by Kuwaiti facilitates faster transactions than were possible investors but were incorporated in the Gulf Emirates under the previous manual system, allowing traders (Al-Sultan, 1989) and, as such, were prohibited from to register their bids and offers which are then trading on the official Kuwaiti market (Al-Qenae, matched according to the priority of prices (Al- 2000). The Kuwaiti over-the-counter (OTC) or ‘Al- Hashel, 2003). With the introduction of the Manakh Stock Market’ developed to trade Gulf automated trading system, investor trust was companies (Elshamy and Al-Qenae, 2005) and gradually regained and the KSE again became active unlisted Kuwaiti Closed Shareholding Companies (Annual Economic Report of the KSE, 1995; (Al-Sultan, 1989). This unregulated market was International Monetary Fund, 2004). popular with investors, generating four times the trading volume of the official market and double the 2.4 The boom, 2003-2005 equity base of the official market (Al-Sultan, 1989). After a meteoric rise, the OTC index fell by 45% in The elimination of Saddam Hussein’s regime in April the six months to August 1982 (Al-Sultan 1989). 2003 coupled with higher international prices for

8 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

crude oil contributed to an economic boom in Kuwait against the KSE was filed by one group attempting to during the 2003-2005period (International Monetary recoup their losses (Al-Atrabi and Al-Sayed, 2008). Fund, 2004; Oxford Business Group, 2006b). Greater As a result, on November 13, 2008, the economic and political security prompted increased Administrative Court in Kuwait suspended all trading investment on the exchange and in new business on the KSE for two working days (Al-Shal Report, initiations. The resultant increase in liquidity and 2008). This action set a striking precedent for the financial growth in Kuwait contributed to the KSE KSE. Trading had not been suspended even during ranking among the best-performing stock markets the worst days of Al-Manakh crisis. Many financial worldwide in 2003 (Oxford Business Group, 2006b). analysts considered that the decision to suspend A sharp rise in market capitalization, new corporate trading was ill conceived and revealed the Kuwaiti listings (specifically IPOs), and market index values regulators’ inability to cope with the crisis. They was experience in 2004 and 2005 (Annual Economic argued that market forces should be allowed to Report of the KSE, 2004; 2005). Further, many prevail even when resultant clearing prices are Kuwaiti firms benefited from profitable extreme and that the losses the Kuwaiti market reconstruction projects in Iraq. incurred were no different than those sustained by other markets around the world (Al-Shal Report, 2.5 The global financial crisis, 2008 2008). They further contended that a decision to suspend market transactions should be treated as a The financial crisis that began in the United States in matter of national security, not jurisprudence (Al- 2008 quickly spread to countries around the world, Shal Report, 2008). leaving collapsed financial markets, economic Kuwait was the last of the GCC countries to recession, rising unemployment, and personal and establish a Capital Market Authority (CMA) when business bankruptcies in its wake. Various factors the Kuwaiti Parliament enacted the Capital Market created the crisis, but a chief cause was the poor Law in February 2010 (Eiman 2010). The law was quality of subprime mortgages in the United States gazetted on March 13, 2011 and its objective is to (Shiller, 2008). Kuwait faced a difficult financial create the CMA as a single, independent, accountable situation, as did many countries around the world authority, managed by a Board of Commissioners (Al-Mutawaa, 2009). The KSE index experienced with the power to develop and regulate the capital large losses relative to some of the leading stock market in Kuwait. This step is anticipated to enhance markets1. That the crisis affected Kuwait, a small transparency, trust and confidence in the Kuwaiti country with no outstanding debt, to the same extent financial system and represents the most recent as many larger countries with more debt, surprised regulatory change at the time of writing. The next many (Al-Mutawaa, 2009). This disproportionately section provides a brief overview of the types of large response in the Kuwaiti economy can be companies that list on the KSE. attributed to four main factors: strong negative investor sentiment on the KSE; a rapid decrease in 3 Equities Trading On the Kse the value of listed investments; lower oil prices and the large exposures of some Kuwaiti financial Shares of companies listed on the KSE are classified institutions to poorly performing foreign investments into eight sectors: banking, investment, insurance, (Al-Mutawaa, 2009). Moosa (2010) is less surprised real estate, industrial, services, food, and non– by the impact of the crisis on the KSE. He shows that Kuwaiti companies (KSE, 2010b). Consistent with the level of integration of the KSE with US markets most investors internationally, shareholders have is not high and argues that Gulf Co-operative Council limited liability and are not responsible for the (GCC) stock markets (including Kuwaiti) were company’s obligations except to the extent of the subject domestic factors that were major contributors face value of the shares subscribed (Law No.15, to the severity of the stock market response. 1960). Article No. 10 of the Ameri Decree (1983, p. In response, Kuwaiti government regulators and 13) specifies the following entities as members of the financial institutions joined forces to take actions to KSE:2 support the economy. The Central Bank of Kuwait  Kuwaiti Public Shareholding Companies (KSCs) (CBK), for example, lowered interest rates on loans  Kuwaiti Closed Shareholding Companies and directed banks to increase their capital to (KSCCs)3 enhance the stability of the banking sector (Al-  Middlemen of the Stock Exchange (stockbrokers) Mutawaa, 2009). KSE investors, however, were not sufficiently impressed by such efforts and a lawsuit 2 Stock exchange members must pay registration fees and an annual subscription fee. 1 The percentage change in the KSE index from closing 3 KSCCs may be either listed or unlisted. KSCCs are prices for 2007 to those of 2008 was -43.12%. In incorporated by an official document issued by the comparison, the change in the FTSE 100 index (London promoters, of whom there must be at least five. Shares of Stock Exchange) for the same period was -31.33%, the KSCCs are freely transferable, subject to the requirement of NYSE Composite index (New York Stock Exchange) was - Kuwaiti ownership of at least 51% (Law No.15 1960). Both 40.89% while the NIKKEI 225 (Tokyo Stock Exchange) lost KCSs (or IPOs) and KSCCs are defined as listing 42.12% over the same period (Al-Mutawaa, 2009). companies.

9 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

Commercial Companies Law No.15 (1960) 4 Regulation And Supervision identifies KSCs as public offerings securities companies. The procedure for listing KSCs is The capital market in Kuwait is regulated and consistent with the initial public offering (IPO) supervised by three enforcement bodies: the Ministry process documented for most developed markets of Commerce and Industry (MoCI), the KSE Market where the company invites a public subscription prior Committee and the Central Bank of Kuwait (CBK). to listing. In contrast to many other markets, a decree The role of each of these institutions, with respect to must be issued and published in the official gazette, equities, is discussed in this section. announcing the incorporation of the KSC before the The MoCI is responsible for the licensing of company can apply to list. Further, the KSC must market intermediaries and for the regulation and have been established for at least one year prior to supervision of the primary market. More specifically, seeking listing (Law No.15, 1960). the ministry oversees Market Committee decisions, The vast majority of companies listing on the such as listing new companies, listing terms and KSE are KSCCs. In the period from 1997 to 2007, conditions, and company mergers (Al-Jarrah, 2008; for example, only seven from a total of 92 listing Oxford Business Group, 2006b). firms in the non-financial sectors were KSCs (or The Market Committee supervises the IPOs). In contrast to the KSCs that list to raise management of KSE and is responsible for setting the capital, KSCCs do not normally offer shares for general rules and policies for the exchange. Article public subscription immediately prior to listing No.5 (1983) stipulates that the Market Committee (National Bank of Kuwait, 2007; Oxford Business must be organized under the chairmanship of the Group, 2006c)4. The objective of listing a KSCC is to MoCI. Internal rules and regulations on matters switch from private to public ownership via the sale including the structure of the stock exchange, of vendor shares to the public. The original financial regulations, registrations, and KSE shareholders decide the selling price of the shares and membership are the domain of the Market proceeds go directly to the vendors (IPO Monitor, Committee. It is also responsible for issues relating to 2010). KSE dealings, securities, registration of stockbrokers The Commercial Companies Law does not and applications for listing including the inspection require the issue of a decree for the incorporation of of applicant company financial statements. Since its closed companies. Listing KSCCs must, however, inception, the Committee has actively issued meet a specific set of requirements issued by the resolutions regulating the market. Market Committee and issue a prospectus5 that Based on the requirements of the Ameri Decree should be available on the first day of trade. The of 1986, the Market Committee created the Kuwait prospectus should include full sets of audited Clearing Company (KCC) in Resolution No.13 financial statements and the auditor’s reports for the (1987) (Al-Qenae et al., 2002; Annual Economic three years prior to listing (KSE, 2010c). The KSCC Report of the KSE, 1987). All investors and traders usually appoints a listing consultant when applying to must hold an account with the KCC to trade on the have shares traded on the KSE. The consultant is KSE. The KCC clears transactions, registers shares, responsible for filing the company’s legal documents resolves obligations and rights arising from market and assists with the preparation of the prospectus. transactions, and specifies the parties and their The consultant also reviews the financial status of the respective rights for each transaction. The KCC also company, including the company’s compliance with provides a central depository service for listed and accounting regulations (Listing Consultants, Personal non-listed companies’ securities and for domestic and Communication, May 27, 2009). foreign investors. In addition, the KCC provides a A distinction claimed by no other stock range of other services including the distribution of exchange is the KSE trading hall for ladies. The profits and the administration of some IPO Ladies Trading Hall opened in January 2003 with the subscriptions (Al Mohasiboon Magazine, 2009b). objective of creating an adequate environment for The CBK was established in 1969 to set the business women and granting women equal rights to rules of the Kuwaiti financial system (Islam, 2003). It men in terms of managing their investments (Annual controls currency management and the organization Economic Report of the KSE, 2003). It contains the of the banking sector (Law No.32, 1968). The CBK facilities needed for women to closely follow market supervises listed banks, investment companies, and activity. exchange houses (Law No.32, 1968). Mutual funds, however, have been supervised by the KSE Market

4 KSCCs have been required by the Market Committee to Committee since July 2005 (Ameri Decree No.158, offer shares for private subscription if the spread of 2005). The CBK has the power to inspect institutions shareholders is insufficient since 2007. under its supervision at any time for compliance with 5 The prospectus must also include general information its provisions, laws, resolutions, and regulations about the company, the history of the company and its (Central Bank of Kuwait, 2009). It plays a substantial affiliates for the last five years, descriptions of company property, the legal status of any cases filed by or against the role in the supervision of KSE listed companies as company and information on the company's shares and shareholders. (KSE, 2010c).

10 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

around thirty percent of these are banks or investment proposed to better organize the KSE it effectively companies6. weakened oversight and illustrates the divergence Prior to 2011, the KSE was the only market in between the technical and the political functions of the Gulf region without an independent regulatory the Minister of the MoCI (Al-Jarrah, 2008). Further body to manage its capital market (Al Mohasiboon conflict is found in the Market Committee’s Magazine, 2006). The primary market for equities is responsibility to regulate and supervise brokerage supervised by the MoCI while the secondary market firms at the KSE while the MoCI is responsible for is supervised by the Market Committee and the CBK licensing the brokers (International Monetary Fund, supervises financial institutions. The absence of one 2004; Oxford Business Group, 2006b). cohesive regulatory body for the KSE has resulted in As the chief clearing and settlement institution, conflict between the regulators and there are many the KCC works under the umbrella of the KSE. historical examples of the regulators being unable to Meanwhile, the KCC also operates as the securities fulfil their duties. Both domestic and international depository and registry. These two functions should observers agree that the KSE suffers from a weak be separated (International Monetary Fund, 2004). In governing structure and that its management and most other exchanges, securities must be deposited regulatory framework contains significant flaws as a with the clearing and settlement institution, while the result of diverse laws and regulating agencies cash transfer must be made directly through financial (Aldaihani and Aldeehani, 2008; Bouresli, 2009; institutions (Kuwait Transparency Association International Monetary Fund, 2004). The KSE Report, 2006). Therefore, it is inappropriate for the enforcement agencies issue numerous interrelated KCC to act as an investment custodian for stock regulations that are practically impossible to track while holding the settlement cash. Moreover, the and thus undermine consistent enforcement KSE owns 27.5% of the KCC (Kuwait Transparency (International Monetary Fund, 2004; Oxford Association Report, 2006). This level of ownership Business Group, 2006b). has the capacity to bias KCC decisions and The practical functioning of the Market demonstrates that the KCC does not have the Committee and the KSE illustrates the confusion requisite independence to duly fulfil its tasks (Kuwait among KSE regulatory agencies. The Market Transparency Association Report, 2006). Committee, charged with managing the detailed In brief, Kuwait’s system is unique and affairs of the KSE, is presumed to oversee the KSE. eccentric compared to other regulatory systems for In practice, however, the Market Committee and the stock markets (Kuwait Transparency Association KSE are effectively a single entity with dual Report, 2006). Most other stock markets have a functions and responsibilities (Bouresli, 2009; Capital Market Authority, which has the executive International Monetary Fund 2004; Kuwait Chamber ability to meet its responsibilities and the full powers of Commerce and Industry, 2006). While the KSE to develop and regulate the market. The and the Market Committee share the same establishment of the CMA may achieve strengthening management body, it will remain difficult for the of the securities regulatory framework for the KSE to Market Committee to uncover and address violations international standards. The next section documents and shortcomings of the KSE (Al-Jarrah, 2008). The the historical development of the listing requirements current practice results in a clear conflict of interest of the KSE and their evolution. and in a division of power that precipitates weak surveillance, inconsistent enforcement, and 4.1 Evolution of the listing requirements contradictory regulations (International Monetary Fund, 2004). The second Kuwait Conference on This section presents an overview of the legislation Transparency in 2008 concluded that the Market and regulations specifically relevant to the official Committee had not been granted the requisite power and parallel markets over the period from 1983 to and responsibility to act as an independent regulatory 2010. This period captures two critical developments agency and to oversee the development of a securities – the Ameri Decree in 1983 which reorganised the market that is efficient, fair, and transparent. It was, KSE following the Al-Manakh crisis and the therefore, recommended that the Market Committee legislation that enables the establishment of the should be separated from the KSE and operate with Capital Market Authority in 2011. its own resources, staff, and authority (Oxford Article No.6, item No.3, of the Ameri Decree Business Group, 2006b). (1983) stipulates that the “Market Committee is Conflict among agencies regulating the KSE is responsible for setting the general rules and policies institutionalised by Article No.5 of the Ameri Decree for the KSE; in particular it shall set the rules and (1983). This article declares that “the Stock procedures for enrolling brokers and listing shares of Exchange shall be managed by a committee, to be Joint stock companies and any other securities in the constituted under the Chairmanship of the Minister of market.” Market entrance guidelines for both the Commerce and Industry.” While this legislation was official and parallel markets are set by the Market Committee. Activity on the parallel market has 6 Of the 185 KSE listed companies as at March 2010, 9 were decreased substantially since the Al-Manakh crisis banks and 49 were investment companies.

11 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

discussed earlier. In December 2011, only 14 activity of at least 5% of its paid-in capital for each companies are listed on the parallel market compared of the two years prior to listing. with 216 on the official market. Listing regulations Resolution No.1 (1997) gave the Market for the official market are discussed in section 4.1.1 Committee the right to exempt or reject any company and those for the parallel market follow in 4.1.2. from listing without justification. The Market Committee strengthened the listing requirements 4.1.1 Listing Requirements for the Official Market once again by replacing Resolution No.1 (1997) with Resolution No.3 in 2004 and Resolution No.7 in Resolution No.1 (1984) was the first following the 2005, respectively. While Resolution No.3 (1998) reorganization of the KSE in 1983. It instituted the identified that ‘a sufficient’ number of shareholders listing requirements and it provided for their was required for listing, Resolution No.3 (2004) enforcement in the Official Market. Consisting of quantified the minimum number of shareholders only four articles, it required that a company seeking required to list a company for the first time. listing on the KSE have a minimum paid-in capital of The introduction of the ‘strategic’ shareholder KD 5 million, have been established for at least three and ‘lock-up’ requirements (Resolution No.7, 2005) years prior to listing and have obtained a profit of at followed a period of rapid growth in the number of least 5%7 over these three years. In addition, it companies listed on the KSE.8 A strategic empowered the Market Committee to exempt some shareholder was defined as “the one who owns, companies from the listing requirements based on the directly or indirectly, 5% or more of a company's nature of their activities and purposes. This capital” (Resolution No.7, 2005). The total legislation instituted paid-in capital, minim firm age proportion of shares held by strategic shareholders in and a profit requirement as the foundations for KSE a company seeking listing could be no less than 25% listing requirements. of the company's capital, whether owned by one or The provision of a prospectus (including the more strategic shareholders. Further, to protect new company’s history and financial status duly investors by guaranteeing the continuing authorized by the company’s management and participation of firm insiders (or share vendors) after external auditor) to the Market. going public, restrictions on shares were imposed. Committee was first required by Resolution Lock-ups are normally defined as agreements No.4 (1988) . Interestingly, Resolution No.4 (1988) made by insiders of stock-issuing firms to abstain specifies that trading in the company’s shares from selling shares for a specified period of time commence at book value. The first trading price can after the issue (Brau et al., 2005). However, lock-ups be different to book value if approval is provided by in Kuwait are mandated by law, so the term ‘lock-up the Market Committee. restriction’ is used here to differentiate these from the The discussion in the remainder of this section voluntary lock-up agreements prevalent in other relates to Table 1 which provides a summary of the markets. Resolution No.7 (2005) required that all listing requirements from 1984 to 2011. Released listing KSCCs retain 25% of the company’s capital, during Kuwait’s rehabilitation and reconstruction specifically the strategic shareholders’ shares, at the period following the Gulf War, Resolution No.1 clearinghouse. Figure 1 shows the lock-up restriction (1993) revoked Resolution No.1 (1984). The main periods and the proportion of strategic shareholder effect of this new legislation was to reduce the shares that can be offered to the market at the barriers to listing. Resolution No.1 (1993) integrated expiration of each. the parallel and official markets and reduced the In addition to being a listing requirement, the required level of paid-in-capital to KD 1 million. The lock-up restriction in the Kuwaiti setting is unique in pre-listing profit requirement was softened to the two further ways. Firstly, there are three fixed constraint that companies seeking listing not have expiration periods, after which strategic shareholders losses in the previous year’s financial statements. are allowed to dispose of their shares. The first Following post-war reconstruction and partial expiration period is after the first listed year, the economic recovery, Resolution No.1 (1997) again second is after the second listed year and the final tightened the KSE listing requirements. One of the expiration period is after the third listed year. most important changes was the requirement for at Secondly, a specific maximum percentage of shares least KD 2 million share capital and at least KD 3 can be disposed of in each period. Fifty-percent of million shareholders' equity. Resolution No.1 (1997) the total restricted shares can be disposed of at the does not elaborate on the difference between ‘capital’ first expiration period, twenty-five percent at the and ‘shareholders’ equity’ so this presumably relates second expiration period while the remaining 25 to an implicit requirement that listing companies percent cannot be sold until the third expiration have retained earnings on the balance sheet. The pre- period. listing profit requirement was reintroduced and specified as net profit from the company’s main 8 There were thirteen new listings on the KSE in 2003, nineteen in 2004, thirty-three in 2005, and twenty-one in 7 Resolution No.1 (1984) did not define 5% profitability. 2006.

12 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

Table 1. Evolution of the Official Market listing requirements for KSCCs – 1984 to 2010

Legislation Capital requirement Profit requirement Additional requirements Resolution No. 1 (1984) 4 million KD No less than 5% profit for three - Companies applying for listing must be established for three years prior to years prior to listing the application Resolution No. 4 (1988) 5 million KD Not less than 6% of operational - Distributed cash dividend of at least 5% during the year prior to listing profit for three years prior to - At least three audited annual financial statements issued prior to listing listing - Prospectus prepared for approval by company management and external auditor - Shares traded based on book value or other value approved by the Market Committee Resolution No. 1 (1993) 1 million KD No reported losses during the - Prospectus prepared for approval by company management and external financial year prior to listing auditor - Shares traded based on book value or other value decided by the Market Committee Resolution No. 1 (1997) 2 million KD Net operational profit not less - If company has effectively increased its capital, one year must elapse prior than 5% of its paid-in-capital for to listing two years prior to listing -Capital must be distributed among a sufficient number of shareholders; if not, the Market Committee may require the company to offer 25% of its capital for private subscription - Prospectus prepared for approval by company management and external auditor Resolution No. 3 (1998) 2 million KD Average operational profit not - Shareholders’ names and ownership percentages to be submitted Amendment to Resolution No. 1 less than 5% of paid-in-capital for immediately after Market Committee listing approval (1997) two years prior to listing - Complete listing procedures within 45 days of listing approval Company must have reported operational profit for the financial year prior to listing Resolution No. 3 (2004) 3 million KD Average operational profit not - If company has effectively increased its capital, one year must elapse prior less than 7.5% of paid-in-capital to listing for two of the three years prior to - If capital is 3 million KD, there must be at least 150 shareholders, each listing with 20,000 shares representing in total at least 20% of the paid-in-capital

13 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

Legislation Capital requirement Profit requirement Additional requirements Resolution No. 7 (2005) 3 million KD - A strategic shareholder is one who owns, directly or indirectly, 5% or Amendment to Resolution No. 3 more of a company's capital shares (2004) - A strategic shareholder in a company seeking listing must hold at least 25% of the company's capital, whether owned by one or a group of strategic shareholders - Lock-up restrictions are imposed. 50% of the strategic shareholder’s shares must not be sold during the first year of listing, 25% during the second year of listing Resolution No. 1 (2007) 10 million KD Not less than 7.5% of paid-in- - If a closed company has increased its capital by more than 50%, one year capital for each of the two years must elapse from the date of notice in the commercial registry to listing prior to listing - Companies must offer 30% of their capital for private subscription and the offer is to be managed by a specialized company that must be independent from the company seeking listing - Shareholders’ equity to be not less than 115% of paid-in capital for each of the last 3 years - Strategic shareholders must hold at least 25% of the company's capital, whether owned by one or more strategic shareholders - 50% of the strategic shareholder’s shares must be retained for the first year after listing, 25% for the second year after listing Resolution No. 2 (2008) 10 million KD Not less than 7.5% of the - If a closed company has increased its capital by more than 50%, one year weighted average of the paid-in- must elapse from the date of notice in the commercial registry to listing capital for each of the last two - 30% of a company’s capital must be distributed among a sufficient years number of shareholders which is specified by the Market Committee. If the percentage is not achieved, the company must offer 30% of its capital for private subscription with the offer managed by a specialized company - ratio of paid-in capital to shareholders’ equity to be not less than 115% of weighted average paid-in capital for the last 2 years - 25% of the company’s capital must be retained at the Kuwait clearing company for two years after the day of listing

14 Risk governance & control: financial markets & institutions / Volume 1, Issue 3, 2011

The concept of strategic shareholders continues capital of the listing company be retained at the for companies trading in the parallel market clearinghouse for two years from the date of listing. (Resolution No.2, 2007). It was, however, abandoned Thus, the contribution of shares required for the for the official market by Resolution No.2 (2008) and restriction can now be met with smaller parcels of replaced by the requirement that 25% of the paid-in shares.

Figure 1. Lock-up expiration periods at the KSE – November 27, 2005 to November 6, 2008

From 2007 onward, companies seeking listing Lock-up restrictions were imposed then altered while must offer 30% of their capital to the market the concept of strategic shareholdings was introduced (Resolution No.1, 2007). These offers are managed then removed. The most volatile listing requirement by specialized companies that must be independent of for the KSE is for pre-listing profits. The pre-listing the listing company. In addition, shareholders’ equity profit requirement started at 5% in Resolution No.1 must be no less than 115% of the paid-in capital in (1984), increased to 6% in Resolution No. 4 (1988), each of the last three years. The most recent alteration changed to zero in Resolution No.1 (1993), and was to listing requirements occurred in 2008 with reintroduced at 5% in Resolution No.1 (1997). From Resolution No. 2. This resolution altered the 2004 until 2011, the pre-listing profit requirement calculation of the base percentage of the total remained stable at 7.5%, with the only change being shareholders’ equity and pre-listing profit from paid- to the base used for calculating the required profit in capital to weighted average paid-in capital. The percentage. ratio of paid-in capital to total shareholders’ equity It has been argued that the organization and was also adjusted to 115% of the weighted average of supervision of the KSE has been largely neglected by the paid-in capital in the past two years. the relevant authorities (International Monetary Fund, In contrast to Resolution No.1 (2007), which 2004). Few countries in the world lack obligates companies seeking listing to offer 30% of comprehensive legislation to organise their stock their capital for private subscription, Resolution No.2 markets but Kuwait’s law facilitating the (2008) requires that 30% of the company's capital be establishment of the CMA is very recent. The distributed among a number of shareholders specified regulatory climate and framework for the KSE is by the Market Committee. If this percentage is not generally inconsistent with the conditions needed to available, the company must offer 30% of its capital achieve the objectives and principles advocated by for private subscription through a specialized the International Organization of Securities company independent from the company seeking Commission (International Monetary Fund, 2004). listing Since its inception, the KSE can only be In summary, the listing requirements of the described as unstable (Aldaihani and Aldeehani, Official Market have been introduced, removed or 2008). This instability can be attributed, in part, to the modified frequently over relatively few years. During process of issuing regulations in response to the the period from 1984 to 2011, the capital requirement political and market crises outlined in section 2 rather was increased, reduced and then increased again. than the early adoption of a coherent regulatory

15 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

framework for the KSE. Al-Nefeesi (2008) argues In 1993, the Market Committee again combined that the KSE has more gaps and deficiencies than do the parallel and official markets with a single set of other emerging markets and that the KSE’s listing requirements (Resolution No.1, 1993). Kuwait development has been hampered by the lack of an continued with one market for seven years until appropriate legal and institutional framework. The Resolution No.4 (2000) re-established the parallel next section will discuss and review the listing market (Annual Economic Report of the KSE, 2000). requirement of the parallel market. The listing requirements for the Parallel Market are summarised in Table 2. 4.1.2 Listing requirements for the parallel market Resolution No.1 (2003) allowed parallel market companies to transfer to the official market once they As discussed in section 2.2, the Al-Manakh market had been listed for one year on the parallel market facilitated trading of GCC companies and KSCCs that and met the official market listing requirements. did not meet the listing requirements of the official Resolution No.2 (2007) introduced a profit market. Following the reorganization of the KSE in requirement for parallel market companies, increased 1983, the Parallel Market was established to replace the required level of capital and mandated that 25% the Al-Manakh market (Annual Economic Report of of the company’s capital, specifically the strategic the KSE, 1988). GCC companies were admitted to shareholders shares, be retained at the clearinghouse. the official market in May 1989 when it was opened Two fixed expiration periods were introduced in to investors who were citizens of the GCC. Cross- Resolution No.2 (2007). Strategic shareholders are listing of GCC companies was also permitted from permitted to dispose of 25% of total restricted shares this time (Annual Economic Report of the KSE, during the second year of listing while the remainder 1988). Following a brief closure, the market was can be sold during the third year of listing. reopened in 1989 with a new set of listing Regulations for the parallel market have requirements that permitted GCC and Kuwaiti undergone numerous reforms. During the past three companies that did not meet the requirements of the decades, this market has been thrice activated, once official market to trade (Annual Economic Report of suspended, and once merged with the Official the KSE, 1989). Market. Listing requirements have been less onerous Resolution No.6 (1989) included the than those of the official market, especially with requirement for a minimum paid-in-capital of KD 1 respect to the pre-listing profit and capital million to trade on the parallel market compared to requirements (Resolution No.1, 1997; Resolution KD 5 million on the official market. The amount of No.2, 2007; Resolution No.34, 2000). As a result, the historical financial information was also lower with parallel market contains higher-risk companies than two audited annual financial reports required prior to does the official market (Al-Nefeesi, 2008). At listing whereas three were required for the official present, the parallel market is best described as a market. While there was a 6% pre-listing profit transitory market where companies can list relatively requirement for listing on the official market, there easily prior to transferring to the official market. was no profitability requirement for the parallel market.

Table 2. The Parallel Market: evolution of listing requirements

Legislation Capital Profit Other requirements requirement requirement Established -Only GCC companies admitted 1984 Parallel Market suspended, 1988 (Annual Economic Report of the KSE 1988) Resolution No.6 1 million KD Not required - Kuwaiti and GCC companies are admitted (1989) - At least two audited annual financial statements issued prior to listing - Shares traded based on book value or other value decided by the Market Committee - Company founders (vendors) retain 30% of their shares for two years Parallel Market merged with the official market, 1993 (Resolution No.1 1993) Separate parallel market re-established, 2000 (Resolution No.4 2000)

16 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

Legislation Capital Profit Other requirements requirement requirement Resolution 0.5 million Not required - Shareholders equity is at least 1 million KD No.34 (2000) KD - At least three audited annual financial statements issued prior listing - Minimum of 50 shareholders - Not more than 50% of shares at listing to be sold in the first year of listing Resolution No.2 3 million KD Average profit - Minimum of 50 shareholders (2007) must be at least - If a closed company has increased its capital by more 5% of paid-in- than 50%, one year must pass from the date of notice capital for the in the commercial registry until listing last two years, - Non-Kuwaiti companies must be listed on their own and company domestic exchanges must have - Strategic shareholders must hold at least 25% of the reported net company's capital, whether owned by one or more profit during the strategic shareholder previous two - 50% of the strategic shareholders’ shares cannot be financial years sold in the first year of listing, 25% during the second year of listing - If a closed company has changed its legal structure, a period of three years must pass from the date of notice in the commercial registry before listing

5 Accounting and Auditing Requirements Since 1991, the compliance of listing companies with IFRS has been monitored by two agencies: the An overview accounting and auditing standards and Control Department of the MoCI and the Surveillance practices is provided in this section. The adoption of Department of the KSE (Al-Bannay, 2002; Alanezi, the International Financial Reporting Standards 2006; Al-Shammari et al., 2008). The Control (IFRS) and the regulatory mechanisms for monitoring Department of the MoCI is responsible for compliance with these are discussed next. The ascertaining compliance with the appropriate fiscal absence mandated auditing standards prior to the regulations and to verify compliance with IFRS (Al- adoption of the International Standards of Audit Shammari et al., 2008). Meeting the Control (ISAs) in 2008 and the requirement for two auditors Department’s mandate to monitor, review, and check are discussed in section 5.2. The organisation of the IFRS compliance for every company in Kuwait accounting profession is then addressed in section appears to be unworkable given deficiencies of 5.3. professional qualifications and the experience of the Control Department staff (Alanezi, 2006) and their 5.1 Financial reporting standards and high workloads (Al-Shammari et al., 2008). monitoring compliance The Surveillance Department of the KSE is also legally responsible for monitoring the compliance of Kuwait is considered a pioneer not only among the listed companies with IFRS. Unlike the Control GCC members,9 but also globally for its adoption of Department, the Surveillance Department staff of the IFRS in 1991 (Al-Shammari et al., 2008). Prior to the KSE are technically qualified and monitor adoption of IFRS, however, the accounting standards compliance with a checklist to ensure all required used in Kuwait were most frequently based on those disclosures are made (Al-Shammari et al., 2007). from the United States, members of the European Listed companies must submit their audited financial Union, or other Arab countries (Elshamy and Al- statements to the MoCI and the KSE within three Qenae, 2005; Shuaib, 1978; Shuaib, 1998). The months of the end of the financial year (Resolution accounting and disclosure practices chosen by No.16, 1987). The requirement for quarterly financial companies varied widely, and the financial reports was introduced in 1998. These must also be information disclosed was quite limited. Further, it filed with the MoCI and the KSE within 45 days of was difficult to ascertain which accounting standards the quarter closing date (Al-Wazzan, 2006). were being used by Kuwaiti companies (Al-Bannay, In general, the MoCI and the KSE are 2002). The lack of uniformity in the application of considered separate agencies. Each has its own accounting standards made comparability between procedures for determining compliance with IFRS. the accounts of companies difficult (Elshamy and Al- However, there is a lack of coordination between the Qenae, 2005). two agencies and different enforcement mechanisms are applied by each ( Al-Shammari et al., 2008).

9 Oman adopted IFRS in 1986 (Al-Shammari et al., 2008).

17 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

5.2 Auditing standards contains the conditions and procedures for auditor registration. It also forbids auditors from undertaking No uniform body of regulated or even generally any additional professional activities that are accepted auditing standards were used in Kuwait incompatible with auditing duties including prior to 1998 (Shuaib, 1998). Most auditors consultation unrelated to accounting, bookkeeping, voluntarily used the International Standards of Audit and preparing financial statements or advertising (ISAs), but there was no legal requirement to do so services in a way that is incompatible with the ethics (Listing Consultants, Personal Communication, May of the profession (Shuaib, 1998). Further, an auditor 27, 2009). Some accounting firms based their audits cannot also be a partner, administrator, employee, or on U.S. and U.K. standards while others did not a relative to anyone in the client’s company. appear to follow any particular standards. In some Sanctions for violations of Law No.5 begin with a cases, financial statements were certified without any warning; proceed to prohibition from practicing effective auditing (Shuaib, 1978; Shuaib, 1998). This auditing for a specific period of time; and end with eclectic approach to audits continued until 2008 when removing the auditor’s name from the registry of Ministerial Resolution No.101 required audits be auditors. conducted in accordance with ISAs. This resolution was clearly overdue when one considers that 5.3 Accounting profession in Kuwait Resolution No.18, mandating the use of IFRS, was issued in 1990. The accounting profession in Kuwait is still in the early stages of development and is far from well 5.2.1 Auditor practice established (Shuaib, 1998). The Kuwait Accountants and Auditors Association (KAAA), established in Consistent with regulation in most markets, the 1973, is the accountants’ professional body. The Company Commercial Law (1960) required that KAAA is a member of the International Federation of listed companies be audited. Interestingly, Law No.51 Accountants (IFAC). This body conducts studies, (1994) mandated at least two auditors from separate prepares research reports, provides consultations, firms serving as joint auditors for KSE listed promotes the exchange of experience, supports the companies. Both auditors must be independent of the development of members’ expertise, widens the company being audited and must be registered with practical background of those working in the field of the MoCI. Licensed auditors have been a requirement accounting, and helps regulate auditors’ qualification since 1981 (Law No.5). Licensed auditors are exams (Kuwait Accountants and Auditors accountants who have passed the auditing practice Association, 2010a). It does not, however, have the professional examination (prepared by the MoCI in power to certify accountants and auditors or to set collaboration with Kuwait University), have a accounting and auditing standards (Al-Bannay, 2002; specified amount of audit experience, be Kuwaiti Alanezi, 2006; Shuaib, 1998). The KAAA has been nationals and be registered with the MoCI. criticized for its inability to set accounting standards, Registration of auditors, prior to Law No.5, required its inefficiency, and its slow progress as an only a bachelors degree in business or, in some cases, accounting association (Shuaib, 1998). practical experience only (Shuaib, 1978). To increase the effectiveness of the KAAA, Even with the two auditor requirement for listed Ministerial Resolution No., 291, which concerns the companies, it appears that auditors will authorize rules of ethical conduct for the auditing profession, financial statements that do not reflect the true was issued in 2006. This resolution charges the position of the company and questions about auditor KAAA with monitoring the implementation of, and conflict of interest are not uncommon. The apparent compliance with, ethical rules of conduct for lack of auditor independence was highlighted by the accountants issued by the IFAC. It also gives the Kuwaiti media as one potential cause that worsened KAAA the right to investigate any violation of these the effect of the 2008 global financial crisis in rules of conduct and to report the complaint to the Kuwait. As one Kuwaiti analyst put it, the “financial Minister of MoCI. crisis is seen everywhere but in the financial reports How effective the accounting profession has of the Kuwaiti companies” (Al Mohasiboon been with respect to the application of standards and Magazine, 2009a, p. 38). While the financial reports the auditing of accounts is an open question. There of Kuwaiti companies indicated their strength, the have been numerous incidents where KSE-listed companies themselves were clearly under financial companies “bent” their application of required pressure. accounting principles to serve their own interests The key legislation for the regulation of auditing rather than those of stakeholders (Al-Bassam, 2006; in Kuwait is Law No.5 (1981). Law No.5 addresses Alanezi, 2006). The IMF (2004) highlighted the the general requirements for registering and ambiguous and inadequate disclosure in financial practicing accounting in Kuwait, the rights and duties statements prepared by some KSE-listed companies. of a public auditor, and the penalties that auditors Researchers, analysts, financial reporters, academics, may incur for violation of this law. This legislation and authors of specialized reports argue that the

18 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

accounting information provided by Kuwaiti firms contends that this variability in efficiency levels would be enhanced by more stringent professional indicates that regulatory changes in Kuwait have not training for accountants (Al-Bassam, 2006; resulted in greater market efficiency (Abdmoulah, International Monetary Fund, 2004). They question 2010). the usefulness of the accounting information that Research into the informational efficiency of the KSE-listed companies provide investors and traders KSE indicates that, at best, the market is weak form given relatively low levels of disclosure, a lack of efficient but may be weak form inefficient in different reliability and the inadequate accounting expertise of sample periods. Caution is advised for testing theories preparers. and models that require the assumption of an efficient market. For a security price to reflect the underlying 6 Implications for Research value of the company, price must incorporate relevant information about the performance of the company The institutional setting in Kuwait has a number of and the market’s assessment of prospects for future implications for capital market based research. The performance. regulatory framework and the types of investors who participate in the market result in informational 6.1.2 Share price and company value inefficiency. Many theories in accounting and finance include the impact of variables on firm value and use A second requirement needed to justify the price as the proxy for value. The validity of price as a assumption that price is an unbiased representation of proxy in this context relies on how much information company value is that of insignificant arbitrage costs is impounded in price and the speed at which this (Lee et al., 1999). Short selling, which is a process occurs. While it is reasonable to assume that mechanism for arbitrage, is prohibited on the KSE the trades of informed investors are responsible for (Hassan et al., 2003; Aldaihani and Aldeehani, setting security prices in developed markets, the 2008). Further, KSE investors cannot create synthetic profile of investors participating in the KSE provides short positions as call options are the only exchange- an interesting context for identifying the factors that traded derivatives and these are limited to companies affect price setting when less sophisticated investors approved by the Market Committee (Resolution have a substantial role in the process. Small IPO No.337, 2004). Where arbitrage is not possible, as is sample sizes and weak form efficiency preclude the case on the KSE, prices will only incorporate the typical underpricing studies. However, the listing views of optimistic investors as pessimistic investors requirements for the KSE provide several interesting are unable to take short positions (Miller, 1977). This research opportunities. These are discussed in the argument sits nicely with the KSE’s history of remainder of the paper. speculative booms and busts. Stiglitz (1989) advocates taxes to reduce 6.1 Informational efficiency on the KSE speculative trading in stock markets and in markets (such as Australia) where capital gains tax is levied Antoniou et al. (1997) argue that the tests for on stock market investments, a concessional rate is market efficiency developed for highly liquid markets applied to investments held for more than one year. with strong participation by informed investors Kuwaiti investors do not pay tax (Global Consultants accessing frequent and reliable disclosures are not 2006) and this institutional feature may contribute to appropriate in the emerging market context. Investors the level of speculation in the stock market. in emerging markets may not meet the assumptions of The lack of effective arbitrage mechanisms rationality and risk aversion that underlie the efficient combined with the relatively low levels of publicly markets hypothesis (EMH) and, therefore, non-linear available information (discussed in section 6.2) models of testing for efficiency are more appropriate indicates that results from value relevance studies for emerging markets (Antoniou et al., 1997). While should be interpreted with caution. The impact of a number of studies show mixed results using linear earnings management or accounting policy choice on tests of informational efficiency on the KSE, these price or return, for example, implicitly assumes that will not be discussed here. price reflects value. This assumption cannot be Hassan et al. (2003) use a non-linear approach justified unless the market is at least semi-strong form and report that the KSE does not achieve the weak- efficient and the KSE has a way to go before this form level of market efficiency. Therefore, the level of efficiency can be achieved. history of past prices for a stock is not fully reflected in the current price. Hassan et al. (2003) do, however, 6.1.3 Investor participation note that efficiency improves in the latter part of the 1990s. Also employing a non-linear approach, Bley and Chen (2006) argue stock markets in the Abdmoulah (2010) examines changes to the level of Middle East have remained virtually invisible to efficiency over time and concludes that the KSE is global investors. Restrictions that these markets have generally weak-form inefficient but that the level of imposed on foreigners, their lack of common efficiency varies across the sample period. He accounting standards, business transparency, and

19 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

economic and political uncertainty are features that increasing (decreasing) accounting change is foreign investors do not find attractive (Bley and announced. A mixed methods approach using Chen, 2006). With the exception of GCC citizens, the quantitative analysis of stock prices and accounting KSE was closed to international investors until 2000. announcements and qualitative analysis with brokers The mean level of foreign investor shareholdings in at the exchange could resolve the issue. Kuwaiti companies was around one percent in 2002 (Al-Shammari et al., 2008). Further, the majority of 6.2 Information availability traders (over 60%) on the KSE are individuals rather than institutional investors (Abdmoulah, 2010). Non-disclosure is considered a key deficiency of the Individual traders tend to be less informed, more KSE. To provide transparency for stakeholders, subject to behavioural biases and face higher information that affects the financial position of a transaction costs (Masih et al., 2010). Al-Shammari company must be disclosed through appropriate et al. (2008) confirm the dominance of individual official and legal channels then be released to the investors on the KSE and their data show the mean market in a timely fashion. While disclosure of legal level of institutional ownership in Kuwaiti companies disputes, for example, is mandated for KSE listed is thirty-nine percent. companies it has become standard practice for the Al-Bassam (2006) contends that the volume of KSE regulators and other stakeholders to read such speculative trades on the KSE reduces the importance information in the media prior to its official disclosed of accounting as an information source for investors. (Kuwait Transparency Association Report, 2006). As Given the informational inefficiency of the KSE, it is a result, shareholders may receive vital financial unlikely that investors could rationally expect to be information only after the share price has changed rewarded for fundamental analysis as there is no and deals have been completed by insiders (Al- demonstrated linkage between price and value. In Nefeesi, 2008). The injunction issued against Mobile their survey of Kuwaiti investor perceptions of Telecommunication Company (MTC) (now Zain) on information usefulness, Nasser et al. (2003) do find April 19, 2006 (Kuwait Transparency Association that investor groups rate financial statements as an Report, 2006, p. 13) provides an example. MTC did important source of information. However, the not disclose the injunction to the KSE so the financial statements rate after ‘other information exchange could not disseminate the information. obtained directly from the company’ for individual Investors were astonished when the verdict from the investors, while stock market brokers rate market case was published in newspapers several weeks rumours as more important than financial statements later. MTC shares were not subject to a trading halt (Nasser et al., 2003). They argue that the importance and ensuing trades allowed insiders to profit from of rumour can be attributed, in part, to the fact that trading on the undisclosed information. brokers and institutional investors spend their Without the enforcement of continuous working days in the exchange building. disclosure requirements for the KSE, researchers will The capacity for market manipulation is experience difficulty determining when specific enhanced where rumours are regarded as a primary pieces of information became available to market source of information. Both market manipulation and participants. This feature has serious implications for insider trading are serious problems on the KSE research questions that are typically answered with an (International Monetary Fund, 2004). In Kuwait, events study methodology – determination of ‘event’ Board members cannot be held accountable for dates would necessarily be arbitrary. While the KSE selectively ‘sharing’ confidential information which has embraced an electronic trading system, the has the capacity to affect share prices and trading electronic dissemination of information to market volume (Kuwait Transparency Association Report, participants lags well behind that observed in 2006). Foreign investors will not find the KSE an developed markets. Announcements arising from attractive opportunity for diversification if their continuous disclosure requirements are released on investment is likely to result in a transfer of their the KSE announcement board for the benefit of wealth to corporate insiders. The number of informed investors in the exchange building but are not stored investors is, therefore, constrained by the on the KSE website. characteristics of the KSE. The KSE began provide annual reports for listed The KSE provides an excellent context for companies electronically in 2008. Prior to this date, testing functional fixation on accounting information the majority of annual reports were available only in releases (c.f. Kaplan and Roll, 1972; Hand, 1990). hardcopy and had limited distribution. The sampling We would expect that a large pool of relatively timeframe for value relevance research would unsophisticated investors using accounting necessarily be constrained to the post 2008 period. It information would be fooled by cosmetic changes in would require a heroic leap of faith to assume that the financial statements. The importance of rumour as information in the financial statements had been source of information, however, provides a ‘instantaneously’ available to the market prior to potentially confounding variable if a negative (or 2008. However, Kuwaiti companies willingly supply positive) rumour is circulating when a profit

20 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

annual reports when requested (Nasser et al., 2003; and guidelines, investors and researchers will have Al-Shammari and Al-Sultan, 2010). difficulty grasping the implications of these brief publications which are not clearly explained. This 6.3 Listing companies lack of clarity became painfully clear to the authors of this paper as we tried to collate and interpret the As discussed in section 3, the majority of new listings regulations governing the KSE across time. on the KSE are KSCCs that do not list to raise new Therefore, we expect the summary of regulations capital and shares are offered at book value or ‘other provided in this paper will ease the path for those approved’ value. Therefore, there is little capacity for interested in capital market research in the Kuwaiti IPO underpricing studies given the very small sample context. sizes. Listing companies make hard copy In conclusion, theories based on the efficient prospectuses available in the exchange on listing day. markets hypothesis will be less able to explain Prospective investors wishing to purchase shares on changes in share prices in the Kuwaiti setting. While the first trading day would be subject to extreme this constrains the application of some theories, information asymmetry as they would have very institutional features offer opportunities for research limited time to process the prospectus information. In that could not be conducted in other markets. The addition, the somewhat arbitrary nature of the importance of individual investors in price setting and application of listing requirements was flagged as a the information they can access when making major issue in the Kuwait Transparency Association investment decisions suggests tests of behavioural Report (2006). The Market Committee has approved theories as a potentially fruitful area for future companies that do not meet the listing requirements research using KSE data. for admission and it is difficult to see how the prospective investors could identify such occurrences References given the short time they have to analyse prospectus data for trading on listing day. 1. Abdmoulah, W. (2010) “Testing the evolving The two auditor requirement of Law No.51 efficiency of Arab stock markets”, International (1981) applies only to the financial statements of review of financial analysis, Vol.19 No.1, pp. 25-34. listed companies, not to those seeking admission to 2. Al-Atrabi, M. and Al-Sayed, M. (2008), "The court the list. Thus, for the historical financial statements suspends the KSE", Al-Quabas, 14/11/2008. 3. Al-Bannay, A. (2002), Reception among accountnats, presented in prospectuses, it would be prudent to auditors and users of IAS in preparing annual spend more rather than less time analysing this accounts: The case of Kuwait, City University information. By implication, the two auditor 4. Al-Bassam, S. (2006), Toward developing the role of requirement suggests that a single auditor is accounting information to rationalize the decisions of insufficient in the Kuwaiti context. This feature casts dealers in the market, Kuwait Chamber of Commerce doubt on the usefulness of pre-listing financial and Industry. statements and provides scope for further research. 5. Al-Hashel, M. (2003), Overreaction, The introduction of lock-up restrictions and heteroscedasticity, and spillovers in stock returns: Evidenc from the Kuwait Stock Exchange, Old changes to these outlined in section 4.1.1 provide an Dominon University. exciting opportunity for listing company research in 6. Al-Jarrah, T. (2008), Reform of the services of the Kuwait. While extant research indicates weak-form Ministry of Commerce and Industry, viewed 19 March inefficiency on the KSE, investigation of trading 2010, < http://www.transparency-kuwait.org/index. volumes at lock-up expiration may provide insights php?pid=25 >, Kuwait city. for the use of signalling by founders in the post- 7. Al-Mutairi, A. (2004), Empirical testing of various listing period. The frequent changes to the pre-listing aspects of investment in the Kuwait Stock Exchange profit and capital requirements provide opportunities (KSE), Cardiff University of Wales. for research to test the efficacy of these regulations. 8. Al-Mutawaa, F. (2009), The global economic crisis One interesting empirical question would be ‘do and its impact on the state of Kuwait, Kuwait Transparency Association, Kuwait. listing companies that meet the profit requirements in 9. Al-Nefeesi, N. (2008), Kuwait Stock Exchange: periods when these are more stringent perform better Problems and solutions, Kuwait Chamber of in the long run?’ Commerce and Industry. 10. Al-Qenae, R. (2000), Financial Information and 6.4 Incoherent regulatory framework regulation in an emerging market: Empirical study of the Kuwait Stock Exchange, University of Esesex. The Kuwaiti regulators continue to promulgate 11. Al-Qenae, R., Li, C. and Wearing, B. (2002), "The regulations that do not meet the standard required for information content of earnings on stock prices: The ISOCO membership (Bouresli, 2009). The Market Kuwait Stock Exchange", Multinational Finance Journal, Vol.6, No.3&4, pp. 197-221. Committee, MoCI, and CBK have issued a large 12. Al-Shal Report (2008), "Suspending the KSE", Al- number of regulations over the last three decades and Quabas, 16/11/2008. these are difficult to track. Although the KSE website 13. Al-Shammari, B., Brown, P. and Tarca, A. (2008), includes some regulations, administrative decisions, "Developement of enforcement mechanisms following

21 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

adoption of International Accounting Standards in the European Financial Management, Vol.3 No.2, pp. Gulf Co-Operation Council Member States", The 175-190. International Journal of Accounting, Vol. 43, No. 4, 36. Bley, J. and Chen, K. H. (2006), "Gulf Cooperation pp. 425-47. Council (GCC) Stock Markets: The dawn of a new 14. Al-Shammari, B. and Al-Sultan. W. (2010) "Corporate era", Global Finance Journal, Vol.17, pp. 75-91. governance and voluntary disclosure in Kuwait", 37. Bouresli, A. (2009), "An analytical study of the International journal of disclosure and governance, organizational supervising structure of the Kuwaiti Vol.7 No.3, pp. 262-280. financial market in the light of the ISOCO principles 15. Al-Sultan, F. H. (1989), "Averting Financial Crisis- and the World Bank report", Journal of the Gulf and Kuwait", The World Bank, Working Paper. Arabian Peninsula studies, Vol.35 No. 133, pp. 43- 16. Al-Wazzan, B. (2006), "The business guide, 131. accountancy: Rapid exchange. Accounting matters:, 38. Brau, C., Lambson, E. and McQueen, G. (2005), Annual Business Economic and Political Review, "Lockups Revisited", Journal of Financial and Vol.1, pp. 177-81. Quantitative Anaysis, Vol.40 No.3, p. 13. 17. Al-Yaqout, A. (2006), The usefulness of quarterly 39. Central Bank of Kuwait (2009), Supervision and financial reports to Kuwaiti financial users in the financial units, viewed 17 March 2010, Kuwait Stock Exchange, Loughborough University. . 18. Al Mohasiboon Magazine (2006), Kuwait Accounting 40. Eiman, A. (2010), "Capital Market Authority Law", and Auditing Association Seminar, viewed 17 March Al-Quabas Newspaper, 05/02/2010. 2010, 41. Elshamy, A. M. and Al-Qenae, R. (2005), "The change . in the value-relevance of earnings and book values in 19. Al Mohasiboon Magazine (2009a), The crisis equity valuation over the past 20 years and the impact revealed: Accounting mistakes made by external of the adoption of IASs: the case of Kuwait", auditors just to satisfy the companies management, International Journal of Accounting, Auditing and viewed 25/03/2010, Performance Evaluation, Vol.2 No. 1/2, pp. 153-65. . 42. Global Consultants (2006), Doing Business in Kuwait, 20. Alanezi, F. S. (2006), An empirical study of Kuwaiti Baker Tilly International. Joint-Stock companies' compliance with the 43. Gomber, P., Lutat, M. and Schubert, S. (2008) "Capital international financial reporting standards, University markets in the gulf: international access, electronic of Newcastle. trading and regulation", in Handbook on information 21. Aldaihani, M. and Aldeehani, T. (2008), "Portofolio technology in finance, Seese, D., Weinhardt, C and optimization models and a tabu search algorithm for Schlottmann, F. (Eds.) Springer-Verlag, Berlin the Kuwait Stock Exchange", Investment Management 44. Hand, J. (1990) “A test of the extended functional and Financial Innovations, Vol.5 No.2, pp. 31-40. fixation hypothesis” Accounting Review, Vol.65 No.4, 22. Alsalman, E. (2002), Empirical issues of financial pp. 740-763. market volatility in Kuwait Stock Exchange, Howard 45. Hassan, K., Al-Sultan, W. and Al-Saleem, J. (2003) University. “Stock market efficiency in the Gulf Cooperation 23. Ameri Decree (1983), Organizing the Kuwait Stock Council Countries (GCC): the case of Kuwait”, Exchange, viewed 02 March 2009, Scientific Journal of Administrative Development, . 46. International Monetary Fund (2004), Kuwait: 24. Ameri Decree No.20 (2000), Direct Foreign Financial Sector Assessment Program-Detailed Investment Law. assessment of observance of standards of codes- 25. Ameri Decree No.158 (2005), Amending the Ameri International Organization of Securities Commission Decree of 1983 (IOSCO)-Objectives and principles of securities 26. Annual Economic Report of the KSE (1987), The regulations, viewed 19 March 2010 Kuwait Stock Exchange. , IMF 27. Annual Economic Report of the KSE (1988), The Country Report No.04/352. Kuwait Stock Exchange. 47. International Monetary Fund (2005), Kuwait:2005 28. Annual Economic Report of the KSE (1989), The article IV consultaiton-staff report; staff statement; Kuwait Stock Exchange. and public information notice on the executive board 29. Annual Economic Report of the KSE (1990/1991), The discussion, viewed 30/03/2010, < Kuwait Stock Exchange. http://www.imf.org/external/country/syr/index.htm?ty 30. Annual Economic Report of the KSE (1995), The pe=9998>, International Monetary Fund. Kuwait Stock Exchange. 48. IPO Monitor (2010), Secondary offerings, viewed 31. Annual Economic Report of the KSE (2000), The 2/04/2010, . 32. Annual Economic Report of the KSE (2003), The 49. Islam, M. (2003), "Regulations and Supervision of Kuwait Stock Exchange. Financial Institutions in GCC Countries", Managerial 33. Annual Economic Report of the KSE (2004), The Finance, Vol.29 No. 7, pp. 17-42. Kuwait Stock Exchange. 50. Kaplan, R. and Roll, R. (1972) “Investor evaluation of 34. Annual Economic Report of the KSE (2005), The accounting information: some empirical evidence”, Kuwait Stock Exchange. Journal of business, Vol.45 No.2, pp. 225-257. 35. Antoniou, A., Ergul, N. and Holmes, P. (1997) 51. KSE (2010a), The KSE history, viewed 02 April 2010, “Market Efficiency, Thin Trading and Non-linear .

22 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

52. KSE (2010b), Companies listed on the KSE, viewed 24 corporate reporting", Managerial Auditing Journal, March 2010, < http://www.kuwaitse.com/PORTAL/A Vol.18 No.6/7, pp. 599-617. /Stock/Companies.aspx>. 65. National Bank of Kuwait (2007), Doing Business in 53. KSE (2010c), Listing documents, viewed 02 April Kuwait, viewed 24 March 2010, 2010, < . 66F04A6F3622/0/doingbusinessNEW.pdf>. 54. Kuwait Accountants and Auditors Association 66. Oxford Business Group (2006a), "The Economy: Time (2010a), Kuwait Accountants and Auditors Association for Change", Annual Business Economic & Political (KAAA) objectives, visions and values, viewed 12 Review: Kuwait, Vol.1, Emerging Kuwait, pp. 37-42. March 2010, Fired Growth", Annual Business Economic & political 55. Kuwait Chamber of Commerce and Industry (2006), Review, Vol. 1, Emerging Kuwait, pp. 75-81. The Kuwait Stock Exchange and the challenge of 68. Oxford Business Group (2006c), "The Business Guide, international islamic financial market growth and Accountancy: Rapid Exchange", Annual Business developement. Economic & political Review, Vol. 1, pp. 177-81. 56. Kuwait Transparency Association Report (2006), The 69. Resolution No.1 (1993), The KSE listing requirements. deficiencies of the KSE and recommendations, viewed 70. Resolution No.1 (1997), The KSE listing requirements. 19 March 2010, , Kuwait Transparency 72. Resolution No.3 (2004), The KSE listing requirements. Society, Kuwait. 73. Resolution No.4 (1988), The KSE listing requirements. 57. Law No. 5 (1981), Practicing the auditing profession. 74. Resolution No.2 (2007), Listing and trading of 58. Law No. 15 (1960), Commercial Companies law. shareholding companies shares in the Parallel Market. 59. Law No. 32 1968, Currency, the Central Bank of 75. Resolution No.3 (2004), KSE listing requirements. Kuwait and the organization of banking business 76. Resolution No.13 (1987), Procedures for the 60. Lee, C., Myers, J. and Swaminathan, B. (1999) "What registration, settlement and ownership transfer of is the intrinsic value of the Dow?" Journal of Finance, shares listed on the market. Vol.54 No.5 pp. 1693-1741. 77. Resolution No.16 (1987), Financial Statements of 61. Masih, M., Alzahrani, M. and Al-Titi, O. (2010) listed companies. "Systematic risk and time scales: New evidence from 78. Resolution No.34 (2000), The KSE listing an application of wavelet approach to the emerging requirements: The Parallel Market. Gulf stock markets", International review of financial 79. Shiller, R. (2008) The subprime solution, Princeton analysis, Vol.19 No. 1, pp. 10-18. University Press, N.J. 62. Miller, E. (1977) "Risk, uncertainty and divergence of 80. Shuaib, A. (1978), "The middle east: Accounting in opinion", Journal of Finance, Vol.32 No.4, pp. 1151- Kuwait and a banker's view of business opportunities 68. in the entire area", Journal of Accountancy, No. 146, 63. Moosa, I. (2010) "Stock market contagion in the early pp. 146-74. stages of the global financial crisis: the experience of 81. Shuaib, A. (1998), "Evolution of accounting standards the GCC countries", International journal of banking in Kuwait", Industrial Bank of Kuwait (IBK) papers, and finance, Vol.7 No.1, pp. 19-34. Series No. 53. 64. Naser, K., Nuseibeh, R. and Al-Hussaini, A. (2003) 82. Stiglitz, J. (1989) “Using tax policy to curb speculative "Users' perceptions of various aspects of Kuwaiti short-term trading”, Journal of financial services research, Vol.3 No. 2-3, pp. 101-115

23 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

IS INSIDER TRADING REGULATION EFFECTIVE? EVIDENCE FROM UK TAKEOVER ACTIVITY

Brendan Lambe*

Abstract

Analysed in this study are the returns on stock prices of target companies surrounding the first publicised dates of completed takeovers in the UK between 2001 and 2010. Two samples are created of 209 and 197 firms for announcement and rumoured dates respectively. Both demonstrate statistically significant cumulative abnormal returns (CARs) prior to the release of information about the impending bid. This paper investigates whether observable factors create this price run-up or if it is the result of disclosed insider trading. Cross sectional analysis of CARs do not corroborate the claim that reported informed trades are the cause of this effect, this may indicate that trading on material non public information goes undisclosed.

JEL Classifications: G18, G28, G14

Keywords: Insider trading, Takeovers, Stock Market, Regulation

* University of Leicester, School of Management, Ken Edwards Building, University Road, Leicester LE1 7RH, UK Tel.: +44 (0) 116 229 7420 E-mail: [email protected]

1 Introduction act as an enforcer. This ‘informed trading hypothesis’ is not in receipt of universal support as several studies Since assuming the guardianship of market integrity point to the influence of publicly discernible signals surprisingly little research has been carried out to on abnormal pricing behaviour prior to an gauge the success of the UK’s Financial Services announcement of corporate restructuring events Authority (FSA) in meeting its enforcement (Jarrell and Poulsen 1989, Neely 1987). Much responsibilities. The role of protecting the markets empirical literature documents evidence suggesting from abuse of various kinds was assumed by the FSA that certain observable changes in company following the introduction of the Financial Services characteristics can demonstrate the increased and Markets Act 2000. probability that this firm could become a likely target Empirical studies that examine instances of for a takeover (Hasbrouck 1985, Schleifer and Vishny abuse focus upon known cases of insider trading 2003, Rossi and Volpin 2004, Powell and Yawson (Meulbrook 1992, Jarrell and Poulsen 1989). 2005, Palepu 1986, Brar et al 2004). Factors such as Research in this area often investigates corporate size and profitability (Sony and Walkling 1993, restructuring events such as mergers, acquisitions and Palepu 1986, Cudd and Duggal 2000), evidence of tender offers and can be encompassed in the catch-all inefficient management and market sentiment term takeover. Announcements relating to these have (Powell 2004, Barnes 1999, 2000, Kennedy and been shown in the literature to have a price altering Limmack 1996), industry disturbance (Gort 1969, effect (Seyhun 1992). Takeovers are distinctive in Palepu 1986) can signal to market participants that a that a reaction is discernible in the value of the target takeover is likely to occur. The pre-event price run-up company’s stock prior to knowledge of the event therefore might not be attributable solely to informed becoming publicly available. A broad body of trading but rather to a combination of influences. This empirical literature has been published both in idea has empirical precedence and supports the Europe and the US which supports this position argument that pre-event activity preceding instances (Keown and Pinkerton 1981, Jenson and Ruback of corporate restructuring is conducted by investors 1980, Seyhun 2000, Bris 2005, among others). who hold public rather than private information. This Presumed in much previous work is that the pre- paper’s main contribution is to attempt to resolve this event price run up occurs as a result of informed debate in the UK within the post Financial Markets trading (Seyhun 1992, Meulbrook and Hart 1997, and Services Act (FMSA) 2000 context. Meulbrook 1992, Korczak et al 2010). If such run ups The period under examination marks the first are observed, this could draw into question the decade of the FSA’s assumption of the role of effectiveness of legislation and the FSA’s capacity to ‘policeman of the markets’. The FSA has made the

24 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

admission that a scarcity of resources prohibits it close of the trading day immediately prior to the date from following up on every suspicious transaction on which the takeover is announced. (Barnes 2009). Therefore this paper sets out to gauge A consensus exists in the literature that the the level of its success and calls into question the abnormal returns signified through a price run-up efficacy of using traditional indicators such as pre- features as a universal characteristic in the target event price run-up when such confounding effects are company’s share price prior to a public shown to be present. The possibility that the price run announcement. Despite the agreement that abnormal up is a consequence of informed but legitimate returns exist, competing reasons are offered to trading is also investigated using disclosed trades explain why this may be the case. The dividing line made by company directors. rests between those who are inclined to argue that Guiding the study are the following research such anticipation is consistent with ideas on semi questions. strong market efficiency and those who believe that it  What is the full extent of abnormal returns is a consequence of the leakage of privately held evident in target companies prior to the price sensitive information. In the former group it is announcement of a takeover? thought that investors pick up on publicly discernible  Can any of the established probability factors be signals that suggest that a merger is imminent, used to explain away unusual activity prior to a adherents of the latter opinion hold the rather more takeover announcement; and can a new factor intuitive belief that the change emanates from trading such as declared insider trades be added to this actions executed by those who know details of the list of explanatory factors within the post FMSA forthcoming deal. 2000 UK context? To investigate these questions I use event study 2.1 Explaining the Run-up analysis to observe abnormal returns prior to the announcement dates. This study also examines Several attempts have been made to attribute this separately the dates in which the takeover first came abnormal pricing behaviour to trading upon privately ‘into play’, that is, the day in which the rumour first held information. Meulbroek and Hart (1997) appeared in the media. Cross-sectional OLS examine the abnormal returns gained through 112 regression analysis is used to determine whether we instances of known illegal insider trading in takeover can make presumptions about the extent of illegal episodes in the US between 1974 and 1989. A insider trading inherent in the UK corporate control sample of mergers where no insider trading restructuring market. was known to have occurred is constructed in an The remainder of this paper is structured as attempt to isolate the effect on the run-up that could follows; first, the literature on insider activity be attributable specifically to this activity. Their pertaining to Mergers and Acquisitions is reviewed. results are consistent with the idea that insider trading Then, a full description of the legal context is contributes to the magnitude of the run-up provided with particular reference to the FSA’s role experienced before a merger is announced. Earlier in policing the markets and enforcing legislation. work by Asquith (1983), examined both the Next, the description of the sample, methods and announcement and outcome of events on 211 target procedures used are provided in addition to the results firms and demonstrated evidence for a positive pertaining to the extent of pre-event abnormal cumulative average abnormal return (CAAR) within pricing. An introduction to the factors which could the target firm of 15% in the 60 days before the offer an alternative explanation is then offered. announcement. Meulbroek (1992) also contributes to Following that is an empirical investigation that this argument through linking positive gains in the isolates factors which could contribute to explaining target run-up prior to merger announcement to days some of the pre-event pricing activity. The article where illegal insider trading was known to have concludes with a discussion of the implications of the occurred. These findings grow more intriguing when findings, limitations of the research and directions for robustness checks control for the presence of media future work in this area. rumours and suitable adjustments are made for the normal premium expectancy. 2 Literature Review A similar performance in CAAR was realised in the work carried out by Keown and Pinkerton (1981). There is considerable evidence to suggest that a price However in this study the authors attribute the cause run-up can be expected within target companies prior as media leakage of the event. Further weight is to the announcement of a merger, acquisition or added to the argument by Neely (1987) that the run- tender offer. Studies suggest that positive gains are up is the product of astute investor practice. In this realisable in stock prices a number of weeks before study of acquired and acquiring US Banks it was the formal announcement date (Jenson and Ruback found that abnormal returns in the target companies 1983, Bradley, Keown and Pinkerton 1981, Jarell, reached 9% in the two weeks prior to the Brickley and Netter 1988). Of these some claim that announcement, in total, over a period of seven weeks as much as half the total premium is attained by the before the event a CAAR of 15.1% is obtained. He

25 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

attributes the price behaviour to buying pressure abnormal returns in such instances to reach 11.2% by created by acquirers before an announcement is close of the last trading day prior to the rumour. The made. There is an acknowledgement by the author general conclusion is that as the rumour date often that both these possibilities are heightened when the precedes the announcement date by as many as two target firms in the sample are large. In smaller months, trading by directors can occur outside of a companies that come under less scrutiny the time when their activities are more likely to be likelihood of these being significant factors scrutinised. dramatically reduces. Within the UK, Korczak et al (2010) provide 2.2 What is the effect of regulation and evidence to demonstrate that insiders do indeed enforcement? display a propensity to trade ahead of company news announcements. Furthermore it is shown that the If we are to accept the insider trading hypothesis, capacity for investors to buy ahead of good news empirical studies suggest that this can have an announcements outweighs their inclination to sell immediate effect on price and that this can occur as prior to the release of negative information. However, early as fifty days before the first public disclosure of the level of insider activity before trading ahead of the takeover (Meulbroek 1992). The effect further bad news is governed by the supposed significance of intensifies in the twenty trading days prior to this the event. In conditions where the effect of the news event (Meulbroek and Hart 1997). The question then is thought to dramatically affect stock prices the arises as to whether regulation is a sufficient deterrent probability of insider selling is shown to decrease. against market abuse of this nature. Although The conclusions reached are that behaviour is toned evidence is mixed some work suggests that informed down when an incident is likely to attract more trading is curbed when insider trading regulations are attention from regulators and market commentators. introduced (Korczak et al 2010). Durnev and Nain The dispute relating to what causes a price run (2005) for example, examine 2,827 firms from 21 up could be resolved in part through examining the countries in a sample of takeovers taken between phenomenon while controlling for a number of 1996 and 1999. They find that on the whole, factors that signal the likelihood that a company restrictions on insider activity do curb levels of could become a target. These could include factors informed trading but that in companies where the such as size, undervaluation or inefficient share ownership is concentrated among large management of the company in question. Work in the shareholders these restrictions become less effective. US stock market was conducted by Jarrell and Certainly the positive attitude among regulators Poulsen (1989) who examine 172 tender offers toward imposing restrictions in this area is gathering between 1981 and 1985 proposing that a statistically into something of a global trend. Bhattacharya and significant CAAR of 11% can be detected as early as Daouk (2002) surveyed 103 countries which have fifteen days before the announcement. This accounts stock markets; they find that 87 of these prohibit the for 40% of the entire premium attained owing to the activity. The principle motivation behind the ban is to takeover. They attribute this to media speculation protect the interests of uninformed investors against concerning the probability of an impending takeover, more informed opposite parties. Despite this, when although they do not rule out the possibility that empirical tests are conducted to ascertain the effect of insider trading activity does not fuel the press regulation, it does not follow that an increase in speculation. The argument is also put forward that the regulation or sanctioning will dampen either the known attempts by the acquirer to gain a foothold in instance or volume of insider trading. Seyhun (1992) the target company, through purchasing shares bears conducts an investigation of open market sales and some relation to the evidence of anticipation. purchases in 8,856 US firms between 1975 and 1991; However, with known cases of insider trading which he finds that despite the overhaul in regulation and are examined in conjunction with the abnormal sanctioning in the 1980’s profits earned by insiders returns a relationship does not appear to be evident. increased from 3.5% in the pre-1980 period to 7% in These conclusions do not however discount the the years following 1984. Further to this, the volume possible influence of unknown instances of insider of known insider trades also increased fourfold. One trading on the recorded run up. caveat here is that the sample involved uses declared During instances where news of the merger has insider trades and that the increased sanctioning could been leaked or has established itself as a rumour there have forced these trades out into the open. is no longer a solid case to suggest that the pre- In a study of European takeovers Bris (2005) announcement trading activity is instigated solely as a shows that a period of intense takeover activity is consequence of privately held information. Evidence generally preceded by an episode of concentrated of activity prior to the rumour date is highly insider trading activity. The consensus among suggestive of insider trading activity. In the empirical regulators is that it appears to be a form of ‘cheating’ literature a number of studies have documented the that is ‘legally forbidden, morally wrong and presence of abnormal returns prior to the rumour economically dangerous’ (Zevitt 1998). Bris (2005) date. Jarrell and Poulsen (1989) ascribe cumulative examines 4541 acquisitions spanning 52 countries

26 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

and demonstrates that there is a direct correlation legislation creates a reduction in terms of the cost of between the severity of punishment, the diligence of capital and volatility in addition an increase in the regulators and the incidence and profitability of liquidity of traded assets. The authors argue that these insider trading. It appears that more stringent changes are the result of increased investor enforcement dramatically discourages this activity. confidence stemming from a belief that the new The findings have theoretical precedence the rational measure could facilitate greater success in economic perspective as espoused by (Becker 1993), prosecution than which had previously achieved. which as the cost of violating the market abuse laws Implicitly suggested here is that it is when increase self-interest dissipates as the marginal enforcement or a promise of successful prosecutions benefits from participating in the activity decreases. manifests, it is then that we see the benefits arising Hence we see a fall in the levels of insider trading. from a regulatory regime. However a word of caution must be introduced at this juncture, if a market abuse is legislated for in a 2.3 UK regulation and enforcement certain market yet it does not become enforced, then the effect on the cost of equity can be more severe The introduction of the Financial Services and than if there had been no law in place. Bhattacharya Markets Act (FMSA 2000) harnessed the FSA with and Daouk (2009) find this to be the case in emerging greater powers of investigation and enforcement than markets where these conditions exist. Explanation is that which it had previously held. The increased offered using the prisoners dilemma analogy, in sanctioning powers in addition to the ability to situations where a good but unenforced law exists penalise failure to co-operate with investigations is some individuals will obey the law while others will perhaps one of the more important evolutionary choose not to comply, therefore creating changes in the market abuse regulation (Rider et al disequilibrium in behaviour. The cost to the law 2009). The Act permitted the FSA to pursue abiders will thus be greater than in the event that a individuals and companies through both the civil and law exists and is enforced or when no law exists at criminal courts, in order to allow the FSA to meet its all, in such circumstances behaviour would follow statutory objectives of protecting market confidence equilibrium. Where enforcement is improved then and reducing financial crime (Section 2(1) FMSA there is some empirical evidence to suggest that the 2000). Sections 167 and 168 of the FMSA granted the proposed aim of regulation can be met. Following FSA powers to conduct investigations on any other work which investigates insider trading individual suspected being involved in market abuse. behaviour around price sensitive disclosure events The terms of the legislation confers a wider remit (Bettis, Cole and Lemmon 2000, Roulstone 2003 and beyond those who owe a fiduciary duty to Garfinkel 1997) Jagonlinzer and Roulstone (2009) shareholders of companies participating in the concur that insiders shift their trades until periods market. As such the net widens to include those who after event related earnings announcements; they go may benefit from the dissemination of the further in suggesting that this results in a fall in information. However, budgetary constraints mean abnormal returns which are gained through insider that not every suspicious trade can be examined. An trading activity. This is in contrast to the findings of incident, if it is to warrant an investigation must meet Seyhun (1992) who, while discovering displacement a pre-designated set of criteria, which in summary, behaviour, does not uncover a marked difference in demand material evidence of a breach of legislation, the gains made by insiders once the trading pattern proof of shareholder loss/detriment, evident risk that has shifted. This evidence is produced using US data the incident could damage investor confidence and following the imposition of the Insider Trading and that it falls within the strategic priority of the FSA at Securities Fraud Enforcement Act (ITSFEA) 1988. the time (Rider et al 2009). Furthermore, the FSA The primary tenets of this act brought an increase in retain the right not to take action in cases where a the sanctions available to authorities in addition to a legal infraction has been identified. This usually broadening of the terms of culpability to incorporate extends to instances that are not deemed to be the firms to which insiders belong, thus incentivising particularly serious or to trades where the individual firms to police the activities of their employees to a or firm involved has self reported. Once necessary greater degree. These findings add further weight to remedial action is undertaken, the FSA may prioritise the argument that it is the threat of enforcement and the objective of maintaining a co-operative and open sanction which become the main drivers which relationship with firms, especially those that mutate trading patterns among insiders. demonstrate the initiative to assume responsibility for Gilbert et al (2007) supports the idea that an its own regulatory infringements. In such cases, the introduction or improvement of the regulatory FSA will refrain from taking further action (Rider et framework specifically relating to enforcement can al 2009). This implies that simply conferring the have a positive effect on markets. In their powers of enforcement are not enough to guarantee investigation of the New Zealand stock market that every case of market abuse is dealt with. The performance following the amendment of existing purpose of this paper is in part to air the question as insider trading regulation they find an amendment to to whether there is a significant failure to enforce

27 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

legislation in that most typical of insider 2.4 Disclosure opportunities namely, trading before takeovers. In a speech delivered at the FSA’s Law Part VI of FMSA 2000 requires companies issuing enforcement conference in 1998, the FSA’s director shares to release information along the regulatory of enforcement, Margaret Cole, confirmed the information service, which when publicised would intention of the FSA to be ‘bolder and more resolute’ lead to a considerable movement in the price of the in pursuing market abuse cases, so as to introduce ‘a underlying security. This is what is deemed in the change in the culture in the city’ (Cole 1998). The legislation to constitute ‘price sensitive information’, focus of this change was to pursue offenders through (Rider et al 2009). Furthermore the information is the criminal courts where a prosecution if successful required to be disseminated as soon as is possible could result in a prison sentence. The rationale behind after the information comes to light internally. The this is that stigmatising offenders with criminality stipulations are that reports of director’s dealings and would have a greater dissuasive effect than simple transactions on the accounts of their spouses and imposition of civil sanctions (Symington 2008). The children are offered in as timely a manner as possible. penalties which the FSA has the power to impose are More explicitly, the director is obliged under the UK wide ranging. It could publish a statement detailing Model code to inform their company of such activity how a person has participated in market abuse rather no later than five working days following the trade. than impose a tangible penalty. Where remedial The firm is then obliged to report this information to action has been undertaken by the transgressor this is the LSE no later than one working day after the often the preferred avenue of pursuit. transaction has occurred. The company must also It can also assist both parties involved to reach ensure that data vendors are informed via the an agreed settlement, where it may consider the regulatory news service feeds. Furthermore the firm individual circumstances of the case and issue an must enter details of the trade in a publicly available appropriate penalty. In particularly serious cases it register within three days of the event (Fidrmuc et al may take action to remove the companies’ business 2006). There is also a requirement put upon the permits or in the case of individuals their ‘approved company issuing the underlying shares to make persons status’, (FSA enforcement guide chapters 8 available to the FSA a full list of persons who may and 9). The authority may also take out a court have access to the information. Individuals closely injunction against individuals or companies to compel connected to the deal such as senior management and these to take a proscribed course of action or prohibit directors are also obliged to disclose transactions of them from further engagement in market abuse. This issuing company shares in their own accounts. The is a particularly powerful sanction as refusal to full rules surrounding disclosure are available in the comply would result in the party in question being disclosure and transparency rules (DTRs) for listed held in contempt of court. Financial penalties when companies; which is contained in the FSA’s applied are determined in accordance with the figure handbook. The statutory authority for these rules is the offending party is thought to have gained as a laid down in part VI of FMSA 2000. result of the transaction. Under sections 201 and 402 of FMSA 2000, the FSA has the power to prosecute 2.5 Self Regulation the offence of misleading statements and practice. Further to this, it has also the power to prosecute Takeover activity in particular draws substantial insider dealing under part five of the 1993 Criminal attention from regulators as corporate restructuring in Justice Act. In determining whether or not to pursue a the markets has shown to be a focus of insider case through the criminal courts factors such as the activity. The aspect of the model code which is seriousness of the offence and the impact this may concerned with takeovers (the takeover code) is have had on the markets. Penalties liable to a person administered by the panel on takeovers and mergers. convicted of the offence of insider dealing are either a The code lays down general principles and practices fine or imprisonment of up to seven years (FMSA which are to be followed by listed companies during 2000, Section 397(8) and the Criminal Justice act times of corporate change. In general the main aim of 1993 section 61). In cases deemed to be of particular setting these principles is to ensure that shareholders severity the FSA may refer the matter to the serious receive fair treatment, adequate information and to fraud office that may pursue a conviction under ensure that no abusive trading occurs ahead of a bid section 2 of the Criminal Justice Act 1987. Such announcement. While the panel itself holds no referrals are made only in cases where the alleged regulatory or sanctioning powers its decisions can be fraud exceeds the value of £1 million or where the acted upon by the FSA, therefore a breach of code nature of the case requires the investigatory power of could result in disciplinary measures, which at the an organisation with a further reach than the FSA extreme might involve a delisting of the company (Rider et al 2009). concerned. Part of the code (Rule 2) requires that the bidding company must declare its interest in the target company should the target’s unusual price movements occur in the target’s share price or

28 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

rumours surface relating to the possibility of a 2.7 The future of enforcement and takeover. These incidents of course must be seen to regulation be attributable to the acquiring company. The code (Rule 2.1) also requires any third party who may have While the policy, as it manifests through the access to such price sensitive information to keep this legislation, appears adequate to deal with market private. abuse in its many forms, evidence of insider activity in the form of abnormal returns remains. In addition 2.6. Enforcement to this there has been a clear admission by the FSA that securing successful prosecutions is fraught with In the years before the FSA received the powers to difficulty (Cole 2007). If changes are to occur, these police and prosecute the variant forms of market are more likely to relate to a strategic rather abuse, responsibility for the task belonged to the legislative change. In a review of ‘suspicious Department of Trade and Industry (DTI). Throughout transaction reports’ (STRs) the FSA confirmed that it the 1980’s 21 acts of insider dealing were had received 266 STRs between July 2005 and successfully prosecuted under the Companies Act October 2006, of these 255 related specifically to (1980/1985). Of these, six were as a result of trading alleged incidences of insider trading (Barnes 2009). by a director or an associate, or were made upon Commentators have argued that the regulation should information provided by a director. The remaining be based upon a set of principles rather than defined offences were committed by individuals who held circumstances and should shift the focus towards close links with companies through either compliance (Barnes 2009). This is achievable through professional involvement or links to other insiders directing attention to the compliance efforts of (Barnes 2009). The prosecution success rate companies deemed likely to be involved in market represents 58% of all cases pursued. Following the abuse. The introduction of STRs in addition to the implementation of FMSA 2000 in January 2001, the implementation of software systems which in real FSA took over the role of policing market abuse from time identify notable changes in the share price and the DTI. With only two criminal convictions secured the volume of transactions for listed companies since 2001, the FSA demonstrably favours the civil (SABRE 2 an acronym for Surveillance and Business route when enforcing the regulations. Since 2001, in Reporting Engine), are new measures which it is the UK there have been 12 successful civil actions hoped will increase the monitoring capacity and relating to market abuse, of these, eight involved effectiveness of the authority. insider trading. In the two cases where criminal proceedings were initiated, the five individuals 3 Data involved received prison sentences of up to two years and one individual received a community service The market for corporate control incorporates a order. In the civil cases fines totalling £27,550,143 number of restructuring processes that fall under the were imposed, the smallest of these was £1,000, umbrella term takeover, consequently in the literature while the largest amounted to £17,000,000 and was this can refer to mergers, acquisitions, proxy contests levied against the Royal Dutch Shell Group in 2004 or tender offers. for providing misleading information to investors In mergers, acquisitions and tender offers the (FSA. The largest fine imposed for insider dealing bidding firm proposes to buy shares in the target firm was levied against GLG partners LP and a Mr Paul for a price that is higher than the target firm’s value, Sabre who were each fined £750,000 respectively and which is thought to be reflective of future income (FSA 2011). Despite these successes the FSA has by generated from the target’s assets following the deal. its own admission much work to do minimise the In the case of Mergers and Acquisitions the extent of insider trading in the UK market. In 2006 a management in the target company is approached study was commissioned by the FSA that examined a about the deal and in order to go ahead approval is total of 769 merger announcements on the London required from the target company’s board of stock exchange from 2000 to 2005 (Dubow and directors. In a tender offer the bidding company Monteiro 2006). The study examines each takeover approaches the shareholders directly to buy the announcement for abnormal returns over a four day shares. The proxy contest is an attempt to win seats event window. Their findings suggest that in 20% of on the Board of Directors usually as a result of an the merger announcements surveyed informed trading activist group emanating from within the shareholder appears to be present. Concerns could be raised with or management groups (Jenson and Ruback 1983). this study that it does not go far enough and that full For the purpose of our investigation we will extent of the activity is not captured. In much of the exclude both tender offers and proxy contests from literature significant abnormal returns are recorded our sample, as a higher degree of legitimate leakage over periods as much as fifty days prior to the is possible to larger groups of people thereby blurring announcement (Meulbroek 1992). the line between what is considered to be public and private information.

29 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

Restricting the sample to Mergers and low points of economic activity within the decade Acquisitions, the deals identified involve 227 target respectively. The number of takeovers in the latter companies which are listed on the London Stock year far exceeds those which occurred the former Exchange. All deals were completed and carried out year indicating that the motivations underpinning in the UK between March 2001 and January 2011. decisions could have been value rather than growth Details of each deal were acquired though the Zephyr based. All offers are in the form of cash or equity database. Figure 1 illustrates the breakdown of the with the exception of one which was financed Mergers and acquisitions as they occurred in each of through the issuing of corporate bonds. Constituents the years through the sample. The most intense for the sample including the dates involved are listed periods in takeover activity occurs in 2006 and 2009; in Table VII in the appendix. these years could be seen as marking the high and

Figure 1. Number of Takeovers per year for the sample

In our analysis we will estimate abnormal announced. From our entire sample, 39 of the 227 returns surrounding two specific events for each of deals have a rumour date that is different from the the deals. The announcement date is often used in the day in which the deal is announced. The mean literature to mark the date at which the information average length between the dates is 231 trading days becomes public. The definition for this date collected although this reduces to 118 when outliers are through Zephyr is the day upon which either a formal removed. With such a large discrepancy in trading offer has been made to merge with or to acquire the days there is a distinct possibility that more informed target company or when one of the parties involved trading could occur prior to a rumour when the has confirmed that the deal is to go ahead. possibility of the trade being linked with the Measuring the anticipatory effect created for a announcement is less. The solution is to test both reason other than public disclosure can only be events independently of one another to ascertain achieved with any degree of accuracy if we factor in whether abnormal returns are evident. Table I the possibility that news of the impending change provides a breakdown of both samples in terms of could itself be viewed as the event. If the markets are size. Although the number of observations in the informationally efficient then the first day in which sample differ the average size of the deal expressed the possibility of a merger or takeover is openly as both the arithmetic mean and the median remain discussed could see the target’s share price move in similar. response. As a consequence, for each deal in the sample we examine the date at which a rumour first appeared in addition to the announcement date. The former date is defined as that day on which the possibility of a forthcoming deal is first mentioned. This may be the first time it is reported in the media, or issued as a company press release. The announcement date doubles as the rumour date in instances where the first indication of the possibility of a move is the day upon which the deal is

30 Risk governance & control: financial markets & institutions / Volume 1, Issue 3, 2011

Table 1. Deal Sizes

Sample Obs Mean SD Median Maximum Minimum Announced 207 261088.7789 1365518.8714 9885.4950 14849363.0000 2.8600

Rumoured/Announced 193 286384.0979 1412209.7042 9527.8800 14849363.0000 2.8600

Presented abow are the deal sizes for the target firms analysed in both samples statistics relating to size are given in £ (thousands)

Prices used to generate returns for each public. Insider trading (IT) takes the value of 1 if company in the sample were obtained from disclosed trades have been made by company insiders Datastream. Once missing observations are removed within a three month period before the event as a consequence of the pre – event estimation period occurrence. The data for this was obtained from a stretching beyond the date of the firm’s establishment sample of declared insider trades provided by the sizes for the announced and rumoured event Directors Deals. MC is the logged pre deal market samples are 207 and 193 firms respectively. capitalisation of the target firm. Net profit margin The accounting information used to construct (NPM) is calculated through dividing the profit the variables that control for hypothesised factors that before tax for the company by the turnover in the signal an increased probability that the company same year for that company. GRM The extent to could become subject to takeover were obtained from which growth and resources are mismatched (GRM) Zephyr. Table II provides the summary statistics for is attained through dividing the previous year’s the variables; these have been shown in previous turnover by the total assets for that year. The book to empirical studies (Palepu, 1986, Sony & Walkling, market value for the firm (BMTV) is computed using 1993, North, 2001 among others) to be possible the by the dividing the Shareholder’s equity by the signalling factors which could offer an explanation Market capitalisation of the firm, these figures are for observed pre-event run ups. taken from the previous year’s annual reports for each Contained in Table 2 are the summary statistics company. All accounting figures in addition to the for the variables which are investigated as possible event dates for each Merger or Acquisition were signalling factors. These are analysed as independent obtained from Zephyr and were taken from the annual variables in a cross sectional OLS regression where reports in the year immediately prior to the event. the dependent variables are the variant measures of Where there is a negative value in shareholder equity event anticipation which represent trading activity (NVSE) this takes the value of 1 if shareholders prior to news of the intended takeovers is made equity value is negative and 0 if this is not the case.

Table 2. Descriptive Statistics

Variable Mean SD Median Maximu Minimu m m IT 0.1131 0.3177 0.0000 1.0000 0.0000 MC - 226128.1 - 21109.8615 156 -8.5568 834.2412 2567725.0000 NPM 10.1097 2.2814 9.6546 17.1282 5.6858 GRM 0.9652 1.2197 0.5670 9.0359 -0.3194 BMTV 0.8870 2.9511 0.4061 12.5345 -21.3334 NVSE 0.0928 0.2909 0.0000 1.0000 0.0000

A further variable named declared insider trades 4 Methodology is also introduced to attempt to explain pre-emption of the event. If an insider’s buy trade has been Analysing trading surrounding event declared to the FSA and made public, then this could dates possibly indicate to outside investors that the company increases in possibility for takeover. In order to analyse the firm returns prior to the dates Information on these trades was obtained from a when takeovers become imminent an event study is dataset provided through Director’s Deals. From this employed following the process set out in McKinlay a dummy variable was constructed where a value of 1 (1997). The method allows for detection of abnormal was assigned if a Director bought shares in the target returns surrounding the day the news is made public company in the period of three months preceding the either through formal announcement or when a event date that is analysed. rumour surfaces in the press. Abnormal returns are calculated according to the benchmarks

31 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

recommended by Brown and Warner (1985). These Where T1 represents the first trading day and T2 is the are the market and the constant mean return model. final trading day within the abnormal return series. Both models use a separate estimation period which The constant mean adjusted model is computed does not overlap with the event window period. For through finding the arithmetic mean return for the this study this begins for both models at day -160 and firm stock i over the estimation period similar to that finishes at day – 61. The event window period spans previously described. The average return figure is from day -60 up to day +10, where the event occurs at then subtracted from the return on each trading day in day t = 0. the event window period. The market model benchmark assumes that security returns are estimable using a single factor market model as follows: The procedures for arriving at AARit and CAAR are similar to those described for the market model. The degree of statistical significance following Where Rit is the rate of return of the common parametric assumptions is ascertained using a t-test stock of target i on day t and Rmt equates to the return similar to that employed in Bialkowski et al 2011. on the FTSE All Share Market index which is sourced This approach enables volatility stemming from the from Datastream. it is the random error term which is event to be incorporated into the test through assumed to have an expected value of zero and is retaining the estimation within the event window uncorrelated both with Rmt and the returns for security period. i over the estimation period. The error term is also assumed not to be auto-correlated and to be homoskedastic. βi represents the slope parameter taken from the OLS regression between returns of stock i and with returns on the market over the Where n1 and n2 represent the start and finish estimation period. Beta is thus a measure of the dates of the event window and σ(AAR) is the standard sensitivity of the returns on the stock to the market deviation from the mean of abnormal returns which is index returns. calculated from daily observations over the period Abnormal returns according to this model are over which the event is measure. assumed to be realised as follows: A further test for significance is carried out using when the parametric assumptions are relaxed. This investigates whether the proportion of positive

CARs can differ from 0.5. (This test is defined in The coefficients ai and βi are estimates and are Białkowski et al 2008). It is computed as follows: the products of an OLS regression between the returns of stock i and the market over the designated estimation period. Abnormal returns are then averaged across the entire sample of target company Where p is the observed proportion of CARs securities to obtain a sample mean. which are positive and N refers to the size of the

sample.

4.1. Controlling for external factors

Where N refers to the number of firms in the In the attempt to investigate these other possible sample and t represents the trading day within the contributory factors to the pre event run-up a cross- event window period. The Cumulative Average sectional OLS regressions model is constructed using Abnormal Return (CAAR) is then computed for the a number of indicators identified through the desired number of days within the event window literature on Mergers and Acquisitions as independent period as follows: variables. The regressions are performed using the

CARs and two separate measures of Run-up Indexes generated from both the Rumour and announcement samples as dependent variables. The models can be

displayed as follows:

32 Risk governance & control: financial markets & institutions / Volume 1, Issue 3, 2011

Where CAR equals to the cumulative abnormal A further variable, negative value of shareholder return in the sample of target companies prior to the equity (NVSE) is added to represent instances where event date, RU1i and RU2i are run-up indexes the value of book to market ratio falls below zero constructed following Bannerjee (2001), the CARs in allowing further power to investigating situations each sample are separated into a pre-event phase where companies are close to insolvency. Regression designated as a Run-up period, an event gain phase coefficients are tested for statistical significance using and a post event drift phase. standard student t tests. The results of the regressions are displayed in Table 3 and are discussed according to each of the factors in turn.

5 Results

5.1 Evidence of price run up surrounding announcement and rumour dates

The Cumulative Average Abnormal Returns (CAAR)

A Run-up index (RU1) is then computed for for the announcement sample is given in Table III and each of the CARs and this is considered to be the Figure 2. The results demonstrate that there is proportion of the Run-up to the entire premium evidence to support the hypothesis that statistically gained by the target company shareholders. significant abnormal returns are realisable before the In the spirit of Bannerjee (2001) it is defined event date. The findings are conclusive when both the thus: market model and constant mean return benchmarks are used. For both these benchmarks CAARS remain consistently positive and maintain this sign from early in the event window period at 55 days prior to the event. Similarly when the daily average abnormal returns (AAR) are counted, in the sixty days prior to

the announcement date AAR is positive for 43 days The second Run-up index (RU2) is calculated for the constant mean return model while the figure is using the CAR from the entire event window period 41 days for the market model. For both models there as the denominator; it can be displayed as: is a drop in CAARs on the day of the announcement,

this is not consistent with much of the literature. It suggests that information pertaining to the event has for the most part been assimilated into the price prior to the announcement being made. Interestingly, in the IT is a dummy variable which takes the value of five days before the event date AAR is consistently 1 if declared insider trading takes place during the negative. If insider trading is behind the abnormal three months immediately prior to the event date used returns then there is a decided absence of activity to generate the dependent variable. The data on immediately preceding the announcement. This could declared insider trades was sourced at Directors Deals signal that in order to exercise caution insiders shy and refers to the dates that shares were bought by away from conspicuously flouting the rules. directors or officers within each of the target Furthermore, the absence of a surge of immediate companies in the samples. MC refers to the lognormal pre-event positive AARs fails to strengthen the idea market capitalization of each firm; this is estimated that the takeovers in question are publicly discernible. through multiplying the share price by the total Reported in the Table 3 are the cumulative number of shares outstanding in the company, this abnormal returns for the event windows surrounding ratio is transformed using the logarithmic process to the announcement of forthcoming Mergers or ensure a normal distribution in the sample. NPM is Acquisitions which have since been completed. The the company’s Net Profit Margin; this is calculated cumulative average abnormal returns are calculated by dividing the profit before tax for the firm by its from prices denominated in UK sterling and is turnover. GRM is a variable taken to represent the expressed in percentage terms; the sample in total mismatch between growth and resources, it is consists of 208 Companies. Panel A contains CAARs computed through dividing the turnover realised in defined over event windows of various sizes for the previous year by the estimated value of the total which abnormal returns are calculated using the assets for that year. constant mean return model providing an expected Finally the variable BMTV represents the book return, this is generated using the estimated average to market value ratio and is computed by dividing the of returns in each company 100 trading days value of the previous year’s shareholder equity by its immediately prior to the first day of each event market value for the same period. The figures for window. In Panel B CAARs are calculated from these variables were taken from the previous year’s returns generated using the Market Model, which annual reports provided through the Zephyr Database.

33 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

estimates returns from coefficients taken from an report the results of a non parametric investigation OLS regression using a estimation period similar to that bases its tests on the hypothesis that cumulative that of the previous model. Listed in the third and abnormal returns are equal to zero when the fourth column of each panel are the t-statistics and p- assumption that returns are normally distributed is values for the sample of CARs, this tests the removed. The null hypothesis attests that the hypothesis that as a collective the mean of the proportion of negative and positive CARs are equally Cumulative abnormal returns is equal to zero. The balanced, where the mean is equal to zero and the final three columns in each panel of the table below standard deviation is equal to one.

Table 3. Cumulative Average Abnormal Returns surrounding Announcement date

Panel A: Constant Mean Return Model Event Proportion of Window CAAR t-stat p-value positive CARs z-stat p-value (-60, -1) 10.5822% 3.4615 0.0010 0.5604 1.7504 0.0400

(-60, 10) 10.7616% 2.9539 0.0043 0.5362 1.0453 0.1479

(0, 10) 0.1794% 0.0954 0.9956 0.4638 -1.0453 0.8521

Panel B: Market Model Event Proportion of Window CAAR t-stat p-value positive CARs z-stat p-value (-60, -1) 9.0314% 2.8735 0.0056 0.5550 1.6007 0.0547

(-60, 10) 8.9455% 2.4165 0.0182 0.5407 1.1798 0.1190

(0, 10) -0.0859% -0.0434 0.9662 0.4641 -1.0403 0.8509

Figure 2. Cumulative average abnormal returns to sample of target companies surrounding announcement date of Takeover

Panel A: Constant-Mean Return Model

14,00%

12,00%

10,00%

8,00%

6,00% 4,00%

2,00%

Cumulative abnormal return 0,00% -60 -56 -52 -48 -44 -40 -36 -32 -28 -24 -20 -16 -12 -8 -4 0 4 8 -2,00% Day relative to the event

34 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

Panel B: Market Model

12,00%

10,00% n 8,00%

6,00%

4,00%

2,00%

Cumulative abnormal retur 0,00% -60 -56 -52 -48 -44 -40 -36 -32 -28 -24 -20 -16 -12 -8 -4 0 4 8 -2,00% Day relative to the event

As targets are held to consistently deliver used includes 192 target companies. Panel A contains positive premia as a result of announcement of a CAARs calculated using the constant mean return proposed deal, cumulative abnormal returns from the model, both event windows defined span from day t = day of the announcement up to the tenth day after the -60 to day t = 0 and from day t = -60 to day t =+10 event are examined separately. These findings show respectively. The benchmark model used in Panel A that with respect to announcements, the market meets provides an expected return equating to an average of a proportion of the deals proposed with genuine 100 trading days immediately prior to the first day of surprise. The presence of a positive CAR however is each event window. In Panel B CAARs are generated suggestive of the possibility that private information using the Market Model to estimate expected returns, drives trading on a number of target company’s this computes an expected return using coefficients shares in the sample prior to the announcement. taken from an OLS regression over an estimation Turning now to the investigation using the first period spanning 100 days and which ends date at which news of the impending deal appears in immediately prior to the first day of the event the public sphere, it can be seen from Table IV and window. The third and fourth column of each of the Figure 3 that both benchmarks demonstrate that panel’s reports the t-statistics and p-values for the positive CARs of a slightly greater magnitude are series of CARs, and tests the hypothesis that the mean attained through the rumoured sample. The similar of the Cumulative abnormal returns is equal to zero. sizes could be explained by the similarities of the two The remaining three columns in both panels below samples, 39 of the 193 firms examined display a report the results of a non parametric z test which rumour date that is separate from the announcement investigates the hypothesis that the mean cumulative date, the difference that does exist however may abnormal returns equates to zero removing the suggest that insider trading could be more likely to assumption that the CARs are normally distributed . occur prior to a rumoured rather than the Under the null hypothesis the proportion of negative announcement date as the possibility of detection and positive CARs share equal weight, upon which would perhaps be less likely. However this assertion the mean is equal to zero and standard deviation is is undermined by the proportion of deals displaying equal to one. positive CAARs which is only slightly greater for the sample that investigates the announcements. A further notable point is that in the 11 days following the event date the CAARs drop in magnitude to the tune of 0.4% for the constant mean return model and 0.69% for the market model, as buying pressure eases following the release of the information (See Table 4). The results recorded in the following table are the cumulative abnormal returns for the event windows surrounding the rumour date for a Merger or Acquisition which has since been successfully completed. The cumulative average abnormal returns are expressed in percentage terms; the entire sample

35 Risk governance & control: financial markets & institutions / Volume 1, Issue 3, 2011

Table 4. Cumulative Average Abnormal Returns surrounding Announcement/Rumour date

Panel A: Constant Mean Return Model Event Proportion of Window CAAR t-stat p-value positive CARs z-stat p-value (-60, -1) 13.8459% 4.8282 0.0000 0.5596 1.6675 0.0477

(-60, 10) 12.3654% 3.6893 0.0004 0.5648 1.8148 0.0348

(0, 10) -1.4805% -1.0068 0.3085 0.4404 -1.6675 0.9523

Panel B: Market Model Event Proportion of Window CAAR t-stat p-value positive CARs z-stat p-value (-60,-1) 11.5756% 4.1741 0.0000 0.5692 1.9523 0.0255 (-60, 10) 9.8910% 3.1100 0.0027 0.5436 1.2220 0.1108 (0, 10) -1.6847% -1.4043 0.1878 0.4410 -1.6586 0.9514

Figure 3. Cumulative average abnormal returns to sample of target companies surrounding rumour/announcement date of Takeover

Panel A: Constant-Mean Return Model

16,00% 14,00% 12,00% 10,00% 8,00% 6,00% 4,00% 2,00% Cumulative abnormal return 0,00% -60 -56 -52 -48 -44 -40 -36 -32 -28 -24 -20 -16 -12 -8 -4 0 4 8 -2,00% Day relative to the event

36 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

Panel B: Market Model

14,00%

12,00% urn

10,00%

8,00%

6,00%

4,00%

Cumulative abnormal ret 2,00%

0,00% -60 -56 -52 -48 -44 -40 -36 -32 -28 -24 -20 -16 -12 -8 -4 0 4 8

Day relative to the event

Note: Portrayed in Figure 2 are the cumulative average abnormal returns in an event window for a sample of companies 6o trading days preceding and 10 days following the first published rumour of a Merger or Acquisition, the sample examines 193 events in the UK markets over a period of 119 months from March 2001 to January 2011.

It would appear that for both the announcement argue that trading on insider information does not and rumour sample statistically significant cumulative occur. A tradition of information leakage (as has been abnormal returns are present before the occurrence evidenced in several of the successful prosecutions by with each date. This is consistent with the hypothesis the FSA) ensures that the net of possible insiders that informed trading using these is taking place prior spreads much wider than those who sit in the to the public dissemination of the information. The boardroom. The implications that can be drawn from magnitude of the run-up is similar to results produced this are that while the reporting system does appear to by Jarell and Poulsen (1989) who record abnormal deter directors and senior executives from trading returns of 11% on a similarly sized sample where the ahead of takeover announcements these measures can event date employed equates to the day on which the directly control only a small number of the group of news is publicly disseminated. people that could have access to this information. If There are a number of alternative explanations informed trading is driving the upsurge in returns which could account for the run-up. A large body of prior to the event then it can only be concluded that literature has been devoted toward determining this is the product of undisclosed illegal insider possible signals of impending takeovers. trading, thus corroborating the observations made in Consequently a number of factors have been the UK context by Korczak et al (2010). identified as publicly observable features of targeted companies, these become signals that point to the 5.2.2. Size increased likelihood that a takeover is imminent. Numerous studies (Sony and Walkling 1993, Palepu 5.2 Controlling for publicly observable 1986, Ambrose and Megginson 1992, Cudd and signals Duggal 2000) have shown that takeovers are more likely to happen to smaller companies. There are a 5.2.1. Declared Insider Trading number of possible reasons for this, either as companies increase in size the pool of possible From the results of the OLS regressions it is clear that acquirers grows smaller. A second reason relates to disclosed insider trading does not play a major role in the fact that the probability of a takeover decreases as contributing to the magnitude or presence of the pre size determined transaction costs increase (Barnes event price run-up. While this particular variable has 2009). Further to this, in comparison with larger not been tested in the literature to date, the evidence companies the capacity of smaller companies to is not convincing, with respect to the sample that dedicate resources to a defensive campaign is much announced trades could be linked to abnormal less. Market capitalisation (MC) is employed to proxy pricing. Therefore it would be safe to assume that for size in the company immediately prior to the company directors and top executives refrain from announcement. Despite the support for this factor in trading prior to the announcement or first indication the literature, it is apparent from Table 3 that the of the possibility that a takeover may occur. These findings do not concur conclusively with the idea that findings do not provide evidence to successfully size has anything to do with the likelihood that a

37 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

takeover will occur. A statistically significant but merger announcements is not apparent through these slight relationship is observed between the second run findings. up measure (RU2) in the announcement sample using the constant mean return benchmark. However, this 5.2.5. Undervaluation does not remain consistent across the remaining models, neither is a relationship present in the It can be often the case that a firm is snapped by a announcement sample. Taken together these results purchaser because its market value falls short of the suggest that the size of the target firms was not acted value of its asset. Undervaluation has been indicated upon as a signalling factor for a possible takeover. in a number of studies to be a signal of inefficient management and could attract those interested in 5.2.3. Profitability stripping and selling off the company’s assets but also firms who could manage the company more This is a factor that allows one to gauge the efficiently to make better use of its assets. probability of a takeover based on the premise that a Consequently, if a firm displays a low Book to firm can be inefficiently managed. Several studies Market value ratio then it is considered to increase in indicate that unprofitable firms are more likely to attractiveness as a takeover target. The findings become the target of a takeover than their profitable presented in Table 3 indicate that there is some counterparts (Singh 1971, 1975, Kuehn 1975, Palepu evidence in support of this idea as the relationship is 1986). It follows intuitively that firm shareholders, statistically significant at 5% and 1% in the rumour being concerned primarily with performance, show model using both benchmarks models respectively. less tendency to resist advances from acquirers. When However no link is evident with the run up ahead of the significance of the relationship with the net profit the announcement date. Based on these results the margin figures (NPM) is examined, it appears that relationship is open to question, while it may in some profitability offers a statistically significant but very instances indicate a motivation for the company to be small explanation using both benchmarks for the taken over it is not does not provide convincing proof CARs in the announcement sample; however, these that undervaluation provides the information that results are not consistent across the models instigates trading ahead of the events. NVSE, which investigated. This indicates that the hypothesis that distinguishes between deals where the BMTV of the unprofitability acts as a signal cannot be supported. In target company is negative. Therefore this offers the rumoured sample no statistically significant some support to the notion that investors have the relationship is detected. capacity to spot a likely takeover ahead of its announcement. The relationship is demonstrated with 5.2.4 Growth Resource Mismatch CARs ahead of the release of information in the rumoured sample. In the event where the growth in turnover is Of the three models employed, the model that mismatched with the assets which the company has at uses the CAR as the dependent variable offers the its disposal then there may be an increased likelihood most explanation as demonstrated through R squared of takeover. There are two types of firms that can values. Across both samples these results are become likely targets, the first has low growth but consistent with the market model delivering greater holds high resources, and the second demonstrates explanatory power than the constant mean return high growth while holding few resources. The benchmark. inclusion of this variable is founded upon Maris’s Results reported in Table 5 are the coefficients (1964) inefficient management hypothesis which and standard errors (in parentheses) taken from OLS allows that the market offers a mechanism by which regressions showing the statistical significance of the poorly managed firms can be transferred to the correlations between the pre-bid price run-up and control of more capable managers. A number of publicly discernible variables that reportedly signify studies have pointed to possibility of this as an increased probability that companies are more likely identifying factor (Palepu 1986, North 2001, Barnes to be targeted for takeover. Each of the dependent 1999). When considering this in a multivariate cross- variables expresses the price run up for the in sectional regression framework no specific sign is percentage terms. Cumulative abnormal returns hypothesised as managers could target firms (CAR) are used up to day t=0. Run-Up Indexes 1 and displaying attractive growth prospects or instead 2 express the cumulative abnormal returns on each could hone in on resource rich firms. For this analysis day prior to the event as a proportion of the entire findings show that as a signal it demonstrates a abnormal returns recorded over the event window statistically significant but slight relationship with the period. The total returns to calculate Run-Up Index 1 second run-up index variable (RU2) in the rumoured cover the entire period running up to day t= 10, for sample. No relationship is detected with the run up Run-Up Index 2 abnormal returns up to day t = 0 are before the intention to merge or acquire a firm is used. All prices used to generate returns were officially announced. The mismatch between growth obtained from DataStream. Observable factors that and resources and its relationship to run-up ahead of are reported in the literature to increase the likelihood

38 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

that a company becomes target for a takeover are the year’s turnover by the total assets for that year. The independent variables in the regressions. Insider Book to market value for the firm (BMTV) is trading (IT) is a dummy variable which takes the generated by dividing the Shareholder’s equity of the value of 1 if declared insider trading occurred within firm by its market capitalisation; this information was the company in a three month period before the event obtained from the previous year’s annual reports for occurred. The information on insider trades was each firm. NVSE represents the instances when the sourced from a sample of declared insider trades firm’s shareholders hold negative equity in the provided by Directors Deals. The variable MC company, taking the value of 1 if this is the case and represents the logged pre deal market capitalisation 0 otherwise. Previous year’s annual reports for each for the targeted firm in the takeover. The net profit company in the sample were used to obtain figures margin (NPM) is computed by dividing the profit that were used to construct the preceding variables. before tax for the firm by its turnover. GRM is a The table below also reports the R squared value for variable taken to represent the growth resource each regression. mismatch; this is calculated by dividing the previous

Table 5. Controlling for other factors attributed to Pre bid target price run-up around Announcement date

Panel A: Constant Mean Return Model Controlling factor CAR Run-Up Index 1 Run-Up Index 2

Intercept -0.5036 746.1637 -1.5257 (0.3916) (675.6857) (1.3016)

IT 0.2004 19.3506 -0.4108 (0.2784) (480.39) (0.9254)

MC 0.04651 -59.2877 0.2257** (0.0340) (58.5870) (0.1129)

NPM 0.0000*** 0.0042 0.0000 (0.0000) (0.0243) (0.000)

GRM 0.0000 -108.787 0.0000 (0.0000) (174.3221) (0.0000)

BMTV -0.0430 21.5953 0.0341 (0.0303) 52.2444 (0.1006)

NVSE 0.0000 0.0000** 0.0000 (0.000) (0.000) (0.000) R-square 0.1712 0.1332 0.0820 Panel B: Market Model Controlling factor CAR Run-Up Index 1 Run-Up Index 2

Intercept -0.2329 -0.1048 0.8991 (0.3606) (0.5600) (0.6973)

IT 0.3512 0.1479 -0.4758 (0.2291) (0.3215) (0.4430)

MC 0.0169 0.0656 0.0034 (0.0314) (0.0475) (0.0608)

NPM 0.0000*** 0.0000 0.0000 (0.0000) (0.0000) (0.0000)

GRM -0.0114 0.1834 -0.0994 (0.0867) (0.1340) (0.1677)

BMTV 0.0179 0.0137 -0.0067 (0.0224) (0.0331) (0.0433)

NVSE 0.0000 0.0000 0.0000 (0.2531) (0.0000) (0.0000)

R-square 0.2066 0.0605 0.0233 *, **, *** denote statistical significance at 10%, 5% and 1% respectively

39 Risk governance & control: financial markets & institutions / Volume 1, Issue 3, 2011

The results recorded in Table 6 investigate the trading occurred within the company in a three month statistical significance of the hypothesised period before the event occurred. The data for this relationship between a measure of pre-bid price was obtained from a sample of declared insider trades change activity and publicly observable factors that provided by Directors Deals. MC refers to the logged have been shown to signify increased probability that pre deal market capitalisation for the target firm. each firm in the sample is likely to become the target NPM is the net profit margin which is calculated of a takeover. The dependent variables are variant through dividing the profit before tax for the expressions of the price run up for the sample company by the turnover for the company. GRM expressed in percentage terms. CAR represents the refers to the growth resource mismatch and is pre-event cumulative abnormal returns up to day t=0. computed by dividing the previous year’s turnover by The Run-Up Indexes are the measure of pre-event the total assets for that year. BMTV is the Book to abnormal returns as a proportion of the total abnormal market value for the firm calculated from the returns over the event window period. With Run-Up previous year’s annual reports for each firm. All Index 1 the total abnormal returns used finishes at day accounting figures in addition to the event dates for t =10, while with Run-Up Index 2 the abnormal each Merger or Acquisition were obtained from returns are used only up to day t =0 in the event Zephyr and were taken from the annual reports in the window. Price information used to calculate returns year immediately prior to the event. NVSE takes the was obtained from DataStream. The independent value of 1 if shareholders equity value is negative and variables in the regressions represent the observable 0 if this is not the case. Regression coefficients factors that reportedly increase the probability that the together with standard errors (in parenthesis) are firm to which they are related becomes a more likely reported for each variable in each of the regressions. target for takeover. Insider trading (IT) is a dummy The R squared value for each of the regressions is variable which takes the value of 1 if declared insider reported below.

Table 6. Controlling for other factors attributed to Pre bid target price run-up around Announcement/Rumour date

Panel A: Constant Mean Return Model Controlling factor CAR Run-Up Index 1 Run-Up Index 2 Intercept 0.1990 0.5650 0.9106 (0.4145) (0.5584) (0.6169)

IT -0.0773 0.2888 -0.0343 (0.2537) (0.3418) (0.3776)

MC -0.0050 0.0215 -0.0336 (0.0367) (0.0500) (0.0550)

NPM 0.0000 0.0000 0.0000 (0.0000) (0.0000) (0.0000)

GRM -0.1737 0.1739 0.3376** (0.0997) (0.1343) (0.1484)

BMTV 0.0718** 0.0166 0.0240 (0.0239) (0.0322) (0.0356)

NVSE 0.0000** 0.0000* 0.0000 (0.0000) (0.0000) (0.0000)

R-square 0.1705 0.1107 0.1090

Panel B: Market Model Controlling factor CAR Run-Up Index 1 Run-Up Index 2

Intercept -0.1997 1.1748*** 0.7698 (0.3261) (0.2578) (0.5435)

IT 0.0318 -0.2085 -0.3019 (0.2062) (0.1630) (0.3436)

MC 0.0176 -0.0215 -0.0103 (0.0292) (0.0231) (0.0486)

40 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

NPM 0.0000 0.0000 0.0000 (0.0000) (0.0000) (0.0000)

GRM -0.0477 0.0191 0.2367** (0.0685) (0.0542) (0.1142)

BMTV 0.0549*** 0.0048 0.0239 (0.0191) (0.0151) (0.0318)

NVSE 0.0000*** 0.0000 0.0000 (0.0000) (0.0000) (0.0000)

R-square 0.2142 0.0606 0.0990

*, **, *** denote statistical significance at 10%, 5% and 1% respectively.

6 Conclusions activity is more plausible ahead of a rumour date because rumours are generated by market watchers This study investigated the pricing behaviour of who notice unusual activity in firms stocks (Pound successfully completed takeover target firm stocks in and Zeckhauser 1990), in the instances where the period surrounding the first announcement or rumours turned out to be true the change noticed appearance of a rumour pertaining to the impending could have been informed. event. Having identified abnormal pricing behaviour As the presence of rumours in the media does it then goes on to rule out other possible explanations not appear to influence abnormal returns the that could account for this phenomenon. temptation is to ascribe the price behaviour to To the best knowledge of the author, this study unidentified insider trades. However, much of the is the first to specifically examine pricing behaviour literature points to the possibility that signals are around both announcement and rumour dates present which appear when a firm experiences a specifically relating to Mergers and Acquisitions in heightened probability of being targeted for takeover. the ten year period since FMSA (2000). It goes Of the variables included in the cross sectional further than previous studies of its kind in that it regression, profitability and value offer the greatest assesses the effectiveness of regulation and possibility that CARs can in part be generated by enforcement initiatives through employing a wider externally observable signals. event window. It also analyses CARs to investigate There is however a lack of consistency across whether these could be attributable to a combination samples and models used, so it would be imprudent to of publicly observable signals. offer full support for the idea that other reasons In both samples the results indicate that there behind creating the abnormal returns can be appears to be activity that suggests informed trading eliminated for consideration. occurs prior to the public release of the information. It is also interesting to note that declared insider The appearance of CARs forerunning announcements trades do not appear to bear any strong relationship is something which has been well documented in the with the price run-up prior to the information release. literature (Seyhun 1992, Meulbrook 1992, Jarrell and Those occupying key positions that would have Poulsen 1989, Korczak et al 2010, among others). access to price sensitive information do not visibly The FSA itself has produced work documenting a trade on this knowledge. As known insiders do not price run-up ahead of announcements (Dubow and carry out the trades it is clear that the private Monteiro 2006). information becomes diffused from its decision The influence of the media rumours on CARs making source. This may occur in a number of ways, can be ruled out when a sample is constructed in the UK, a number of successful prosecutions have consisting of rumour dates and announcement dates been brought against individuals who have received for which no previous indication of an impending bid tip offs that a takeover may occur or who have had exists. The presence of a run-up prior to a rumour access to this knowledge while acting as an date is something which has been confirmed by intermediary. Furthermore the opportunities to Pound and Zeckhauser (1990). However as resulting disseminate the information to individuals to act on CARs are similar for both samples this means that the behalf of the senior people within a company who assertions by Neely (1987) , Keown and Pinkerton are compelled to disclose trading activities still (1981) and Jarrell and Poulsen’s (1989) that the run- remain. Although the legislation is designed to take up is in some way a product of the media generated these conditions into consideration following up on rumour can be ruled out. The similarity in CARs is every suspicious trade and tracing the information on explained by the fact that in both samples the which this was based back to its source would be an majority of event dates are the same and that in all almost impossible task. firms the takeovers are eventually completed. Insider

41 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

In conclusion, this study demonstrates that 14. Brar, G, Giamouridis D. and Liodakis, M. (2009) despite enhanced legislation and powers of Predicting European takeover targets. European enforcement insider trading ahead of mergers and Financial Management. Vol. 15, Issue no. 2, pp. 430– acquisitions continues to occur in the UK. In light of 450. this evidence the efficacy of the current system for 15. Bris, A. (2005) Do insider trading laws work? European Financial Management. Vol. 11 Issue no. 3 preserving market integrity must come into question. pp. 267–312. The question policymakers are left with is, if market 16. Brown, S. and Warner, J. (1985) Using daily stock abuse legislation proves ineffective should the returns: The case of event studies. The Journal of response be to strengthen legislation and/or to endow Financial Economics. Vol. 14 pp. 3–31. the FSA with greater powers? Certainly a situation 17. Cole, M. (2008) Speech given at Cambridge where disequilibrium of information between symposium on economic crime. 1 September 2008. investors is allowed to exist that will serve only to 18. Cudd, M. and Duggal, R. (2000) Industry increase the cost of equity for all market participants. distributional characteristics of financial ratios: An acquisition theory application. Financial Review. Vol.

35 Issue no. 1 pp. 105–120. References 19. Dubow, B. and Monteiro, N. (2006) Measuring market cleanliness, FSA Occasional Paper. March 1. Ambrose, B. W. and Megginson, W. L. (1992) The 20. Durnev, A. A. and Nain, A.S. (2007) Does insider role of asset structure, ownership structure, and trading regulation deter private information trading? takeover defences in determining acquisition International evidence, Pacific-Basin Finance Journal. likelihood. Journal of Financial and Quantitative Vol. 15, Issue no. 5, pp. 409-433. Analysis. Vol. 27 Issue no. 4 pp. 575–89. 21. Fidrmuc, J.P. Goergen, M. and Renneboog, L. (2006) 2. Asquith, P. (1983) Merger bids, uncertainty, and Insider trading, news releases and ownership stockholder returns. Journal of Financial Economics. concentration, The Journal of Finance, Vol. 61 Issue Vol. Vol. 11, No. 1-4, pages 51-83. no. 6 pp. 2931-2973. 3. Bainbridge, S. (2000) Insider trading: An overview. 22. Finnerty, J. E. (1976) Insiders and market efficiency, Encyclopaedia of Law and Economics. Vol. 3 pp. Journal of Finance, Vol. 31, Issue no. 4 pp. 1141–48. 772–812 23. Financial Services Authority, (2011) Fines table, 4. Bainbridge, S. (1995) Incorporating state law fiduciary Available at www.fsa.gov.uk . Accessed October 12th duties into the federal insider trading prohibition. 2011. Washington & Lee Law Review. No. 52 pp. 1189. 24. Garfinkel, J. (1997) New evidence on the effects of 5. Banerjee, A. and Eckard, E.W. (2001) Why regulate federal regulations on insider trading: The insider insider trading? Evidence from the first great merger trading and securities fraud enforcement act wave (1897-1903).The American Economic Review. (ITSFEA), Journal of Corporate Finance, Vol. 3 Issue Vol. 91 Issue no. 5 pp.1329-1349. no. 2 pp. 89-111. 6. Barnes, P. (1999) Predicting UK takeover targets: 25. Gilbert, A. Tourani-Rad, A. and Wisniewski, some methodological issues and an empirical study. T.P.(2007) The impact of regulatory change on insider Review of Quantitative Finance and Accounting. Vol. trading, Management International Review, Vol. 27, 12 Issue no.3 pp. 283–301. Issue no. 5, pp. 745-765. 7. Barnes, P. (2000) The identification of UK takeover 26. Givoly, D. and Palmon, D. (1985) Insider trading and targets using published historical cost accounting data: the exploitation of inside information: some empirical Some empirical evidence comparing logit with linear evidence. The Journal of Business. Vol. 58 Issue no. 1 discriminant analysis and raw financial ratios with pp. 69-87. industry-relative ratios. International Review of 27. Gort, M. (1969) An economic disturbance theory of Financial Analysis. Vol. 9 Issue no. 2 pp. 147–62. mergers, Quarterly Journal of Economics, Vol. 83 8. Becker, G. S. (1993) Nobel Lecture: The economic Issue no. 4 pp. 624–42. way of looking at behaviour. The Journal of Political 28. Hasbrouck, J. (1985) The characteristics of takeover Economy. Vol. 101 Issue no. 3 pp. 385-409. targets and other measures, Journal of Banking & 9. Bettis, J. C. Coles, J.L. and Lemmon, M.L. (2000) Finance. Vol. 9 Issue no. 3, pp. 351-362. Corporate policies restricting trading by insiders, 29. Jagonlinzer, A.D. and Roulstone, D.T. (2009) The Journal of Financial Economics, Vol.57 issue no.2 pp. effects of insider trading regulation on trade timing, 191-220. litigation risk, and profitability. Unpublished working 10. Bhattacharya, U. and Daouk, H. (2002) The world paper. Stanford University. price of insider trading. The Journal of Finance. Vol. 30. Jarrell, G. A. and Poulsen, A. B. (1989) Stock trading 57 Issue no. 1 pp. 75–108. before the announcement of tender offers: insider 11. Bhattacharya, U. and Daouk, H. (2009) When no law trading or market anticipation? Journal of Law, is better than a good law. Review of Finance. Issue no. Economics and Organization. Vol. 5 Issue no. 2 pp. 13 pp. 577-627. 225-48. 12. Bialkowski, J, Gottschalk, K and Wisniewski, T. 31. Jarrell, G. Brickley, J. and Netter, J. (1988) The (2008) Stock market volatility around national Market for Corporate Control: The Empirical Evidence elections. Journal of Banking & Finance. Vol. 32 Issue Since 1980, Journal of Economic Perspectives. Vol. 2 no. 9 pp. 1941-1953. Issue no. 1 pp. 49-68. 13. Bradley, M. (1980) Interfirm tender offers and the 32. Jensen, M.C. and Ruback, R. S. (1983) The market for market for corporate control. Journal of Business. Vol. corporate control: The scientific evidence, Journal of 53 Issue no. 4 pp. 345. Financial Economics. Vol. 11 No. 1-4 pp 5-50. Kennedy, V. A. and Limmack, R. J. (1996) Takeover

42 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

activity, CEO turnover, and the market for corporate 46. Powell, R. and Yawson, A. (2005) Industry aspects of control, Journal of Business Finance and Accounting, takeovers and divestitures: Evidence from the UK. Vol. 23 Issue no. 2 pp. 267–85. Journal of Banking & Finance. Vol. 29, Issue no. 12 33. Keown, A.J. and Pinkerton, J. M. (1981) Merger pp. 3015-3040. announcements and insider trading activity: An 47. Powell, R. G. (2004) Takeover prediction models and empirical investigation, The Journal of Finance. Vol. portfolio strategies: a multinomial approach. 36 Issue no. 4 pp. 855-869. Multinational Finance Journal, Vol. 8 Issue no. 12 pp. 34. Korczak, A. Korczak, P. and Lasfer, M. (2010) To 35–74. trade or not to trade: The strategic trading of insiders 48. Rider B, Alexander, K, Linklater, L. and Bazely, S. around news announcements, The Journal of Business (2009) Market abuse and insider dealing. London; Finance and Accounting. Vol. 37 Issue no. 3-4, pp. Tottel publishing. 369-407. 49. Roulstone, D. T. (2003) The relation between insider- 35. Kuehn, D. (1975), Takeovers and the theory of the trading restrictions and executive compensation, firm, London. Journal of Accounting Research, Vol. 41 Issue no. 3 36. Macey, J. R. (1991) Insider trading: economics, pp.525-552. politics, and policy. Washington: AEI Press. 50. Rossi, S. and Volpin, P. F. (2004) Cross-country 37. MacKinlay, C. A. (1997) Event studies in economics determinants of mergers and acquisitions. Journal of and finance. Journal of Economic Literature. Vol. 35 Financial Economics. Vol. 74, Issue no. 2 pp. 277-304. Issue no.1 pp. 13-39. 51. Seyhun, N. (1992) The effectiveness of insider-trading 38. Manne, H. G. (1966) Insider trading and the stock sanctions. Journal of Law and Economics. Vol. 35 pp. market. New York: Free Press. 149–182. 39. Marris, P. (1964) The experience of higher education. 52. Seyhun, N. (2000) Investment intelligence from London: Routledge & Kegan Paul. insider trading. MIT Press. MA. 40. Meulbroek, L. (1992) An empirical analysis of insider 53. Shleifer, A. and Vishny, R. W. (2003) Stock market trading, The Journal of Finance. Vol. 47 Issue no.5 driven acquisitions, Journal of Financial Economics. pp.1661-1699. Vol. 70, Issue no. 3 pp. 295-311. 41. Meulbroek, L. K. and Hart, C. (1997) The effect of 54. Singh, A. (1971) Takeovers. Cambridge University illegal insider trading on takeover premia. European Press. Finance Review. Vol. 1 Issue no.1 pp. 51-80. 55. Singh, A. (1975) Takeovers, economic natural 42. Neely, W. P. (1987) Banking acquisitions: acquirer selection and the theory of the firm: Evidence from the and target shareholder returns. Financial Management. postwar United Kingdom experience. The Economic Vol. 16 Issue no. 4 pp. 66-74. Journal. Issue no. 85 pp. 497-515. 43. North, D. S. (2001) The role of managerial incentives 56. Song, M. H. and Walking, R. A. (1993). The impact of in corporate acquisitions: the 1990s evidence. Journal managerial ownership on acquisition attempts and of Corporate Finance. Vol. 7 Issue no. 2 pp.125-149. target shareholder wealth. Journal of Financial and 44. Palepu, K. (1986) Predicting takeover targets: A Quantitative Analysis. Vol. 28 Issue no. 4 pp. 439-457. methodological and empirical analysis. Journal of 57. Symington, J. (2008) The FSA and enforcing the Accounting and Economics. Vol. 8, Issue no. 1 pp. 3- market abuse regime. Speech given at City and 35. financial market abuse conference. 6 November 2008. 45. Pound, J. Zeckhauser, R. (1990) Clearly Heard on the London. Street. The Journal of Business. Vol. 63 Issue no. 3 pp. 291-308.

43 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

Appendix

Table A. Deal List

Outlined below is the constituent list for the sample of announced mergers and the sample of the rumoured/announced mergers together with the release dates of the information. The announced sample list contains dates on which the impending takeover was officially released. For the Rumoured/Announced sample list the dates relate to the day the news reached the public domain, the announcement date is used when no rumour for the impending deal emerged prior to the official release of the information.

Announced Rumoured/Announced

1 Lloyds Banking Group plc 03/11/2009 1 Lloyds Banking Group plc 10/08/2009

2 Xstrata plc 29/01/2009 2 Xstrata plc 29/01/2009

3 Fitbug Holdings plc 19/10/2010 3 Fitbug Holdings plc 19/10/2010

4 Burberry Group plc 17/11/2005 4 Burberry Group plc 14/10/2004

5 Stagecoach Group plc 10/09/2004 5 Stagecoach Group plc 10/09/2004

6 Foreign & Colonial Investment Trust plc 02/07/2004 6 Foreign & Colonial Investment Trust plc 02/07/2004

7 Billiton plc 18/03/2001 7 Billiton plc 18/03/2001

8 Mecom Group plc 09/03/2007 8 Mecom Group plc 09/03/2007

9 3i Group plc 08/05/2009 9 3i Group plc 27/04/2009

10 Hammerson plc 09/02/2009 10 Hammerson plc 09/02/2009

11 Premier Foods plc 12/07/2006 11 Premier Foods plc 12/07/2006

12 Segro plc 04/03/2009 12 Segro plc 18/02/2009

13 Raven Russia Ltd 31/03/2006 13 Jarvis 17/04/2005

14 Intermediate Capital Group plc 02/07/2009 14 Raven Russia Ltd 31/03/2006

15 National Express Group plc 11/11/2009 15 Intermediate Capital Group plc 02/07/2009

16 William Hill plc 27/02/2009 16 National Express Group plc 06/05/2009

17 Aga Foodservice Equipment 19/10/2007 17 William Hill plc 12/02/2009

18 Cookson Group plc 29/01/2009 18 Aga Foodservice Equipment 06/07/2007

19 Grainger plc 05/11/2009 19 Cookson Group plc 29/01/2009

20 Inchcape plc 19/03/2009 20 Grainger plc 05/11/2009

21 Northgate plc 08/11/2010 21 Inchcape plc 12/01/2009

22 ACP Capital Ltd 20/03/2007 22 Northgate plc 08/11/2010

23 Quintain Estates and Development plc 05/11/2009 23 Genting Singapore 24/11/2004

24 Gasol plc 15/12/2005 24 ACP Capital Ltd 20/03/2007

25 Marston's plc 18/06/2009 25 Quintain Estates and Development plc 18/03/2009

26 Lupus Capital plc 19/03/2007 26 Gasol plc 15/12/2005

27 Biocompatibles International plc 19/11/2010 27 Marston's plc 18/06/2009

28 Avis Europe plc 25/06/2010 28 Lupus Capital plc 19/03/2007

29 Mecom Group plc 22/05/2009 29 Biocompatibles International plc 20/09/2010

30 Wichford plc 20/06/2005 30 Avis Europe plc 25/06/2010

31 Galliford Try plc 10/09/2009 31 Mecom Group plc 23/03/2009

32 Burford Capital Ltd 24/11/2010 32 Wichford plc 20/06/2005

33 Borders & Southern Petroleum plc 26/11/2009 33 Galliford Try plc 10/09/2009

34 Speedy Hire plc 28/05/2009 34 Burford Capital Ltd 24/11/2010

35 European Nickel plc 18/05/2006 35 Borders & Southern Petroleum plc 26/11/2009

36 GTL Resources plc 08/08/2005 36 Speedy Hire plc 28/05/2009

37 Northgate plc 10/07/2009 37 European Nickel plc 18/05/2006

44 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

38 Workspace Group plc 27/01/2009 38 GTL Resources plc 08/08/2005

39 BowLeven plc 09/06/2009 39 Northgate plc 10/07/2009

40 Conygar Investment Company plc 15/09/2009 40 Workspace Group plc 14/01/2009

41 Jarvis plc 27/05/2005 41 BowLeven plc 09/06/2009

42 Zimbabwe Platinum Mines Ltd 30/06/2003 42 Conygar Investment Company plc 15/09/2009

43 Sterling Energy plc 14/08/2009 43 Zimplats Holdings 30/06/2003

44 John David Group plc 11/05/2005 44 Sterling Energy plc 14/08/2009

45 Blackstar Investors plc 20/07/2006 45 John David Group plc 11/05/2005

46 Wichford plc 05/08/2009 46 Blackstar Investors plc 20/07/2006

47 Conygar Investment Company plc 29/01/2007 47 Paypoint 16/11/2003

48 Cookson Group plc's plastic mouldings businesses 02/01/2002 48 Conygar Investment Company plc 29/01/2007

49 Finsbury Emerging Biotechnology Trust plc 02/05/2006 49 Cookson Group plc's plastic mouldings businesses 02/01/2002

50 Rockhopper Exploration plc 26/10/2009 50 The Biotech Growth Trust 10/03/2006

51 Blackstar Investors plc 03/01/2006 51 Blackstar Investors plc 03/01/2006

52 Synchronica plc 29/03/2007 52 Synchronica plc 15/03/2007

53 Cove Energy plc 18/09/2009 53 Cove Energy plc 18/09/2009

54 Brammer plc 06/10/2009 54 Brammer plc 06/10/2009

55 Midas Income & Growth Trust plc 27/01/2006 55 Midas Income & Growth Trust plc 04/01/2006

56 Avesco plc 29/03/2007 56 Avesco plc 29/03/2007

57 National Bus Company 03/09/2004 57 National Bus Company 03/09/2004

58 Johnson Service Group plc 11/06/2008 58 Johnson Service Group plc 11/06/2008

59 E2V Technologies plc 29/10/2009 59 E2V Technologies plc 29/10/2009

60 Real Estate Investors plc 20/12/2006 60 Real Estate Investors plc 20/12/2006

61 Energiser Investments plc 24/12/2009 61 Energiser Investments plc 24/12/2009

62 PSG Solutions plc 30/04/2009 62 PSG Solutions plc 20/03/2008

63 Vertu Motors plc 28/05/2009 63 Vertu Motors plc 28/05/2009

64 President Petroleum Company plc 30/09/2010 64 President Petroleum Company plc 30/09/2010

65 Regent Inns plc 20/10/2009 65 Clyde Process Solutions plc 23/03/2007

66 Clyde Process Solutions plc 23/03/2007 66 Vernalis plc 11/02/2010

67 Vernalis plc 11/02/2010 67 Discover Leisure plc 19/06/2003

68 Discover Leisure plc 07/06/2007 68 Rurelec plc 13/12/2005

69 Rurelec plc 13/12/2005 69 Cove Energy plc 11/03/2010

70 Cove Energy plc 11/03/2010 70 Phytopharm plc 03/12/2009

71 Phytopharm plc 03/12/2009 71 Vernalis plc 29/04/2009

72 Vernalis plc 29/04/2009 72 SkyePharma plc 01/09/2008

73 SkyePharma plc 01/09/2008 73 Islamic Bank of Britain plc 27/07/2010

74 Islamic Bank of Britain plc 27/07/2010 74 Nwide.Accid.Repr.Svs. 14/03/2002

75 Colliers Cre plc 06/10/2009 75 Colliers CRE plc 30/09/2009

76 Rexam plc's two food flexibles businesses 18/03/2002 76 Rexam plc's two food flexibles businesses 18/03/2002

77 Western & Oriental plc 15/05/2007 77 Western & Oriental plc 15/05/2007

78 Hampson Industries plc 01/12/2006 78 Hampson Industries plc 01/12/2006

79 Millwall Holdings plc 20/03/2006 79 Millwall Holdings plc 20/03/2006

80 Tottenham Hotspur plc 05/07/2007 80 Tottenham Hotspur plc 07/06/2007

81 Fitzwilliam Capital plc 27/07/2005 81 Getmobile 27/07/2005

82 First Property Online plc 28/11/2000 82 First Property Online plc 28/11/2000

83 Lupus Capital plc 03/03/2006 83 Lupus Capital plc 01/02/2006

45 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

84 Catalyst Media Group plc 12/03/2007 84 Blavod Black Vodka plc 15/09/2003

85 Blavod Black Vodka plc 23/12/2003 85 Sterling Energy plc 22/12/2003

86 Sterling Energy plc 25/09/2003 86 Asset Management Investment Company plc 26/10/2010

87 Asset Management Investment Company plc 07/12/2010 87 Impax Group plc 02/08/2006

88 Impax Group plc 02/08/2006 88 Millwall Holdings plc 17/11/2010

89 Millwall Holdings plc 17/11/2010 89 MJ Gleeson Group plc's Internal Plant Hire Operations 01/07/2005

90 MJ Gleeson Group plc's Internal Plant Hire Operations 01/07/2005 90 Matra Petroleum plc 03/11/2005

91 Matra Petroleum plc 13/03/2006 91 Axis-Shield plc 07/10/2009

92 Axis-Shield plc 07/10/2009 92 API Group plc 17/12/2007

93 API Group plc 17/12/2007 93 Watford Leisure plc 14/03/2006

94 Argo Real Estate Opportunities Fund Ltd 02/10/2009 94 Ventus Vct 05/04/2005

95 Watford Leisure plc 14/03/2006 95 Xtract Energy plc 07/08/2006

96 Xtract Energy plc 07/08/2006 96 Jarvis Securities plc 31/03/2009

97 Jarvis Securities plc 31/03/2009 97 Ventus Vct 06/04/2006

98 Optare plc 03/09/2009 98 Core Vct IV 05/04/2007

99 Omega Diagnostics Group plc 17/11/2010 99 Delcam plc 30/03/2007

100 Delcam plc 30/03/2007 100 Brainspark plc 28/12/2001

101 Golden Prospect Precious Metals Ltd 09/04/2009 101 Forum Energy plc 02/07/2008

102 Brainspark plc 28/12/2001 102 Kiotech International plc 03/11/2006

103 Forum Energy plc 25/07/2008 103 Kiotech International plc 03/10/2005

104 Diamondcorp plc 15/03/2010 104 Inditherm plc 20/11/2003

105 Kiotech International plc 03/11/2006 105 SynAIRgen plc 27/05/2009

106 African Copper plc 21/05/2009 106 Noble Investments (UK) plc 04/11/2005

107 Inditherm plc 20/11/2003 107 Eredene Capital plc 10/04/2006

108 SynAIRgen plc 27/05/2009 108 Capital Management & Investment plc 11/03/2010

109 Noble Investments (UK) plc 04/11/2005 109 Summit Corporation plc 11/12/2009

110 Eredene Capital plc 10/04/2006 110 Silverdell plc 31/03/2009

111 Capital Management & Investment plc 01/04/2010 111 Matra Petroleum plc 07/07/2009

112 Summit Corporation plc 11/12/2009 112 FFastFill plc 24/02/2005

113 Chalkwell Investments plc 01/11/2010 113 Accumuli plc 01/04/2004

114 Kiotech International plc 05/08/2009 114 Global Energy Development plc 09/11/2010

115 Silverdell plc 07/04/2009 115 Archipelago Resources plc 06/06/2005

116 Matra Petroleum plc 07/07/2009 116 Archipelago Resources plc 30/09/2004

117 FFastFill plc 01/04/2004 117 Octopus Eclipse Vct 4 05/10/2005

118 Accumuli plc 09/11/2010 118 Anglo Pacific Group plc 07/10/2004

119 Global Energy Development plc 06/06/2005 119 SWP Group plc 01/04/2004

120 Archipelago Resources plc 24/12/2004 120 Vyke Communications plc 22/04/2010

121 Gemfields Resources plc 21/10/2008 121 Mid-States plc 12/05/2010

122 Anglo Pacific Group plc 07/10/2004 122 Octopus Eclipse Vct 2 18/03/2005

123 SWP Group plc 30/04/2004 123 Densitron Technologies plc 20/11/2003

124 Vyke Communications plc 22/04/2010 124 Inditherm plc2 13/12/2006

125 Mid-States plc 12/05/2010 125 Gasol plc2 25/07/2006

126 Lombard Medical Technologies plc 09/01/2009 126 Energy Technique plc 29/03/2004

127 Densitron Technologies plc 20/11/2003 127 Zoo Digital Group plc 30/07/2007

128 Inditherm plc 13/12/2006 128 Proximagen Neuroscience plc 05/06/2009

129 Gasol plc 25/07/2006 129 Archipelago Resources plc 27/01/2010

46 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

130 Energy Technique plc 30/03/2004 130 Tottenham Hotspur plc 06/11/2008

131 Zoo Digital Group plc 30/07/2007 131 Framlinton Aim VCT 2 19/12/2005

132 Access Intelligence plc 18/11/2004 132 Goldstone Resources 20/09/2005

133 Proximagen Neuroscience plc 05/06/2009 133 Downing Abst Vct 1 30/09/2001

134 Cyan Holdings plc 14/08/2008 134 Clinical Computing plc 30/10/2007

135 Archipelago Resources plc 27/01/2010 135 Greggs plc 16/04/2004

136 Empresaria Group plc 30/04/2009 136 Condor Resources plc 27/06/2008

137 Tottenham Hotspur plc 06/11/2008 137 Artisan (UK) plc 22/06/2009

138 Sirius Exploration plc 14/07/2009 138 Infoserve Group plc 29/01/2010

139 GoldStone Resources Ltd 20/09/2005 139 Energy Technique plc 30/09/2005

140 Downing Healthcare Protected Venture Capital Trust 30/09/2001 140 Abbeycrest plc 28/08/2009

141 Clinical Computing plc 30/10/2007 141 Woburn Energy plc 10/12/2008

142 Greggs plc 16/04/2004 142 Feedback plc 05/06/2007

143 Condor Resources plc 27/06/2008 143 Metrodome group 01/08/2007

144 Watermark Global plc 20/05/2009 144 Property Recycling Group plc 24/02/2009

145 eXpansys plc 29/05/2009 145 Energy Technique plc 13/04/2006

146 Artisan (UK) plc 22/06/2009 146 Works Media Group plc 23/11/2006

147 Infoserve Group plc 29/01/2010 147 Energy Technique plc 29/03/2001

148 Energy Technique plc 30/09/2005 148 Arcontech Group plc 21/09/2009

149 Abbeycrest plc 28/08/2009 149 CLS Holdings plc 12/06/2009

150 Woburn Energy plc 17/12/2008 150 Oak Holdings plc 02/03/2010

151 Feedback plc 05/06/2007 151 Pittards plc 23/12/2009

152 Property Recycling Group plc 24/02/2009 152 African Copper plc 09/05/2006

153 Energy Technique plc 13/04/2006 153 UMC Energy plc 16/10/2009

154 Works Media Group plc 23/11/2006 154 Cookson Group plc 11/03/2009

155 Energy Technique plc 29/03/2001 155 Messaging International plc 22/03/2007

156 Arcontech Group plc 21/09/2009 156 1pm plc 12/03/2010

157 CLS Holdings plc 12/06/2009 157 Deltex Medical Group plc 28/10/2003

158 Oak Holdings plc 02/03/2010 158 Dominion Energy 11/01/2006

159 Pittards plc 23/12/2009 159 Charles Street Capital plc 30/06/2010

160 African Copper plc 09/05/2006 160 1pm plc 24/10/2007

161 UMC Energy plc 16/10/2009 161 Servoca plc 06/03/2009

162 Cookson Group plc 11/03/2009 162 Coolabi plc 19/05/2006

163 Caspian Holdings plc 22/05/2009 163 Radicle Projects plc 15/12/2009

164 Messaging International plc 22/03/2007 164 John David Group plc 19/01/2006

165 1pm plc 12/03/2010 165 Highams Systems Services Group plc 30/09/2008

166 Deltex Medical Group plc 28/10/2003 166 Parallel Media Group plc 11/09/2006

167 Charles Street Capital plc 30/06/2010 167 Solo Oil plc 16/11/2009

168 LP Hill plc 21/08/2009 168 RGI International Ltd 25/03/2009

169 1pm plc 24/10/2007 169 Eatonfield Group plc 28/10/2009

170 Servoca plc 06/03/2009 170 Tri-Star Resources 19/10/2009

171 Coolabi plc 19/05/2006 171 Deo Petroleum plc 05/05/2010

172 Radicle Projects plc 15/12/2009 172 Metrodome Group plc 23/08/2007

173 John David Group plc 19/01/2006 173 Capcon Holdings plc 01/07/2009

174 Highams Systems Services Group plc 30/09/2008 174 Millennium & Copthorne Hotels plc 21/03/2003

175 Parallel Media Group plc 11/09/2006 175 Archipelago Res.4 21/11/2003

47 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

176 Solo Oil plc 16/11/2009 176 Johnston Press plc 14/05/2008

177 RGI International Ltd 25/03/2009 177 Pathfinder Minerals plc 05/03/2010

178 Eatonfield Group plc 28/10/2009 178 Caledonian Trust plc 12/10/2004

179 Mastermailer Holdings plc 02/04/2009 179 Mediazest plc 17/08/2009

180 Deo Petroleum plc 05/05/2010 180 White Young Green plc 30/10/2009

181 MeDaVinci plc 08/07/2009 181 Jersey Electricity Company Ltd 30/11/2004

182 Metrodome Group plc 16/05/2008 182 Specialty Scanners plc 09/09/2003

183 Capcon Holdings plc 01/07/2009 183 Xstrata plc 20/02/2002

184 Millennium & Copthorne Hotels plc 21/03/2003 184 Avis Europe plc 25/09/2002

185 Johnston Press plc 14/05/2008 185 Workspace Group 01/06/2001

186 Pathfinder Minerals plc 05/03/2010 186 Zimplats Holdings Limited 30/06/2005

187 Caledonian Trust plc 12/10/2004 187 St. James's Place Capital plc 21/03/2006

188 Mediazest plc 17/08/2009 188 Hightwz Group 27/03/2006

189 White Young Green plc 06/01/2010 189 F&C Capital and Income Investment Trust plc 22/05/2006

190 Leed Petroleum plc 06/11/2009 190 St. James's Place plc 25/10/2006

191 Rangers Football Club plc 23/04/2003 191 British Airways plc 12/11/2009

192 Hot Tuna (International) plc 13/08/2009 192 Redstone plc 24/08/2010

193 KleenAir Systems International plc 13/11/2009 193 Archipelago Res.6 24/01/2011

194 Jersey Electricity Company Ltd 30/11/2004

195 Specialty Scanners plc 09/09/2003

196 Avis Europe plc 25/09/2002

197 Workspace Group 01/06/2001

198 National Express Group plc's Airlinks coach business 07/01/2005

199 Zimplats Holdings Limited 30/06/2005

200 St. James's Place Capital plc 21/03/2006

201 F&C Capital and Income Investment Trust plc 22/05/2006

202 Mitchells & Butlers plc's 21 pubs 31/08/2006

203 St. James's Place plc 25/10/2006

204 British Airways plc 08/04/2010

205 London & Stamford Property Ltd 05/08/2010

206 Associated Network Solutions plc 23/09/2004

207 Redstone plc 24/08/2010

48 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

THE CASH-CDS BASIS FOR SOVEREIGN COUNTRIES: MARKET STRATEGY, PRICE DISCOVERY AND DETERMINANTS

Alessandro Carboni* and Andrea Carboni**

Abstract

We study the cash-CDS basis and its implication for market strategies and price discovery, together with the role of credit risk common factors. A positive net income is derived with a negative basis, once funding costs are considered. There exists an arbitrage opportunity for Greece in 2010, with a negative basis of more than 100 bp. Our comparison with three different basis shows that while converging markets seem adopt the same strategy, in particular for Portugal, Ireland and Greece. Results for price discovery show that the CDS market moves ahead the bond market. Finally, our empirical analysis shows that the global risk factor contributes to increase the basis, while the banking sector vulnerability proxy offers a negative contribution.

JEL Classification: G00, G10, G12, G14

Keywords: Credit Default Swaps, Asset Swap, Price Discovery, Basis, Limits To Arbitrage

* Corresponding author, University of Siena, Piazza San Francesco, 8, 53100, Siena, Italy Fax: +39 0577 232757 Tel.: +39 0577 232756-5105 E-mail: [email protected] ** University of Siena, Piazza San Francesco, 8, 53100, Siena, Italy

1 Introduction value of a specified reference asset, with the protection buyer that pays a periodic fee (spread) or a Credit risk indicators have received much attention one-off premium to a protection seller, while the during the financial crisis that began in the summer of protection seller makes the payment when a credit 2007. The bail-out of Lehman Brothers (15° event occurs. The premium is set as a percentage September 2008) has shown the importance of amount of protection bought. A CDS can be viewed financial market liquidity and has demonstrated that as an insurance contract against a risky event on a risk management is dangerous if inappropriately reference entity. A simple CDS structure is shown in used. During the crisis OTC credit derivatives came Figure 1. under attack because they were identified as the main According to (ISDA, 2003) credit events can be contributors to the widespread turmoil, creating a new classified in: 1) Bankruptcy; 2) Obligation kind of dimension, namely counterparty credit risk. acceleration; 3) Obligation default; 4) Failure to pay; Hence, the need to provide more information through 5) Repudiation / Moratorium and 6) Restructuring. the creation of trade reporting for regulatory The contract provides protection against credit authorities, as suggested for example by (Banque de events that can even occur before the end of the France, 2010) and (IFSL, 2009), as well as to contract. In this case, there is the settlement payment understand what is the most informative credit risk made by the seller according to the contract indicator, especially during a crisis. On the other settlement option. hand, it could be also interesting to go into market Credit derivatives specify physical or cash perspective, by analyzing both trading strategies for settlement. In the physical settlement [1], on credit risk and their technical issues. Once arbitrage occurrence of a credit event, the buyer delivers the opportunities exist, they could be affected by some reference asset to the seller, in return for which the common factors, as suggested for example by seller pays for the face value of the delivered asset to (Carboni and Carboni, 2010) and (Ejsing and Lemke, the buyer (Choudhry, 2006). The contract may 2009), among others. specify a number of alternative assets (called Credit default swaps (CDS) are the most deliverable obligations) that the buyer can deliver [2]. common type of credit derivatives. A CDS is a When more than one deliverable obligation is bilateral contract that provides protection on the par specified, the buyer will invariably deliver the

49 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

cheapest asset on the list of eligible assets: this asset swap is a combination of an interest rate swap provides the concept of cheapest-to-deliver option, and a bond, and it is used to alter the cash flow which is an embedded option afforded by the profile of the underlying security: an investor can buy protection buyer [3]. On the other hand, in the cash for example a fixed rate bond and then hedge out settlement option, the contract specifies a (almost all of) the interest rate risk by swapping the predetermined payout value when a credit event fixed payments into floating ones. Hence, the investor occurs. Generally, the protection seller pays the buyer takes on only the credit risk on the new security, the difference between the nominal amount of the which is equivalent to buy a floating rate note issued default swap and the final (market) value of the by the same entity [6]. For assuming this credit risk, reference asset, determined by means of a poll of the investor earns a corresponding excess spread dealer banks. This last value can be viewed as the known as the asset swap spread. recovery value of the asset [4]. For a simple example Following (Bomfin, 2005), the typical terms of we follow (O'Kane and Sen, 2004). Suppose that an this agreement are as follows. The market calls this investor sells protection on $10mm notional to a 5- contract par asset swap: year horizon on a credit risky issuer with a spread of  The investor (the asset swap buyer) agrees to buy 200 bp. The buyer pays approximately $50,000 every from the dealer (the asset swap seller) a fixed-rate quarter. The payments stop if the issuer defaults prior bond issued by the reference entity, paying par for to maturity, when the protection delivered by the the bond, regardless of its market price. seller is par minus the recovery rate. If we assume a  The investor agrees to make periodic payments 40% recovery rate, than the investor would lose equal to the coupon of the reference entity to the $6mm. The CDS spread is the spread which seller. In return, the dealer agrees to make floating determines the cash flows paid by the buyer of the rate payments (based on a fixed spread over contract. In this sense, the spread is the compensation LIBOR) and the notional principal is the same as for taking the risk of incurring in the loss given the par value of the reference bond. In this case, default, when a credit event occurs. In mathematical the interest rate swap embedded in the contract terms, the spread is the sum that makes the expected allows the investor to receive LIBOR (L in the present value of the two lags the same, at the figure) plus a spread (A, the asset swap spread) origination of the contract [5]. against the payment of a fixed coupon. According to (O'Kane and Sen, 2004), the CDS  As in any other kind of swap, the spread A is set spread is the best measure of credit risk for at least so that at the initiation, both legs have the same four reasons. First, the CDS contract is most pure expected value, namely, the market value of the credit risk, while the asset swap incorporates both asset swap contract is zero. This has an important credit and interest rate risk. Second, it is a swap drawback. On the one hand, if the reference bond contract, where the stream of cash flows ceases to is trading below par, dealers must be compensated exist following a credit event, compensating the for selling the bond to the asset swap buyer for buyer according to the settlement rule. Third, this is a less than its market value. On the other, when a flexible measure, so it can be possible to buy or sell bond is trading above par, A must be such that the CDS without restrictions. Fourth, the CDS market is dealer's position in the swap has a sufficient relatively liquid, so CDS spreads should reflect the positive value at the initiation of the swap. This market price of risk. allows a compensation for the dealer who sells the Asset swaps are a common form of derivative bond "at a loss". contracts written on fixed-rate debt instruments. An

Figure 1. Credit Default Swap

50 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

During the life of the contract, there are two swap lives on, the investor loses the source of funding possibilities. If the entity does not default, at maturity from the coupon, as well as the claim on the par value date, the investor receives the par value of the of the bond. He receives only the bond's recovery reference bond, while the interest rate swap vanishes value upon the entity's default. An alternative option with no exchange of notional principals. Otherwise, if in the event of default may be the termination of the default occurs during the life of the contract, the asset swap with an opposite sign operation.

Figure 2. Asset Swap Contract

Source: Bomfin (2005)

We have called A the asset swap spread. But swap spread can be viewed as a measure of credit what is its meaning as a credit risk indicator? quality. We can anticipate this useful relation: The rest of the paper is organized as follows: Section 2 deals with technical frameworks for the PPLIBOR FULL pricing of CDS and asset swap spread, also providing (1) A  the definition of the no-arbitrage relation; Section 3 PV 01 offers a description of the trading strategy for the

negative basis; Section 4 presents an econometric where PLIBOR is the value of the bond's cash flows analysis between CDS and asset swap spread, in discounted at LIBOR, PFULL is bond's market price, terms of price discovery, on the one hand, while a while PV01 is the LIBOR discounted value of 1 bp study of the main determinant of the basis, on the coupon stream. other. Finally, Section 5 concludes. An appendix with If the asset defaults immediately after the data description, time series graphs and tables is initiation of the swap, the investor (who has paid 100 provided at the end. for the asset swap) has an asset which can be sold at its recovery value R in the market, and an interest rate 2 Theoretical Framework swap that is worth 100 - P, with P the full price of the bond. Hence, the investor's loss is (100 - R) - (100 - 2.1 Credit Default Swap Spread P) = P – R, namely the difference between the bond full price and its recovered value. If the price of the Like any other swap, a CDS consists of two lags, one asset swap is par, then the loss on immediate default related to the expected value of the premiums paid by is 100 - R, similar to a default swap. the protection buyer and the other related to the If we hold the credit quality of the asset constant expected value of payments in the case a credit event and increase its price, by using for example a higher occurs. When a credit event occurs, the payoff of a coupon, the loss on default is greater and the asset CDS at time t is usually the face value of the swap spread should increase. On the other hand, let reference obligation minus its market value, as we us assume to allow credit quality to change, fixing have already described above. When the recovery rate both the LIBOR curve and the coupon. In this case, is not zero, (Hull and White, henceforth HW, 2000) bond price falls only because of an increase in the stated that the best assumption about the claim made issuer's credit risk and vice-versa. Therefore, an FULL by the bondholders in the event of default is that this increase in the bond price P in (1) reduces the claim equals the face value of the bond plus accrued asset swap spread and vice-versa. That is why asset interest [7]. Therefore, the payoff from a typical CDS is:

51 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

Contrarily to (Duffie and Singleton, 1997), (2) L  RL 1  At   L  1  R  A t R  , (Jarrow and Turnbull, 1995) and (Lando, 1998) who where L is the notional principal, R is the recovery use the hazard rate, (HW, 2000) use the default rate and A(t) is the accrued interest on the reference probability density to express their relations. obligation at time t calculated as a percent of the face They assume q(t) constant and equal to qi during value. (HW, 2000) evaluate a single name CDS by period (ti-1, ti) and set: assuming that default events, Treasury interest rates and recovery rates are mutually independent and by ti   (5)ijv ( t ) F j ( t )  RC j ( t ) dt also assuming that there is no counterparty default ti1   risk [8]. They provide a two stages procedure. In the , first, they calculate the risk-neutral probability of default at future times from the yield on bond issued where v(t) is the present value of $1 received at time t by the same reference entity, while in the second, with certainty, Fj(t) is the forward price of the j-th they evaluate the expected present value of both lags bond for a forward contract maturing at time t,  in the CDS. assuming that the bond is default free; R is the As in (Choudhry, 2006) and (HW, 2000), we expected recovery rate for holders of the j-th bond in introduce some notation: the event of default at time t and Cj(t) is the claim  T: Life of the CDS made by the holders of the j-th bond, if there is  q(t): Risk-neutral default probability density at default at time t. (HW, 2000) extract qj by inverting time t the total present value of the losses on the j-th bond:   R : Expected recovery rate of the reference j obligation in the risk-neutral world (6) G B  q   u(t): Present value of payments at the rate of $1 j j i ij i1 per year on payment dates between time zero , and time t  e(t): Present value of an accrual payment at time obtaining: j1 t for the period (t - t*), where t* is the payment G B  q  date immediately preceding time t j ji1 i ij (7) qi   v(t): Present value of $1 received at time t  jj ,  w: Total payments per year made by the protection buyer where Bj is the price of the bond today, Gj is the price  s: Value of w that causes the CDS to have a of the j-th bond today if there is no probability of value of zero (CDS spread) default and β is a parameter to be estimated [9].  π: The risk-neutral probability of no credit event Coming back to pricing, if a credit event occurs during the life of the swap prior to maturity, for example at time (t < T), the  A(t): Accrued interest on the reference present value of the payments is w[u(t) + e(t)], while obligation at time t as a percent of the face if there is no default during the life of the contract, value. the present value is wu(T). Hence, the expected The value of π is one minus the probability that present value of the payments (i.e. premium leg) a credit event will occur by time T. It can be becomes: calculated from the risk-neutral default probability density q(t): T (8)w q ()()() t u t e t  dt  w u () T 0 T (3)  1  q ( t ) dt . 0 Considering the assumptions above, the risk- . neutral expected payoff from the CDS becomes: Before going on to CDS pricing, it is useful to provide some definitions for the risk neutral probability density q(t). (HW, 2000) define q(t)Δt as (9) 1 1 A ( t ) R  1  R  A ( t ) R the probability of default between time t and t+Δt as , seen at time zero. On the other hand, the hazard rate while the present value of the expected payoff from h(t) is the default probability between time t and t+Δt the CDS (i.e. protection leg) is: as seen at time t, assuming no default between time zero and time t. The two variables are related T according to this relation: (10) 1R  A ( t ) R  q ( t ) v ( t ) dt 0   t .  h() d  (4)q ( t ) h ( t ) e 0 .

52 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

The value of the CDS is the difference between profit if the bond were to be resold in the open the present value of the protection and the premium market. The second term is the market value of the leg. We have: embedded IRS, which may have either positive or negative market value. Considering that the market TT (11) 1R  AtRqtvtdt ()  ()()  w qt ()()() ut  et  dt  wuT () value of the bond is given to both the investor and the 0   0 . dealer (as well as LIBOR and bond's coupon), the main issue of negotiation would be the IRS The CDS spread s is the value of the premium w component, namely the spread A over the LIBOR that that allows both legs to be equal, namely allows the will be a part of the floating payment made to the difference in (11) to become zero: investor (Bomfin, 2005). The value of the spread is obtained by imposing the market value of the AP T    equals to zero at inception.  1R  A ( t ) R q ( t ) v ( t ) dt The market value of the bond is: (12) s  0   T q()()()() t u t e t  dt  u T N 0   . (15)VB (0, N ) D (0, i ) C  D (0, N ) P i   i1  If an investor forms a portfolio of a credit , default swap and a T-year par yield bond issued by B the same reference entity, he can replicate the T -year with V (0, N) the market value at time 0 of a fixed- Treasury par yield, in the absence of arbitrage rate bond maturing at time N, C the fixed coupon, opportunity. According to (Choudhry, 2006), this D(0,i) the discount factor, P the face value, while δi is difference is called cash-CDS basis. the accrual factor (for example 0.5 if the bond pays In this case, if y is the yield to maturity (YTM) coupons semiannually). on corporate bond, and x the YTM on Treasury bond, To define the market value of the interest rate we have: swap VIRS(0, N), we can consider that the buyer pays fixed and receives floating: this is equivalent to say (13) y s  x that the buyer sells a fixed-rate bond and buys a . floating-rate one. These considerations allow us to Obviously, if y - s is significantly greater than x, price both fixed and floating legs. it is profitable to buy the T-year par yield bond issued Hence, the fixed leg is equal to: by the reference entity, buy the default swap and sell the T -year Treasury par yield. This is the negative N  (16)VXL (0, N ) D* (0, i ) C P basis strategy suggested by (Choudhry, 2006): an i   i1  investor aims to earn a risk free return by buying and , selling identical credit risk across different markets. On the other hand, if y - s is significantly less than x, with the same coupon, notional principal and it is profitable to short the T-year par yield bond, payment dates as the underlying bond. (Bomfin, short the credit default swap and long the T -year 2005) states that, in this relation, there is not an * Treasury par yield. exchange of notional amounts and that D (0, i) is a different discount factor, reflecting the credit quality 2.2 Asset Swap Spread of the counterparties in the swap. On the other hand, the floating leg is: As we have already mentioned, an asset swap (AP) is N a "package" involving a fixed-rate bond (B) and an LL **  (17)V (0, N ) D (0, i )i  F (0, i  1, i )  A    P interest rate swap (IRS): the first is bought by the i1  investor from the dealer for par, while through the , second the investor can swap fixed for floating cash * flows. Following (Bomfin, 2005), from investor's where A is the spread over LIBOR and F (0, i-1, i) is perspective, the market value of the asset swap is: forward LIBOR as seen at time zero, for a deposit to be made at time i-1 with maturity at time i. Finally, AP B IRS the market value of the IRS for the investor is: (14)VNVNPVN (0, ) (0, )    (0, ) , (18)VNVNVNIRS (0, ) LL (0, )  XL (0, ) where P is the face value of the bond and VY(·) is the . market value of Y, with Y =AP, B or IRS. The value of the AP at inception can be decomposed in two parts. Rewriting (17) as: The first (in squared parentheses) illustrates that even NN  if the buyer pays for par, the market value of the bond LL ***  (19)VN (0, ) DiFiiADiP (0, )i (0,  1, )      i (0, )    could be different, so he can incur either in a loss or a i1 i  1   ,

53 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

VNVNBB* (0, ) (0, ) * (23) A  adding and subtracting D (0, N)P and rearranging N * (19), (Bomfin, 2005) arrives to: i D(0, i )  P i1 . N The par asset swap spread becomes positive for LL **  B* B (20)V (0, N ) 1  Ai D (0, i )   D (0, N )  P V > V , assuming that the discount factors on the i1  . reference entity are lower than the same evaluated on the LIBOR curve: this means that the entity has a Finally, the value of the asset swap spread is lower credit quality than that embedded in LIBOR, obtained by solving the price equation for A, after obtaining a positive spread. Opposite considerations substituting the relations above: are true for a negative spread.

N B** XL 2.3 No-Arbitrage Relation (21) 0V (0, NA ) i DiPV (0, )   (0, NDNP )  (0, ) i1 . Pricing considerations lead us to the definition of the The last two terms represent the present theoretical no-arbitrage relationships between credit discounted value of the cash flows from the bond default swaps and asset swaps. Following (De Wit, underlying the AP, with discount factors constructed 2006) and (O'Kane and McAdie, 2001), we can on LIBOR. Denoting this quantity VB*, we rewrite demonstrate that for an investor who funds himself at equation (21) as: LIBOR, a combined position of buying protection in a CDS and entering into an asset swap is fully hedged N in any state of the world. BB** In both figures, the strategy leads to a credit (22) 0V (0, N )  V (0, N )  Ai D (0, i )  P i1 , risk-free position: the CDS premium should match by which the asset swap spread is derived easily the asset swap spread, assuming that both instruments as [10]: have the same remaining maturity.

Figure 3. No default situation

Source : (De Wit, 2006)

54 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

Figure 4. Default situation.

Source : (De Wit, 2006)

The natural market strategy is to buy the cash and the CDS contracts. According to (J.P.Morgan, bond, swap it in the asset swap market and buy 2009) "Before the 2008-2009 liquidity crisis a buyer protection using CDS. The investor receives floating of a CDS protection on a single name would make coupons and pays the CDS premium. If the difference regular payments of the CDS full running spread to between the two premiums is not zero, as in practice, the protection seller. However, over the last year it there is an arbitrage opportunity. Specifically, when has become common practice for CDS protection to the CDS spread is greater than the spread on the asset be bought with an upfront payment, followed by a swap, we have a positive basis with the credit standard fixed coupon. Buyers of protection are also derivative that trades higher than the asset swap. required to give a proportion of the notional, known When the basis is negative, the credit derivative as the margin, to the dealer to act as collateral". trades tighter than the asset swap. When the difference between CDS and asset swap is As for the CDS-cash basis, if the difference negative, as shown in figure (2) for Greece, an between CDS spread and the asset swap spread is not investor can buy a CDS, repo the bond in order to zero, the arbitrageur can trade across the cash and the reduce the funding costs and enter into a par asset synthetic market realizing a profit. In particular, when swap, realizing a profit. We consider the funding the basis is positive, he sells the cash and sells costs of the bond and the CDS and finally compute protection on that bond, while if the basis is negative the income for this strategy. he makes an opposite strategy. As suggested by (Choudhry, 2006), the CDS-bond basis is usually 3.1 Asset Swap Funding positive, due to the net impact of factors driving the basis, while negative basis lasts for brief periods [11]. The par asset swap requires that the bond have to be An important feature to remember is that the priced at par. However this price can be split into two asset swap contract is a par asset swap. Therefore, components: the bond dirty price P and the remaining there is a need for the cash bond to be priced at or 100-P. We assume that investor funds every two very near par (see both figures above). However, components. The bond dirty price is financed through most corporate bonds trade significantly away from a secured repo. The repo counterparty will lend to the par, with the consequences that: i) the asset swap investor at a reduced rate of LIBOR (L) plus the repo price is an inaccurate measure of credit risk and ii) rate (R), but will also hold the bonds as collateral. the CDS-asset swap measure is an unreliable basis. Typically the counterparty will only lend the investor This is true because when the underlying bond of the a proportion of the bond dirty price, while the asset swap is above par, the swap price will unfunded part is the haircut h. Therefore, the funding overestimate the level of credit risk, while if the bond for repo is: is below par, the asset swap will underestimate the credit risk. (24)Fundingrepo  1  h L  R P /100 . 3 Market Strategy for Negative Basis The haircut must be funded at LIBOR plus and Trade unsecured funding rate (F):

The no-arbitrage relation described above does not consider funding costs related to both the asset swap

55 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

(25)Fundinghaircut  h L  F P /100 (26)Fundingap  L  F 1  P /100 . . The remaining 100-P is financed at the same To sum up, the total funding cost for entering an rate as the haircut: asset swap is:

(27)Funding 1  h L  R P /100    h L  F P /100    L  F 1  P /100 

3.2 CDS Funding

(30)Fundingcds  U L  F  m 1  U F The investor who wants to buy protection through a . CDS contract has to pay an upfront followed by a Comparing equations (27) and (30) it is evident standard fixed coupon. Moreover, he has to post a that CDS funding costs are smaller than asset swap margin for the notional at risk (i.e. the total notional funding costs, because both the CDS margin and minus any upfront) [12]. We assume that both the upfront are fractions of the notional. upfront U and the margin m are financed at LIBOR plus the unsecured funding rate. We also assume that 3.3 Total Funding Costs and Income the accrued interest is zero. The upfront funding cost is: We can now compute the total funding costs and the consequent net income for a negative basis trade. We (28) Fundingupfront  U L  F  assume that the investor uses equal notionals on bond . and CDS. The income from a negative basis trade is The margin funding cost is: the income from the asset swap minus the CDS coupon:

(29)Fundingmar  m 1  U L  F  m 1  U L  m 1  U F . (31) Incomebasis L  S asw  C By combining the funding for the CDS upfront . and for the CDS margin we obtain the total funding The total funding cost for the strategy is: cost for the CDS:

(32)Fundingtot  L  F  1  h F  R P /100    U L  F  m 1  U F  .

The net income from the negative basis strategy is:

Incomenet Income basis  Funding tot (33) S asw  C F m 1 U F U L F  1 h F R P /100.

This equation is obtained by canceling out the Figure (3) shows the dynamics of the 5 year cash- LIBOR on the notional amount. Note that the LIBOR CDS basis together with 6 possible cases of net dependency is very small because it is paid on the annual income from the negative basis strategy. In CDS upfront, which itself is a fraction of the notional. order to evaluate the presence of arbitrage opportunities, we use equation (33) by considering 3.4 An Example for Greece these chosen values:

1) upfront=0.1, haircut=0.1, spread=125bp, margin=0.1 and bond price=80 2) upfront=0.1, haircut=0.1, spread=125bp, margin=0.1 and bond price=99 3) upfront=0.1, haircut=0.1, spread=125bp, margin=0.1 and bond price=102.5 4) upfront=0.1, haircut=0.1, spread=200bp, margin=0.1 and bond price=80 5) upfront=0.3, haircut=0.3, spread=125bp, margin=0.3 and bond price=80 6) upfront=0.1, haircut=0.05, spread=125bp, margin=0.3 and bond price=80.

A positive income is obtained in 5 out of 6 from the strategy requires at least 100 bp of negative cases: during the end of April up to 7 May 2010 the basis [13]. However, our results deserve caution maximum net income obtained is 91.70 bp. This because data on both CDS premia and asset swap confirms the assumption by Choudhry that negative spread are mean values of all contracts on the same basis lasts for brief periods. Even if the basis is entity provided by Datastream [14]. negative since the end of February, it is not sufficient to create a net income: in our example net income

56 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

4 Lead-Lag Analysis interbank rate (Libor or Euribor) and the 3m Eurepo rate general collateral. All data are gathered from An interesting analysis with both CDS and asset swap Datastream. spread could be the study of the lead-lag relations. As We construct the cointegrating vector with three suggested by the empirical literature (Blanco et al., different relations: i) CDS and asset swap spread; ii) 2004), (Zhu, 2004) and (Coudert and Gex, 2010), CDS, asset swap spread and 3m Libor or 3m Euribor; among others, the short-run relation between CDS iii) CDS, asset swap spread, and the difference and bond spread are bivariate. Hence, the Granger between 3m Euribor and 3m Eurepo general causality test does not give a direct answer to the collateral. For i) and ii) we run the VECM and study causality relation. We care about this problem and we the price discovery process, while for iii) we estimate concentrate only on the long-run relation by using a the cointegrating vector and compare the different Vector error correction model (VECM). For our basis. empirical analysis we use 5 year CDS spread and the Relations i) and ii) require the estimation of the mean between asset swap spread with 3-5 and 5-7 following VECMs: year maturities. Our sample spans from the Lehman Brothers bailout to the end September 2010, with a weekly frequency. For interbank rates we use 3m

p p

pCDS,1 t  p CDS ,101,1 t     p AP t     1, j  p CDS , t  j    1, j  p AP , t  j   1 t j1 j  1 p p (34) pAP, t  2 p CDS , t 1   0   1 p AP , t  1    2, j  p CDS , t  j    2, j  p AP , t  j   2 t j1 j  1 and

p p p pCDS,1 t p CDS ,101,12 t    p AP t    Rate t  1    1, j p CDS , t  j   1, j p AP , t  j   1, j Rate t  j  1 t j1 j  1 j  1 p p p pAPt, 2 p CDSt , 1   0  1 p APt ,  1   2 Rate t  1    2, jCDStj p ,    2, jAPtj p ,    2, j Rate tj   2 t j1 j  1 j  1 p p p (35)Ratet3 p CDS , t 1   0  1 p AP , t  1   2 Rate t  1    3, j p CDS , t  j   3, j p AP , t  j   3, j Rate t  j  3 t j1 j  1 j  1 where Rate is the 3m Libor or 3m Euribor. Results indicated in Table (7). Even in this case the CDS from the Johansen lambda trace test in Tables (2) and market is the leader in terms of price discovery, even (3) suggest that in 10 out of 20 countries if there are cases where the bond seems move ahead cointegration holds for i), while in 19 out of 20 of the CDS market. countries for ii). Hence, the inclusion of a liquidity An interesting comparison would be realized by proxy in the traditional cointegrating relation seems computing the basis with the estimated cointegrating restore the long-run relationship. In Table (5) we vector for CDS and asset swap spread, together with report the estimated cointegrating vectors from the 3m Euribor and Eurepo general collateral. For some Johansen methodology and the Dynamic ordinary countries we note a convergence in the three different least squares (DOLS) by (Stock and Watson, 1993). basis, particularly evident for Greece, Ireland and We can see that only for Greece and Portugal the Portugal. For the first one, the period involved is coefficients seem in line with the traditional from February 2010 when the basis starts to become definition of the basis. However, Portugal denotes a negative; moreover, there is a quasi equivalence constant which is significantly different from zero. during end of April - beginning of May and during When the funding costs proxy (3m Libor or Euribor our last period. For Ireland there is a similar pattern: and Repo) is considered, we can see that it is the equivalence is more pronounced during the end of significantly different from zero, while Euribor and April, but also during end of September with an Repo have different signs. extraordinary convergence among our different basis. The price discovery analysis for the traditional Portugal is only involved for the end of April, cointegrating vector is shown in Table (6). When beginning of May. This pattern could enforce the cointegration holds, results seem confirm that on presence of negative basis strategy in the credit risk average the CDS market is the leader in terms of market and this can be confirmed by the case of price discovery, with respect to the bond market. Greece in Graph (3). However, there are cases where the relation is A final exercise is dedicated to the determinants unclear. Estimated results from equation (35) are of the cash-CDS basis with the presence of funding

57 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

costs. Following the spirit of (Carboni and to enlarge the basis, the markets risk aversion does Carboni,2010) and (ECB, 2010), among others, credit not offer a direct contribution, while the banking risk during financial turmoil is explained by three sector vulnerability proxy offers a negative common factors. The first is reasonably approximated contribution, in particular for Ireland. by the iTraxx Europe which could be considered as a global risk factor: an increase in the iTraxx Europe References should create an increase in the basis. The second is the VSTOXX which measures the markets' risk 1. Banque de France (2010), Financial Stability Review, aversion: an increase in the risk aversion should July increase the basis. The third is the ratio of the 2. Blanco, R., Brennan, S. and Marsh, I.W. (2004), “An Eurostoxx equity index for banks over the overall empirical analysis of the dynamic relationship between Eurostoxx equity index in order to deal with banking investment-grade bonds and credit default swaps”, Banco de Espana, Documento de Trabajo, No. 0401. sector prospects: when this ratio falls, the market is 3. Bomfin, A.N. (2005), Understanding Credit assessing more vulnerabilities to the banking sector, Derivatives and Related Instruments, Elsevier then to the rest of the economy; hence the greater the Academic Press. ratio, the lower the basis. 4. Carboni, A. and Carboni, A. (2010), “A note on banks By using weekly data spanning from January and sovereign CDS premia dynamics: market reactions 2009 to September 2010, for the European countries and determinants”, in our sample, we estimate the following equation http://papers.ssrn.com/sol3/papers.cfm?abstract_id=19 through OLS: 62805. 5. Chrouhy, M. (2006), The Credit Default Swap Basis,

Bloomberg Professional. (36) Basis     iTraxx Europe    VSTOXX    Ratio   1 2 3 . 6. Coudert, V. and Gex, M. (2010), “Credit default swap and bond markets: which leads the other?”, Banque de Results from Table (8) suggest that assumptions for France, Financial Stability Review, Derivatives, No. our proxies hold. The global risk factor seems 14. increase the basis, the markets' risk aversion factor is 7. Credit Suisse (2007), “Credit derivatives handbook”, Fixed Income Research. a positive determinant, except for Italy, while the 8. Credit Suisse (2010), “Sovereign CDS primer”, Fixed perception of the riskiness of the banking sector is a Income Research. negative determinant, except for Norway. An 9. De Wit, J. (2006), “Exploring the CDS-bond basis”, interesting case is Ireland with a coefficient for the National Bank of Belgium, Working Paper, No. 104. bank ratio triple with respect to other countries. 10. Duffie, D. (1999), “Credit swap valuation”, Financial However, the presence of residual autocorrelation Analyst's Journal, Vol.55, No. 1, pp. 73 - 87. does not allow us to conclude in a clear-cut way. 11. Duffie, D. and Singleton, K.J. (1997), “An econometric model of the term structure of interest- 5 Conclusions Rate swap yields”, The Journal of Finance, Vol.52, No.4, pp. 1287-1321. 12. Ejsing, J.W and Lemke, W. (2009), “The Janus- This paper explores the cash-CDS basis for sovereign Headed salvation, sovereign and bank credit risk entities during post Lehman Brothers bailout. After a premia during 2008-2009”, ECB Working Paper, No. technical description of both CDS and asset swap 1127. contract as credit risk indicators, pricing 13. Eichergreen, B., Mody, A., Nedeljkovic, M. and considerations lead to the definition of the no- Sarno, L. (2009), “How the subprime crisis went arbitrage relation. Moreover, we explain how to global: evidence from bank credit default swap realize a positive net income, after taking into spreads”, NBER Working Paper, No. 14904 account funding costs for both the asset swap and the 14. European Central Bank (2010), “Monitoring working group: determinants of government bond spread CDS. The example of Greece during 2010 helps dynamics”, Prepared by Banca d' Italia, Banque de explaining that income from negative-basis- strategy France, Banka Slovenije and Ceska narodni banka}, exists. Our study confirms a positive net income with August. a negative basis of at least 100 bp. Through the use of 15. Gonzalo, J. and Granger, C.W.J. (1995), “Estimation the lead-lag analysis, data show that the CDS market of Common Long-Memory Components in is the leader in terms of price discovery of credit risk, Cointegrated Systems”, Journal of Business and even if some countries do not present clear results. Economic Statistics, ERS-2002-23-F\&A. The empirical evidence for the cash-CDS basis 16. Hamilton J.D. (1994), Time Series Analysis, Princeton compared to its fitted versions, once funding costs are University Press. 17. Hasbrouck, J. (1995), “One security, many markets: considered, demonstrates that when different basis determining the contributions to price discovery”, converge, market seems adopt basis strategy. This is Journal of Finance, Vol. 50, No. 4, pp. 1175-1199. true in particular for Portugal, Ireland and Greece 18. Hull, J. and White, A. (2000), “Valuing credit default when the basis is negative. Finally, we investigate swaps I: no counterparty default risk”, Journal of whether common risk factors are useful determinants Derivatives, Vol. 8, No. 1, pp. 29-40. for cash-CDS basis. Our empirical analysis shows that the global risk factor iTraxx Europe contributes

58 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

19. Hull, J. and White, A. (2001), “Valuing credit default Notes: swaps II: modeling default correlations”, Journal of Derivatives,, Vol. 8, No. 3, pp. 12-22. 1. There is a third type of settlement, called digital, 20. IFLS (2009), Derivatives 2009, June. where the seller pays a fixed percentage (decided at 21. ISDA (2003), Credit Derivatives Definitions. the issue of the contract) on the notional. 22. Jarrow, R.A. and Turnbull, S.M. (1995), “Pricing 2. See ISDA (2003) for specific contractual issues. derivatives on financial Securities subject to credit 3. See (Bomfin, 2005), (Choudhry, 2006) and risk”, The Journal of Finance, Vol. 50, No.1, pp. 53- (Jankowitsch et al., 2007) for a more specific 85. reference. 23. J.P. Morgan (2009), “The Bond-CDS funding basis”, 4. Intuitively, for 1 of notional value, the seller pays the Europe Credit Derivatives Research, 25 September loss given default LGD = (1 - RR), where RR is the 2009. market or simply the recovery rate of the reference 24. Lando, D. (1998), “On Cox process and credit risky asset. securities”, Review of Derivatives Research, No. 2, pp. 5. We will come back to the theoretical relation in the 99-120. next section. 25. Longstaff, F.A., Pan, J., Pedersen, L.H. and Singleton, 6. See (O'Kane, 2001). K.J. (2007), “How sovereign is sovereign credit risk?”, 7. In other studies, like for example (Duffie and NBER Working Paper, No. 13658. Singleton, 1997), the value of this claim equals the 26. Mody, A. (2009), “From Bear Stearns to Anglo Irish: value of the bond immediately prior to default, while how eurozone sovereign spreads related to financial (Jarrow and Turnbull, 1995) stated that it is equal to sector vulnerability”, IMF Working Paper, May, No. the value of the bond in the no-default case. 09/108. 8. However, in (HW, 2001) they formulate pricing 27. O'Kane, D. (2001), “Credit derivatives explained, relations overcoming this assumption. market, products and regulation”, Structured Credit 9. See (HW, 2000). Research, Lehman Brothers. 10. In equation (1) we have expressed this notation from 28. O'Kane, D. and McAdie, R. (2001), “Explaining the dealer's position. Moreover, PLIBOR and Pfull basis: cash versus default swaps”, Structured Credit correspond respectively to VB*(0, N) and VB(0, N). Research, Lehman Brothers Fixed Income Research. 11. Even if one could expect the basis to be negative due 29. O'Kane, D. and Sen, S. (2004), “Credit spreads to financing costs associated with the cash bond explained”, Quantitative Research Quarterly, Lehman position. Brothers Fixed Income Quantitative Research. 12. This margin protects the seller in case of default of the 30. Sgherri, S. and Zoli, E. (2009), “Euro area sovereign protection buyer. The seller will return the margin at risk during the crisis”, IMF Working Paper, October, default of the CDS reference entity, at maturity or in No. 09/222. case of unwind. In our case the investor receives 31. Stock, J. H. and Watson, M. W. (1993), “A simple LIBOR on the posted margin. estimator of cointegrating vectors in higher-order 13. The maximum income is realized together with a integrated systems”, Econometrica, Vol. 61, No. 4, pp. negative basis of 167 bp, that falls to 5 bp the day of 783-–820. the rescue package announcement for Greece on 10 32. Zhu, H. (2004), “An empirical comparison of credit May 2010. See J.P. Morgan (2009) for a definition of spreads between the bond market and the credit default the breakeven basis. swap market”, BIS Working Paper, No. 160. 14. Full description of the data is provided in the Appendix.

59 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

Appendix

A. Data Description

Table 1. CDS premia and asset swap spread

CDS Premia Asset swap spread Name Code Name Code AUSTRALIA SEN 5YR CDS AUGVTS5(SM) Australian Government 3-5 Yrs G2T0(ML:ASWPS) AUSTRIA SEN 5YR CDS OEGVTS5(SM) Australian Government 5-7 Yrs G3T0(ML:ASWPS) BELGIUM KINGDOM SEN 5YR CDS BGGVTS5(SM) Austrian Governments 3-5 Yrs G2H0(ML:ASWPS) DENMARK SEN 5YR CDS DNGVTS5(SM) Austrian Governments 5-7 Yrs G3H0(ML:ASWPS) FINLAND SEN 5YR CDS FINLDS5(SM) Belgian Governments 3-5 Yrs G2G0(ML:ASWPS) FRANCE SEN 5YR CDS FRGVTS5(SM) Belgian Governments 5-7 Yrs G3G0(ML:ASWPS) GERMANY SEN 5YR CDS BDGVTS5(SM) Danish Governments 3-5 Yrs G2M0(ML:ASWPS) GREECE SEN 5YR CDS GRGVTS5(SM) Danish Governments 5-7 Yrs G3M0(ML:ASWPS) IRELAND (REP.OF) SEN 5YR CDS IRGVTS5(SM) Dutch Governments 3-5 Yrs G2N0(ML:ASWPS) ITALY SEN 5YR CDS ITGVTS5(SM) Dutch Governments 5-7 Yrs G3N0(ML:ASWPS) JAPAN SEN 5YR CDS JPGVTS5(SM) Finnish Governments 3-5 Yrs G2K0(ML:ASWPS) NETHERLANDS SEN 5YR CDS NLGVTS5(SM) Finnish Governments 5-7 Yrs G3K0(ML:ASWPS) NEW ZEALAND SEN 5YR CDS NZGVTS5(SM) French Governments 3-5 Yrs G2F0(ML:ASWPS) PORTUGAL SEN 5YR CDS PTGVTS5(SM) French Governments 5-7 Yrs G3F0(ML:ASWPS) SPAIN SEN 5YR CDS ESGVTS5(SM) German Federal Governments 3-5 Yrs G2D0(ML:ASWPS) SWEDEN SEN 5YR CDS SDGVTS5(SM) German Federal Governments 5-7 Yrs G3D0(ML:ASWPS) UNITED KINGDOM SEN 5YR CDS UKGVTS5(SM) Greek Governments 3-5 Yrs G2GR(ML:ASWPS) USA - TREASURIES SEN 5YR CDS USGVTS5(SM) Greek Governments 5-7 Yrs G3GR(ML:ASWPS) Irish Governments 3-5 Yrs G2R0(ML:ASWPS) Irish Governments 5-7 Yrs G3R0(ML:ASWPS) Italian Governments 3-5 Yrs G2I0(ML:ASWPS) Italian Governments 5-7 Yrs G3I0(ML:ASWPS) Japanese Governments 3-5 Yrs G2Y0(ML:ASWPS) Japanese Governments 5-7 Yrs G3Y0(ML:ASWPS) New Zealand Governments 3-5 Yrs G2Z0(ML:ASWPS) New Zealand Governments 5-7 Yrs G3Z0(ML:ASWPS) Norwegian Governments 3-5 Yrs G2J0(ML:ASWPS) Norwegian Governments 5-7 Yrs G3J0(ML:ASWPS) Portuguese Governments 3-5 Yrs G2U0(ML:ASWPS) Portuguese Governments 5-7 Yrs G3U0(ML:ASWPS) Spanish Governments 3-5 Yrs G2E0(ML:ASWPS) Spanish Governments 5-7 Yrs G3E0(ML:ASWPS) Swedish Governments 3-5 Yrs G2W0(ML:ASWPS) Swedish Governments 5-7 Yrs G3W0(ML:ASWPS) Swiss Governments 3-5 Yrs G2S0(ML:ASWPS) Swiss Governments 5-7 Yrs G3S0(ML:ASWPS) U.K. Gilts 3-5 Yrs G2L0(ML:ASWPS) U.K. Gilts 5-7 Yrs G3L0(ML:ASWPS) U.S. Treasuries 3-5 Yrs G2O2(ML:ASWPS) U.S. Treasuries 5-7 Yrs G3O2(ML:ASWPS)

Source: CMA and Merrill Lynch. Provider Datastream. Daily data from 15 September 2008 to 27 September 2010.5 year asset swap spread are computed as mean values of asset swaps between 3 and 5 years and between 5 and 7 years.

60 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

B. Graphs

Figure 1. Sovereign 5 years CDS premia (straight line) and asset swap spread (dotted line)

15 September 2008 - 27 September 2010

61 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

Figure 2. Sovereign Basis: 5 years CDS premia vs asset swap spread

15 September 2008 - 27 September 2010.

62 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

Figure 3. Income from negative basis strategy for Greece

January - September 2010. Different parameter values. In the circle net positive income from the strategies labeled with *. See Section (3).

63 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

Figure 4. Cash-CDS basis vs fitted basis

The grey line is the basis CDS vs asset swap spread taking care of funding cost (Euribor vs Repo rate). The dotted line is the same basis with coefficients estimated from Johansen methodology with a constant. The black line is the basis CDS vs Asset swap spread. Weekly data spanned from January 2009 to September 2010.

64 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

B. Tables.

Table 2. Johansen lambda trace test without and with a constant (first and second row, respectively)

Country none 95% crit at most 1 95% crit BELGIUM* 12.71 12.28 0.01 4.07 15.50 20.16 2.33 9.14 GREECE* 13.16 12.28 0.72 4.07 19.57 20.16 2.35 9.14 IRELAND 6.10 12.28 2.22 4.07 9.45 20.16 3.02 9.14 ITALY 10.46 12.28 0.40 4.07 17.59 20.16 6.41 9.14 PORTUGAL* 13.78 12.28 0.72 4.07 21.92 20.16 1.61 9.14 SPAIN 5.32 12.28 0.15 4.07 17.67 20.16 2.77 9.14 FRANCE* 15.80 12.28 0.17 4.07 18.50 20.16 2.75 9.14 NETHERLANDS 12.22 12.28 0.73 4.07 16.96 20.16 4.72 9.14 AUSTRIA* 12.74 12.28 1.42 4.07 25.17 20.16 8.72 9.14 FINLAND 7.25 12.28 0.84 4.07 14.75 20.16 6.39 9.14 SWEDEN* 18.42 12.28 2.75 4.07 21.37 20.16 5.70 9.14 GERMANY* 14.33 12.28 0.90 4.07 18.70 20.16 4.62 9.14 JAPAN 11.68 12.28 0.05 4.07 18.31 20.16 5.79 9.14 UK* 13.25 12.28 0.16 4.07 20.53 20.16 6.83 9.14 USA 20.80 12.28 3.06 4.07 25.03 20.16 7.02 9.14 NORWAY 6.40 12.28 0.03 4.07 11.95 20.16 5.28 9.14 SWITZERLAND 8.24 12.28 0.43 4.07 14.72 20.16 4.87 9.14 AUSTRALIA* 12.02 12.28 2.23 4.07 21.12 20.16 5.39 9.14 NEW ZELAND* 12.68 12.28 1.31 4.07 18.45 20.16 7.07 9.14 DENMARK 8.74 12.28 0.86 4.07 14.48 20.16 4.61 9.14

Cointegrating vectors with CDS premia and asset swap spread. * indicates cases when cointegration holds.

65 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

Table 3. Johansen lambda trace test without and with a constant (first and second row, respectively)

Country none 95% crit at most 1 95% crit at most 2 95% crit BELGIUM* 35.87 24.21 11.34 12.28 0.00 4.07 40.16 35.07 14.99 20.16 2.59 9.14 GREECE* 42.74 24.21 18.28 12.28 0.44 4.07 45.13 35.07 20.46 20.16 1.57 9.14 IRELAND* 44.69 24.21 19.77 12.28 1.36 4.07 45.65 35.07 20.01 20.16 1.59 9.14 ITALY* 33.18 24.21 7.52 12.28 0.81 4.07 39.89 35.07 13.88 20.16 4.76 9.14 PORTUGAL* 46.48 24.21 22.23 12.28 0.64 4.07 51.05 35.07 25.28 20.16 1.63 9.14 SPAIN* 31.70 24.21 5.93 12.28 0.05 4.07 47.61 35.07 20.09 20.16 1.25 9.14 FRANCE* 39.64 24.21 11.64 12.28 0.31 4.07 42.55 35.07 14.33 20.16 2.91 9.14 NETHERLANDS* 35.73 24.21 11.36 12.28 0.70 4.07 41.19 35.07 16.44 20.16 5.39 9.14 AUSTRIA* 34.44 24.21 9.60 12.28 1.95 4.07 49.86 35.07 24.61 20.16 7.14 9.14 FINLAND* 36.25 24.21 10.21 12.28 0.41 4.07 44.95 35.07 18.92 20.16 5.93 9.14 SWEDEN* 39.68 24.21 13.99 12.28 5.05 4.07 52.34 35.07 24.81 20.16 5.21 9.14 GERMANY* 37.11 24.21 11.40 12.28 1.97 4.07 43.51 35.07 17.73 20.16 6.94 9.14 JAPAN 18.16 24.21 5.75 12.28 0.61 4.07 24.86 35.07 11.85 20.16 5.12 9.14 UK* 46.73 24.21 12.06 12.28 2.77 4.07 55.97 35.07 20.61 20.16 9.20 9.14 USA* 42.97 24.21 22.14 12.28 3.11 4.07 49.40 35.07 26.58 20.16 6.45 9.14 NORWAY* 44.81 24.21 11.62 12.28 0.12 4.07 51.13 35.07 17.72 20.16 4.82 9.14 SWITZERLAND* 26.77 24.21 7.31 12.28 0.83 4.07 41.90 35.07 20.79 20.16 4.67 9.14 AUSTRALIA* 28.49 24.21 8.25 12.28 2.10 4.07 40.64 35.07 20.37 20.16 5.60 9.14 NEW ZELAND* 29.33 24.21 11.43 12.28 1.65 4.07 35.20 35.07 17.27 20.16 4.98 9.14 DENMARK* 47.19 24.21 17.29 12.28 0.13 4.07 52.94 35.07 22.92 20.16 4.09 9.14

Cointegrating vectors with CDS premia, asset swap spread and Libor or Euribor. * indicates cases when cointegration holds.

66 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

Table 4. Johansen lambda trace test without and with a constant (first and second row, respectively)

Country none 95% crit at most 1 95% crit at most 2 95% crit at most 3 95% crit BELGIUM* 138.45 40.10 41.69 24.21 10.37 12.28 0.01 4.07 147.68 53.94 50.84 35.07 14.95 20.16 2.43 9.14 GREECE* 130.67 40.10 41.29 24.21 18.38 12.28 0.49 4.07 142.40 53.94 51.58 35.07 19.97 20.16 1.94 9.14 IRELAND* 151.44 40.10 52.14 24.21 19.97 12.28 1.73 4.07 152.68 53.94 52.99 35.07 20.18 20.16 1.85 9.14 ITALY* 125.01 40.10 35.25 24.21 7.74 12.28 0.54 4.07 135.25 53.94 45.48 35.07 13.97 20.16 4.07 9.14 PORTUGAL* 136.60 40.10 46.54 24.21 22.46 12.28 0.77 4.07 154.71 53.94 63.68 35.07 24.53 20.16 2.12 9.14 SPAIN* 129.33 40.10 35.62 24.21 6.16 12.28 0.14 4.07 151.33 53.94 57.53 35.07 18.46 20.16 1.66 9.14 FRANCE* 152.22 40.10 46.20 24.21 12.67 12.28 0.08 4.07 160.90 53.94 54.01 35.07 15.03 20.16 2.32 9.14 NETHERLANDS* 151.05 40.10 40.32 24.21 10.32 12.28 0.59 4.07 158.66 53.94 47.81 35.07 14.91 20.16 5.12 9.14 AUSTRIA* 147.23 40.10 41.75 24.21 9.72 12.28 1.81 4.07 164.03 53.94 56.96 35.07 24.51 20.16 7.11 9.14 FINLAND* 148.86 40.10 40.38 24.21 10.00 12.28 0.38 4.07 158.57 53.94 49.74 35.07 18.45 20.16 5.74 9.14 SWEDEN* 134.21 40.10 40.66 24.21 14.07 12.28 4.31 4.07 150.44 53.94 56.69 35.07 27.16 20.16 4.54 9.14 GERMANY* 148.51 40.10 44.33 24.21 11.57 12.28 1.90 4.07 156.50 53.94 52.05 35.07 17.34 20.16 6.69 9.14 NORWAY* 152.33 40.10 46.72 24.21 8.87 12.28 0.12 4.07 157.36 53.94 51.69 35.07 13.83 20.16 3.81 9.14 SWITZERLAND* 52.99 40.10 25.29 24.21 10.18 12.28 0.78 4.07 69.12 53.94 41.24 35.07 22.12 20.16 7.17 9.14 DENMARK* 140.15 40.10 45.03 24.21 19.04 12.28 0.05 4.07 146.91 53.94 51.78 35.07 23.71 20.16 3.92 9.14

Cointegrating vectors with CDS premia, asset swap spread, Euribor and Repo rates. * indicates cases when cointegration holds.

67 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

Table 5. Estimated cointegrating vectors for different basis

BASIS BASIS WITH LIBOR BASIS WITH LIBOR AND REPO COUNTRY ASW CONST. ASW LIBOR CONST. ASW LIBOR REPO CONST. BELGIUM -4.755*** - 0.713 -1.509*** - 1.428 -14.923*** 23.293*** - -6.822 4.367*** -288.078 2.284 -19.647*** 30.481*** 31.621 -1.306 - -1.473*** -0.079 - GREECE -1.097*** - -0.926*** -1.199*** - -1.083*** 20.282*** -34.327*** - -0.938*** -1.765*** 48.464 -0.981*** 17.095*** -27.546*** -120.826 -0.988*** - -1.012*** -0.111 - IRELAND 0.151 -4.257*** - -0.821*** -9.483*** 14.117*** - -3.534 -37.271*** 2575.468 -0.721*** -8.785*** 13.183*** -27.177 -0.920** -0.096 - ITALY 2.202 -3.739*** - -0.726 -18.208*** 28.970*** - -2.885 2.064*** -128.045 -0.773 -16.276*** 25.934*** -10.082 -1.473*** -0.028 - PORTUGAL -1.517*** - -0.871*** -1.396*** - -3.135 90.683*** -150.089*** - -1.245*** -38.430*** -1.315*** 0.077 -30.679** -1.371*** 12.139*** -19.384*** -84.209 -1.175*** - -1.233*** -0.037 - SPAIN 0.027 -2.433 - -0.988*** -14.563*** 23.097*** - -1.231*** 0.179*** -77.472*** -1.039*** -8.184*** 13.052*** -36.465 -1.119*** -0.008 - FRANCE 3.544*** - 3.068** -0.268 - 1.625** -5.423*** 8.490*** - 4.428** -0.398 27.999 2.386** -6.966*** 10.768*** 28.257 -0.028 - 0.297 -0.077 - NETHERLAND -0.239 -0.746*** - 0.289 -4.922*** 7.435*** - 0.952 -1.827*** 97.194 0.166 -4.501*** 6.816*** -6.727 -0.768 -0.104 - AUSTRIA -8.281*** - -0.620 -1.183*** - 0.893 -10.296*** 15.463*** - -2.133*** -77.742*** -3.380 2.649*** -259.610** -0.117 -5.936*** 8.944*** -37.297** -1.763*** - -1.715*** -0.053 - FINLAND -0.162 -0.438*** - 0.178 -3.326*** 5.000*** - -0.149 -0.456*** 1.665 0.052 -2.771*** 4.176*** -7.855 -0.420 -0.061 - SWEDEN 2.214*** - -0.251 -0.798*** - 0.513 -5.248*** 8.081*** - 2.164*** -1.562 -2.545 1.275** -194.999*** 0.328 -4.691*** 7.246*** -10.619 0.804 0.030 0.088 - GERMANY 0.962 - 0.372 -0.359*** - 0.566*** -2.388*** 3.675*** - 0.566 -0.431*** 12.289 0.474** -2.228*** 3.439*** -5.231 0.271 0.443 0.035 - JAPAN

UK 2.162*** - -0.393 -1.169*** - 1.016*** -46.119*** -0.675 -0.622*** -57.327* -0.173 - -0.297 -0.088 - USA 74.407*** 45.837*** - -16.115*** -5.736*** -241.369** 0.199 -0.175 - NORWAY -0.346 -0.646*** - 0.104* -1.863*** 2.805*** - -0.488 -0.469*** -24.890 0.073 -1.795*** 2.706*** -3.103 -0.140 -0.055 - SWITZERLAN 0.782*** -0.142 - 6.074*** 15.380*** -20.044*** - 1.320*** -0.560*** 50.184*** 3.485** 5.368*** -7.633*** 55.752 1.195** -0.626*** - AUSTRALIA 0.184 -0.936*** - 9.575*** 411.071*** 0.775 -0.981*** 31.113 0.339 - -0.496 -0.588 - NEW ZELAND 10.350*** - 1.971 -1.128* - 3.102 -1.315* 21.011 -0.760 - -1.909*** -0.710*** - DENMARK 0.011 -0.492*** - 0.195 -4.426*** 6.749*** - 0.222 -0.583*** 17.139 0.231 -4.515*** 6.881*** 2.530 -0.414 -0.058 -

CDS coefficients are normalized to one. */**/*** stand respectively for 10%, 5% and 1% significance values.

68 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

Table 6. Price discovery analysis

COUNTRY λ1 λ2 Gonz-Grang HAS1 HAS2 MID BELGIUM 0.0250359990** 0.0285654077*** 1 0.94343 0.48467 0.71405

0.0144291787 0.0153871196 1 0.91754 0.40532 0.66143 GREECE 0.009799481 0.154041295* 1 0.23231 0.99891 0.61561

0.045211540 0.142527980* 1 0.42314 0.95043 0.68679 PORTUGAL 0.122189238** 0.189028945*** 1 0.75463 0.71853 0.73658 0.191784952* 0.348389262*** 1 0.60236 0.84216 0.72226 0.098874919** 0.125101349*** 1 0.94401 0.45425 0.69913 FRANCE -0.029582241*** -0.022155539*** 0 0.59858 0.46168 0.53013

0.0016844543 0.0025448032 1 0.93347 0.79199 0.86273 AUSTRIA 0.023553594*** 0.0106848476*** 0 0.7594 0.3875 0.57345 0.034423913 0.075027280*** 1 0.92617 0.97174 0.94896 0.006922680 0.010378748* 1 0.97271 0.92882 0.95077 SWEDEN -0.056280198*** -0.025757233*** 0 0.39482 0.49893 0.44688 -0.057638970*** -0.025904349*** 0 0.38638 0.49006 0.43822 -0.032573462* 0.002915748 0.08216 0.02891 0.02333 0.02612 GERMANY -0.064987177*** -0.029042301 0 0.14634 0.41851 0.28243

-0.016745861 0.0046220782 0.21631 0.10036 0.00544 0.05290 UK -0.006203019 -0.040927906*** 1 0.96296 0.98724 0.97510 -0.027422671 -0.057589911*** 1 0.81524 0.89024 0.85274 -0.002859413 0.0013273502 0.31703 0.26161 0.20987 0.23574 AUSTRALIA -0.010373012** -0.011250403*** 1 0.72264 0.77148 0.74706 -0.016373036 0.007400989 0.31131 0.32389 0.3525 0.33820 NEW ZELAND -0.000531672 -0.010731823*** 1 0.99999 0.9989 0.99945

-0.014187582 0.007612627 0.34920 0.35468 0.39976 0.37722

Cointegrating vector with CDS and asset swap spread. */**/*** stand respectively for 10%, 5% and 1% significance values

69 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

Table 7. Price discovery analysis

COUNTRY λ1 λ2 Gonz-Grang HAS1 HAS2 MID BELGIUM -0.004917977 -0.007143918* 1 0.99852 0.65686 0.82769 0.0018519978 0.0030698675** 1 0.99732 0.74060 0.86896 0.016915001 0.017089042 1 0.88833 0.96687 0.92760 GREECE -0.022638249 0.005084217 0.18340 0.00662 0.84218 0.42440 -0.013022458 0.002256271 0.14767 0.00428 0.82861 0.41645 0.004582476 0.178740691* 1 0.21431 0.99984 0.60708 IRELAND -0.004594014 -0.002618797 0 0.43606 0.01781 0.22694 -0.000606615 -0.000295620 0 0.32386 0.00030 0.16208 0.001386371 0.0405022487* 1 0.70294 0.99942 0.85118 ITALY -0.006045622* -0.003703919* 0 0.89876 0.00000 0.44938 0.010122336 0.007181742* 0 0.98465 0.30559 0.64512 0.014646430 0.019443774 1 0.80516 0.83794 0.82155 PORTUGAL -0.005736466 -0.003938810 0 0.43431 0.12396 0.27914 0.153036687 * 0.254486022*** 1 0.68582 0.78091 0.73337 0.142946418** 0.188780537*** 1 0.90729 0.52614 0.71672 SPAIN -0.004254044 -0.002268188 0 0.45617 0.00837 0.23227 0.050154740 0.0755999763 1 0.89879 0.76343 0.83111 0.024414644 0.021142810 0 0.94207 0.22881 0.58544 FRANCE -0.017308066** -0.016347816*** 0 0.73966 0.59135 0.66551 -0.011947340** -0.012133498*** 1 0.77656 0.63256 0.70456 -0.008887458 -0.004563276 0 0.40426 0.21311 0.30869 NETHERLANDS-0.018780105** -0.012015779** 0 0.60263 0.22172 0.41218 -0.007139568** -0.005379989** 0 0.72290 0.33783 0.53037 -0.011699924 0.002203484 0.15849 0.04323 0.35916 0.20120 AUSTRIA -0.024828629** -0.011443638*** 0 0.76910 0.39515 0.58213 0.007990630* 0.005733342*** 0 0.96769 0.70229 0.83499 -0.007810859 0.008751398 0.52839 0.56895 0.92404 0.74650 FINLAND -0.022254410** -0.020054708** 0 0.52073 0.35734 0.43904 -0.021209230** -0.019454354** 0 0.53162 0.36790 0.44976 -0.014732322 0.002785220 0.15900 0.03054 0.14233 0.08644 SWEDEN -0.028121912*** -0.012000824** 0 0.36653 0.46009 0.41331 0.0051453428 0.008286678*** 1 0.88197 0.92015 0.90106 -0.016865509 0.011465120* 0.40469 0.64310 0.64061 0.64186 GERMANY -0.016712461* -0.016142433* 0 0.39631 0.68556 0.54094 -0.012871199 -0.014452655** 1 0.45251 0.73713 0.59482 -0.027102902 0.0083721949 0.23600 0.05429 0.00580 0.03005 UK -0.010004775* -0.010713732** 1 0.57702 0.67801 0.62752 -0.024193523** -0.019478528** 0 0.43873 0.55538 0.49706 -0.009813183 0.0017528337 0.15155 0.04686 0.02647 0.03667 USA 0.000106038 0.000355797** 1 0.88265 0.93159 0.90712 -0.000974074 0.002554125 0.72392 0.91991 0.88049 0.90020 -0.028315002* -0.016304000 0 0.24101 0.32615 0.28358 NORWAY -0.007811670** -0.022851394* 1 0.40001 0.39101 0.39551 -0.010461878** -0.026580242 1 0.33464 0.32053 0.32759 -0.007885424 -0.018812974 1 0.31656 0.28237 0.29947 SWITZERLAND -0.056891584* 0.0396195819*** 0.41052 0.84798 0.65247 0.75023 -0.200080687*** 0.0447313788* 0.18272 0.19018 0.05515 0.12267 -0.014492136 -0.002706978 0 0.10808 0.33993 0.22401 AUSTRALIA -0.023176454** -0.016625787*** 0 0.54876 0.60528 0.57702 -0.020219100** -0.016132145*** 0 0.59789 0.65891 0.62840 -0.023750653* -0.001342544 0 0.00807 0.00617 0.00712 NEW ZELAND -0.003995346 -0.010378396* 1 0.94272 0.93711 0.93992 -0.002207595 -0.008560914* 1 0.97564 0.97087 0.97326 -0.03091934* -0.001786780 0 0.01875 0.01123 0.01499 DENMARK -0.046491644*** -0.020793150 0 0.16843 0.13623 0.15233 -0.037398688*** -0.019218362* 0 0.10546 0.18111 0.14329 -0.018926206 0.003278996 0.14767 0.59010 0.09763 0.34387

Cointegrating vector with CDS, asset swap spread and 3m Euribor. */**/*** stand respectively for 10%, 5% and 1% significance values.

70 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

Table 8. Basis vs common risk factors

Country /VariableConstant Δ iTraxx EuropeΔ VSTOXX Δ Ratio R-sq DW STAT LB(1) LB(4)

Austria 1.091 0.631*** -0.247 -2.268** 0.409 1.325 0.001 0.000

Belgium 1.730*** 0.245** -0.049 -0.439 0.106 1.768 0.269 0.058

Denmark 0.534 0.476*** -0.030 -0.796** 0.389 2.033 0.821 0.011

Finland 1.015** 0.236* -0.003 -0.173 0.157 1.557 0.041 0.001

France 1.221** 0.232 -0.011 -0.941* 0.264 1.745 0.244 0.128

Germany 0.815* 0.264** 0.116 -1.410*** 0.459 1.590 0.072 0.005

Greece 1.077 0.666* 0.638 1.760 0.032 1.415 0.007 0.044

Ireland 0.986 0.337 -0.087 -3.505* 0.140 1.233 0.000 0.000

Italy 1.059 0.309 -1.089** -1.750*** 0.148 1.802 0.377 0.099

Netherlands 0.849 0.196** 0.096 -1.185*** 0.341 1.725 0.353 0.072

Norway 1.258** -0.047 -0.568 0.842*** 0.162 1.860 0.659 0.707

Portugal 1.664 0.085 1.461** 0.124 0.101 1.647 0.091 0.000

Spain 1.258 0.106 0.379 -0.908 0.081 1.877 0.561 0.003

Sweden -0.096 0.536*** -0.452 -0.851* 0.333 1.442 0.022 0.002

Switzerland 0.277 -0.077 -0.367 -3.444*** 0.266 1.369 0.014 0.062

Dependent variable: Δ basis. Period: January 2009 - Sept 2010. HAC standard errors. */**/*** stand respectively for 10%, 5% and 1% significance values.

71 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

IN DEFENSE OF SHOCK THERAPY: POST-SOCIALIST TRANSITION OF THE CZECH REPUBLIC

Scott A. Beaulier*, Peter J. Boettke**, Leonid A. Krasnozhon***

Abstract

Popov (2007, 2000), Kolodko (2000), and Stiglitz (1999) argue that a shock therapy approach has a negative effect on post-socialist transition. Their benchmark for shock therapy, however, refers to the debate on the speed of market reforms. We propose that a more meaningful benchmark is the experience of the Czech Republic, Russia, and other transition economies which share similar approach to the market reforms, but have solved political economy problems of credibility and commitment differently. We compare the Czech Republic’s economic, political, and social performance to these benchmarks in all other post-socialist countries since they began their transitions. We find that the Czech transition is a consistent success because the Havel shock therapy has solved the political economy problems of reform’s credibility and state’s commitment to reform****.

JEL classifications: O52, P20, P27, P52

Keywords: The Czech Republic, Shock Therapy, Transition Economies, Post-Communist Countries

* Executive Director and Adams-Bibby Chair of Free Enterprise, Manuel H. Johnson Center for Political Economy, Sorrell College of Business, Troy University, Troy, AL, 36082, USA ** Corresponding author, University Professor, Department of Economics, George Mason University, Fairfax, VA, 22030, USA E-mail: [email protected] *** Department of Political Science, the University of North Texas, Denton, TX, 76020, USA **** We thank Bryan Caplan, Tyler Cowen, Christopher Coyne, Steven Daley, Steven Eagle, and Richard Wagner and seminar participants at the Mercatus Center for comments. David Lipka provided excellent research assistance. We also thank the Fund for American Studies and the Mercatus Center for essential research support. The standard disclaimer applies.

1 Introduction important service in shifting focus from the debate on the speed of market reforms to a new debate on a role Did post-socialist economies that adopted a shock of political economy forces in post-socialist therapy approach to their transition from socialism transition. We agree with Popov (2007) that the fail or succeed? A debate on the speed of the current thinking about post-socialist transition is liberalization (shock therapy versus gradualism) trapped in a mindset of the debate on the speed of the suggests the former (see Popov, 2007, 2000; Boettke, market reforms, while political economy problems 2001; Kolodko, 2000; Stiglitz, 1999). Popular are underestimated. We, however, disagree accounts of Russia’s transition demonstrate that the respectfully with Professor Popov that the speed of Yeltsin shock therapy deteriorated economic, reform has different effect on transition economies political, and social conditions (Hoffman, 2004; depending on a transition stage. It is not really an Goldman, 2003). Others consider the Czech issue of gradualism versus shock therapy, but an issue Republic’s transition a disappointment of shock of reform’s credibility and commitment to reform. therapy in Eastern Europe (Stiglitz, 2002). Kolodko The same speed of reform will have different effects (2000) writes that the expectations of post-communist if reform has different levels of credibility and reformers were too optimistic and individuals in post- commitment. communist countries would have been better off if Moreover, the issue of speed provides a limited market reforms had progressed gradually. point of comparison for assessment of the market Important recent work by Popov (2007) offers reforms. A more reasonable analysis would be a an alternative view on the debate between gradualism juxtaposition of the Czech Republic and other post- and shock therapy. Professor Popov argues that the communist countries which share similar approach to shock therapy has a negative impact on transition the market reforms, but have succeeded or failed to economies in the first stage of transition - recession, solve political economy problems of credibility and while the gradual approach has a positive effect on commitment. It is especially meaningful to use the economic performance in the second stage of Czech Republic as a point of comparison because transition - recovery. His work also performs an popular accounts call the Czech transition a main

72 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

disappointment of the shock therapy in Central 2 The Havel Shock Therapy Europe (Stiglitz, 1999). In our analysis we focus on economic It is often pointed out that the former Czechoslovakia performance and quality of economic, political, and had more favorable economic conditions than other social institutions in the Czech Republic and other post-socialist countries. The former Czechoslovakia, post-communist countries. We also examine political however, had the most socialistic economy of Eastern economy forces which drove the transition in the Europe where 98% of economy was in public Czech Republic and Russia. We use two important ownership (Hazlett, 1996: 98). Freedom of speech points for a comparison of post-communist countries. and occupational choice were also heavily restricted First, the Czech Republic was one of the eight post- in the former Czechoslovakia. There were severe communist countries which joined European Union shortages of durable goods, and most of the growth in (EU) in 2004. The EU membership serves as a very the economy was being driven by military important indicator of economic, political, and social expenditures. development in post-communist country. To become The post-communist countries which had a a member of EU, a country has to comply with the handicap over other transition countries were the EU principles of freedom, democracy, human rights, former Yugoslav republics. The former Yugoslavia and rule of law. In addition, a country has to meet the had already had working market socialism before Copenhagen criteria, according to which a candidate other post-socialist economies started their transition country must be a stable democracy with a strong rule (Leeson and Trumbull, 2006). Nonetheless, both the of law and working market economy. Second, the Czech Republic and Slovakia have emerged from the Czech Republic is the first and only post-communist transition as upper middle-income countries and country that has graduated from the European Bank became members of the EU in 2004, while the only for Development and Construction’s (EBRD) class of former Yugoslav republic that joined the EU was transition economies. The EBRD graduation also Slovenia. The Czech Republic even succeeded in the serves as important indicator of an advanced level of graduation from the EBRD’s class of transition transition economy. economies. EU membership and EBRD graduation Our findings show that the Czech Republic had marked the end of the Czech Republic’s transition a successful transition from socialism. First, the from socialism. The Havel shock therapy that was Czech Republic performed as well as the countries in driving force behind the successful transition the post-communist EU group that outperformed the deserved the main credit. rest of the post-communist countries. Second, the Using Professor Popov’s terminology, the Czech Czech Republic performed well above average even Republic’s went through a typical two-stage in comparison with the former Yugoslav countries transition: first, recession, and, second, recovery. that had already had a more advanced economic Economic collapse or supply-side recession following system, market socialism, before their transition the end of socialism was prominent in every former experiments began. Third, the Havel shock therapy Soviet and Soviet-bloc countries (Shleifer and transformed a socialist country into a stable Treisman, 2005). Like every one of the former democracy with a strong rule of law and working socialist economies, the Czech economy experienced market economy, while many post-communist a contraction in the early 1990s. Many critics of the countries are still in the gray zone of state-managed shock therapy were quick to point to the recession economy mixed with incoherent authority regime. stage as evidence of failed transitions. Stiglitz (2002: The Havel shock therapy was the main political 186) criticized the Havel shock therapy and pointed economy force behind the Czech transition. Unlike out that relative to where the Czech Republic was in the Yeltsin shock therapy, the Havel shock therapy 1989 it had fallen behind. was a success because it solved the political economy The Czech Republic’s contraction was very problem of reform’s credibility and state’s shallow while many other post-socialist economies commitment to reform. Our findings are unapologetic like Russia had much deeper contractions. The Czech to further criticism of either the Havel shock therapy Republic ended its contraction by 1991, while or the Czech Republic’s transition. Russia’s continued another seven years (Leeson and The paper proceeds as follows. Section 2 Trumbull, 2006). Moreover, the Czech economy demonstrates how the Havel shock therapy solved the returned to pre-transition level twice as much faster political economy problems of credibility and as Russia’s did. Why did the Havel shock therapy commitment. Sections 3 and 4 compare the Czech succeed while the Yeltsin shock therapy Republic’s economic performance, democracy, and underperformed? The answer lies in political social conditions to these benchmarks in other post- economy of these reforms. A juxtaposition of the communist countries since they began their Havel and Yeltsin shock therapies serves as very transitions and until the 2004 EU enlargement and the important point of inference to compare these reforms 2008 EBRD graduation, respectively. Section 5 and explain their different outcomes. contains the concluding remarks. The Havel shock therapy was a genuine legitimate reform lead by Vaclav Klaus and Vaclav

73 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

Havel, himself. Unlike any other politicians, Havel president of the former and left the office in 2003. and Klaus were outspoken advocates of laissez faire Since Havel was president of both the former capitalism and liberal democracy. Havel who was a Czechoslovakia and the Czech Republic, the Havel hero of the dissident movement that brought down the shock therapy also affected the Slovak Republic. communist regime had a very strong public support. Thus, the Havel shock therapy had a positive impact Klaus who was the de facto leader of the Czech on both the Czech and Slovak transitions. Why was transition not only cited liberal free market the Havel shock therapy a success? The answer lies in economists like Adam Smith, Friedrich Hayek, the Havel government’s ability to solve political Milton Friedman, and James Buchanan in his public economy problems of reform’s credibility and states’ speeches but also had good personal relationships commitment to the new status quo. with Friedman and Buchanan. Klaus (1997) wrote Popov (2007) asserts that the debate on the that the Havel government saw the nature of the speed of liberalization underestimated a role of state’s reform clearly by accepting “Adam Smith’s institutional capacity for good economic teaching—his vision of a free, democratic, and performance. He argues that whether a state’s efficient society in which the citizen, not an institutional capacity to enforce rule of law is strong enlightened monarch or an elitist intellectual, is or weak can determine a transition success or failure. king”. The legitimate democratic and free market We argue that, primarily, reform’s credibility and platform of the Havel shock therapy sent a strong state’s commitment to reform determined institutional signal to citizens that issues such as corruption, capacity of transition states. At the dawn of transition illegitimate reforms, and government interventionism a regime uncertainty was a central issue of concern would not be potential stumbling blocks for the with general public (Higgs, 1999). Both authority and Czech transition.1 economy regimes were uncertain because a politics of The year 1989 was arguably the most important discretionary power was unpredictable. year of the 20th century. The Soviet bloc was falling Gorbachev’s perestroika that was full of apart because the Soviet satellite states were breaking economic zigging and zagging served as a reminder away. Nobody, however, expected rapid change with of state’s inability to convey any kind of commitment a stronghold communist regime in control of to reform. Liberal intellectuals like Andrey Sakharov Czechoslovakia. But, on November 17th, 1989, a expressed their concerns with regime instability. student demonstration commemorating the 50th They were uncertain that the zigs permitted today anniversary of the Nazi attack on Czech universities would not be superseded by repressive zags turned into a march for “genuine perestroika” tomorrow. Sakharov warned that “today it is (Wheaton and Zdenek, 1992: 47). The student protest Gorbachev, but tomorrow it could be somebody else. served as a tipping point after which opposition There are no guarantees that some Stalinist will not broadened. The Communist regime that was mainly succeed him” (Kaiser, 1991: 245). The fate of instituted by the 1968 Soviet Invasion desperately Gorbachev’s reforms was sealed by his inability to clung to power in the remaining weeks of November, make credible commitment to the status quo change. but, in early December, the Federal Assembly The political instability of failed reforms and deflated established a government of National Understanding expectations on the part of the population undermined to preside over the transition period (i.e. December institutional capacity of the Gorbachev regime 1989 - June 1990). In one remarkable month, (Boettke, 1993). “As Gorbachev moved back and Czechoslovakia went from being a communist forth from one comprehensive reform to another, he country dealing with a minor dissident movement to a became more and more uncertain about subjecting the non-communist government with liberal democrat Soviet Union to the type of shock therapy such and former dissident Vaclav Havel as President and reforms would inevitably necessitate. He also liberal free-market economist Vaclav Klaus as Prime concluded that unless reined in, the reform process Minister (Dlouhy, 2001). would ultimately shrink his powers and those of the During the transition period, Vaclav Havel was Soviet Union over central economic control, thus appointed president of the independent Czechoslovak reducing the Soviet Union to an ineffective economic state. In June 1990, Havel became Czechoslovakia’s entity” (Goldman, 1991: 222). The state’s failure to first post-communist elected president. In January convey the credible commitment to economic 1993, the former Czechoslovakia disintegrated into liberalization that was necessary to reform the Soviet two independent countries: the Czech Republic and system proved to be perestroika’s undoing (Boettke, the Slovak Republic, while Havel remained the 2001: 169). Thus, reformers’ political economy problem that

1 For an alternative and less favorable interpretation of the Havel government and other post-communist Klaus’s tenure, see Sima and Stastny (2000). According to governments faced was far more difficult than Sima and Stastny, Klaus’s libertarian rhetoric did not signaling a credible commitment to reform to the spillover into libertarian policy. We concur with their main population. Reformers also had to solve the basic point, but maintain that while Klaus’s policies were not ideal from a free market, libertarian standpoint, they were effective paradox of governance - establishing binding in producing economic growth given the transitional constraints on their behavior. Moreover, reformers constraints placed on any leader.

74 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

had to solve these two problems simultaneously. An making credible commitment to reform. Figure 1 issue of simultaneity posed the basic reformer’s illustrates the basic reformer’s dilemma. dilemma – solving the paradox of governance and

Figure 1. Reformer’s Dilemma: Credible Commitment Game with Informational Signal

In Figure 1, player 2 is the citizenry and player 1 presence of discretionary policy. It will be the case of is the reformer – government decision-maker. The the Yeltsin shock therapy. government announces a new reform to liberalize Thus, the reformer’s dilemma suggests that a economy. Player 2 must choose to enter the game, In, Pareto-improvement solution is to convey a credible or stay Out. The paradox of governance suggests that commitment to economic reform and establish the state can Renege on made promises and benefit binding constraints. Each of post-communist from confiscating the wealth of the citizenry. Since experiments in market reforms discovers the basic player 2 knows the sequentially rational moves of point: rules must be established that effectively player 1, player 2 will choose to stay Out. In this case constrain discretionary behavior. A transparent legal the reform will stall unless the effective binding system, independent judiciary, and strong rule of law constraints are in place. Given a weak institutional must be in place before private sector economic capacity of transition state, effective binding experimentation can be expected to yield the constraints are very improbable. Popov (2007) argues promised welfare gains. Without these constraints the that, if a rule of law is weak, authoritarian regimes do government has incentive to use its discretionary a better job in maintaining efficient institutions than power to affect reforms. The discretionary behavior democracies. It could be a possible solution to the on the part of the government fails to produce the reformer’s dilemma, unless player 1 can convey stable regime that is necessary for economic information about his credible commitment to the prosperity. Boettke (2001) writes that “whereas the reform. instability of the 1920s in the Soviet Union led to The logic and structure of the commitment game Stalinism, the instability of the late 1980s led to the gives player 1 two options: make a credible or dissolution of the Soviet Union. The liberalization incredible commitment to economic reform. A choice efforts in the 1950s and 1960s at liberalization failed of commitment has the same payoff if the citizenry and were quickly reversed. Even under Yeltsin’s choose to enter the official economy. The credible experiment in free-market shock therapy, the new commitment to reform, however, significantly government has failed to establish the sort of binding reduces the state’s payoff if it decides to renege on its political and legal commitments required”. promises. Based on player’s 1 signal and prior During reform period political and legal rules information about the reformer, the citizenry must must be status quo breaking. Every successful decide to enter economy or stay out. If player 1 transition from socialism like in Poland shows that signals a credible commitment to a status quo change, restitution and lustration can break the status quo player 2 will choose to enter the official economy and effectively. Both rules establish trust between player 2 will commit to reform as in the case of the citizenry and state and bind the latter to the reform Havel shock therapy. In case of incredible effectively. By burning the bridge between the old commitment, citizens can predict that the state is and new regimes, lustration and restitution commit more likely to renege because there are no significant the state to reform in advance. constraints to reduce the state’s payoff from breaking Lustration laws are the most commonly used its promise. The citizenry will choose to stay out if political and legal rules for screening and the state’s commitment lacks credibility like a “prosecuting” former Communist leaders, public

75 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

employees, and candidates for public office (Ellis, new ruling elite without the communist past, and 1996). These laws have several goals: protecting establishing accountability of government. emerging democracy from the old regime, bringing 11 In October 1991 the Czech Republic and most prominent advocate of the lustration who was Slovakia were one of the first post-communist the member of the Russian parliament was countries to adopt lustration. Between 1991 and 1997 assassinated in 1998. Rothbard (1995) writes that “in total number of people who had been screened was attempting to be congenial statesmen, as opposed to 303,504, while 70 percent of them had been screened counter-revolutionaries, the reformers not only failed by the end of 1993 (Williams, 1999; Ellis, 1996). to punish the Communist rulers with, at the least, the Five percent of screenings (15,166 people) received loss of their livelihoods, they left them in place, certificates with positive results and thus barred their insuring that the ruling “ex”-Communist elite would recipients from holding public office (Williams, be able to resist fundamental change.” 1999). In contrast, the Havel government promised and Restitution was also important change in the adopted both restitution and lustration laws that status quo because it gave private property rights to signaled a credible commitment to the economic citizens with the least favorable attitudes towards the liberalization. Klaus (1997) argued that that communist regime. Moreover, a new propertied class advocates of the gradualism had never understood of citizens led to the development of a collective that restitution, lustration, and privatization action by those individuals to seek protection from “irrefutably signaled that transformation is a serious the government interventionism. Thus, restitution thing”. Restitution solved the fundamental political reduced state’s discretionary power effectively. economy dilemma of an economic system that was: Between October 1990 and May 1991 the Havel “A government strong enough to protect property government adopted the restitution laws. By the end rights and enforce contracts is also strong enough to of 1996 over 1.2 million hectares of agricultural land, confiscate the wealth of its citizens” (Weingast, almost a third of the country’s total, had been 1995). returned to private owners (Dlouhy, 2001). The Havel Moreover, the Yeltsin shock therapy was never government also transferred almost one million implemented as originally announced (Boettke, dollars the Czech Jews who lost their property during 2001). In January 1992 the Yeltsin government under the communist regime (Kraus, 1999). the orchestration of Yegor Gaidar embarked on a The Russian state’s failure to institute restitution liberalization reform that was far more ambitious than and lustration signaled citizens that private and public the Gorbachev’s. From the start the Yelstin predation of wealth creation is to be expected, and government moved towards partial liberalization of thus, citizens had to escape to the unofficial economy. prices. For example, sour cream prices were Though the Yeltsin government considered restitution liberalized, but milk prices were not. If in January and even drafted lustration law, the former was never 1992 only fourteen products were under price and introduced and the latter was actually reversed. Soon output controls, by the summer of 1992 that figure after the failed 1991 coup d’état, the Yeltsin had risen to twenty-four. In addition to food, a wide government banned the Communist Party. But in variety of other products were under state control, November 1992 the Constitutional Court ruled that including energy. Moreover, the Yeltsin government the Communist Party could be reestablished at local continued to provide consumer and producer levels in. By the end of 1992 Yeltsin refrained from subsidies while tax revenues continued to slip in the initial idea of reforming the legal system and kept the unofficial economy. Partial price liberalization and ex-KGB agents in the Ministry of Interior, including policy of soft-budget constraints resulted in the most notorious of them, Viktor Cherkesov, who microeconomic inefficiencies at the enterprise level arrested Andrei Sakharov (Goldman, 2003). After that in turn generated macroeconomic imbalances in Yeltsin finally broke his promise of lustration law, the economy. Continued subsidies and growing the Russian parliament passed the anti-lustration law unofficial economy swelled the fiscal responsibilities that prohibited providing public information about the of the Yeltsin government and undermined Russia’s KGB informants and agents. Galina Starovoitova, the fiscal stability. The Yeltsin shock therapy was ill- conceived and contradictory from the start (Boettke, 11 For a review of lustration laws implemented in post- 2001). communist countries, see Ellis (1996). Lustration requires The Yeltsin government clearly lacked binding investigation and removal from public offices of former and credible commitment to the status quo change. In Communist party officials, members of the former secret police, and collaborators. In the Czech Republic and Slovak the wake of the unexpected election results of Republic the lustration law banned former Party officials, December 1993, the Yeltsin government moved members of Czechoslovak Secret Police (StB), the People’s toward a more gradual reform program with major Militia, the National Security Corps, and collaborators from concessions to state enterprises and state farms. holding a wide range of public offices (Leff, 1996). Lustration could also solve another political economy problem - Economic zigging and zagging of the Yeltsin shock Rousseau paradox: good law needs good people and good therapy clearly reminded Gorbachev’s perestroika. It people need good law caused public distrust and regime uncertainty.

76 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

Rothbard (1995: 396-397) writes that “in other words, By the middle of the 1990s a great divide except for the Czech Republic, where feisty free- between the Havel and Yeltsin shock therapies was market economist and Prime Minister Vaclav Klaus obvious. The former solved the political economy was able to drive through rapid change to a genuine problems of credibility and commitment, but the free market, and, to some extent, in the Baltic states, latter failed. While the Czech Republic appeared to be the reformers were too nice, too eager for one of the successes among the transition countries, “reconciliation,” too slow and cautious. The result with the start of recovery in 1993, rapid growth by was quasi-disastrous: for everyone gave lip-service to 1995, and very low unemployment, Russia’s the rhetoric of free markets and privatization, while in economy went in a deep recession (EBRD, 1999). reality, as in Russia, prices were decontrolled while Between 1991 and 2010 the Czech economy grew at industry remained in monopoly government hands”. an average rate of 1.8 percent, while Russia’s average The fate of the Yeltsin shock therapy was sealed by rate of economic growth was .63 percent (WDI, his inability to solve the political economy problem 2011). of reform’s credibility and state’s commitment to reform. The political instability, public distrust, and 3 The 2004 EU Enlargement growing unofficial economy undermined institutional capacity of the Russian state. The EU membership is a very important indicator of Finally, a phasing-in of reforms was another economic, political, and social development in post- important political economy issue because it raised an communist country. It is reasonable to assume that issue of reform’s credibility and state’s commitment each post-communist country which is located in to reform. Unless reformers moved quickly without Europe pursues the EU membership as a political phasing-in, problems of credibility and commitment economy objective of transition. To join EU, every would halt the reform process (Boettke, 2001). The country must meet the EU standards of freedom, phasing-in approach that reminded the Soviet-style democracy, human rights, and the rule of law. In economic planning caused public distrust towards addition, each country has to meet the Copenhagen market reforms. A rapid transition without phasing-in criteria, according to which a candidate country must sent a strong signal to citizens about a degree of be a working market economy with a stable discretionary policy. Boettke (1993) argues that the democracy and strong rule of law. most successful transition economies will be the ones In 2004 EU welcomed eight post-communist who minimize the omnipresence of state the most. If countries to become EU members. The post- state’s role in economy becomes weaker, it leaves communist EU group included three former Soviet more economic freedom for private sector. republics (Estonia, Latvia, and Lithuania), four The Czech privatization that went without former satellites of the USSR (Poland, the Czech phases signaled that government interventionism Republic, Hungary and Slovakia), and a former would be limited. And the Czech economy greeted Yugoslav republic (Slovenia) (EU, 2004). We use the the status quo change by a quick recovery in the 2004 EU enlargement as a meaningful point of middle of the 1990s. In contrast, the Yeltsin shock comparison between the Czech Republic and other therapy, including privatization, had several official post-communist countries of the EU group. phases. The phasing-in of the Russian privatization Figure 2 shows the EU group’s per capita GDPs signaled that the government intervention would be in 2005 purchasing power parity (PPP) dollars as unlimited. To avoid government embezzlement, the compiled by the World Bank (2011). Leeson and Russian economy mainly moved to the black market. Trumbull (2006) recommend using the year just prior Rothbard (1995) writes that “the reformers didn’t to transition as the base year to control for dates of move fast enough, worrying about social disruption, transition start. Here we follow Roland (2000) in and not realizing that the faster the shift toward dating transitions. Thus, the base year for Hungary freedom and private ownership took place, the less and Poland is 1989, for the Czech Republic and would be the disturbances of the transition and the Slovakia is 1990, for the FSU group is 1991, and so sooner economic and social recovery would take on. Overall, Figure 2 describes a change in per capita place.” income in the first eighteen years of transition.

77 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

Figure 2. GDP Per Capita in the EU Group, First 18 Years of Transition (constant prices)

Sources: Calculated from World Bank, World Development Indicators 2011.

From Figure 2 we can draw a number of former Soviet Union (FSU) countries. Here we follow insights. First, the Czech Republic started transition Popov (2007) in dividing transition in two stages: with the second highest per capita income, second recession and recovery, but we separate between start only to Slovenia, and joined the EU in the same of recovery and end of recovery - return to pre- position. Second, the Czech Republic’s and transition level. We can also make several important Slovenia’s contractions were very shallow while observations from Figure 3. First, soon after the other post-socialist economies like the Baltic States shallow contraction, the Czech economy started had much deeper contractions. It is very important recovering on the fourth year of the transition. fact that the Czech economy holds up quite well in Second, the Czech economy returned to the pre- comparison with Slovenia’s economy because the transition level on the sixth year of transition. As latter had much more advanced economic system (i.e. compared to the Czech Republic, Russia had much socialist market economy) before the start of deeper contraction. It took Russia’s economy eight transition. years to reach the second stage of transition and Figure 3 summarizes data about start of another seven to reach pre-transition level. Thus, the recovery and return to pre-transition level. Presented Czech economy returned to pre-transition level twice data include not only the EU group but also the as much faster as did Russia’s.

Figure 3. Economic Recovery of Post-Communist Countries (years after transition start)

Sources: Calculated from World Bank, World Development Indicators 2011.

78 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

Figure 3 also shows that, on average, the the Czech transition (Green, 1999). Popular outcry countries in the EU group started their economic over shock therapy’s supposed failures led to political recoveries after almost four years of transition and turmoil in several post-communist countries. For reached pre-transition levels in eight and a half years. example, Hungary’s conservative prime minister, It took the countries in the FSU group almost six Viktor Orban, was nearly ousted by opposition years to enter the second stage of transition. Some of socialists in 2002 because Hungary’s reforms had them completed the economic recovery after almost widened the gap between rich and poor (Szamado, thirteen years of transition, while few of them like 2002). Thus, an issue of corruption deserves more Georgia and Ukraine have not fully recovered yet. analysis. Thus, Figures 2 and 3 clearly show that the Czech The Transparency International that studies Republic’s transition is a success, while Russia’s is changes in the corruption perception index (CPI) in well below average. the post-communist countries defines a rampant level Another very popular benchmark for of corruption as a score of three or less than three on comparison of the transition countries is the level of a one to ten scale. Figure 4 demonstrates a correlation corruption. While most FSU countries remain highly between per capita income and corruption for every corrupt at both economic and political levels, the post-communist country in 2004. We can see that Czech government managed to create and secure most FSU countries had rampant corruption in 2004 favorable and relatively corruption-free conditions for but the level of corruption was one of the lowest in the development of private sector. Nonetheless, mass the Czech Republic. The countries which had a lower media sources like the International Herald Tribune level of corruption were Estonia, Lithuania, and still emphasized the corruption in Vaclav Klaus’s Hungary. Those countries, however, had lower GDP economic transition and blamed it for the failure of per capita than had the Czech Republic.

Figure 4. Corruption and GDP per Capita in Post-Communist Countries, 2004

Sources: Transparency International, Global Corruption Report, 2004; WDI, 2011. Data are unavailable for Serbia and Turkmenistan.

Furthermore, Table 1 summarizes average liberalization, commercialization, and economic and political performance of transition decentralization. The Heritage Foundation’s index of countries between 1991 and 2004.12 Here we compare economic freedom and the polity score measure a the Czech Republic’s quality of economic and quality of economic and political institutions, political institutions to these benchmarks in Russia respectively. and other post-communist countries. The EBRD’s As shown in Table 1, the Czech Republic had privatization and infrastructure reform scores much more successful transition than had the rest of measure advancement of the market reforms. The post-communist countries, including Russia. And former evaluates privatization and the latter assesses these differences are statistically significant. On average, the Czech Republic’s per capita income was twice as much as per capita income in post- 12 For a description of variables, please, see Table 2 in the Appendix. communist countries. The Czech economy’s average

79 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

rate of growth exceeded the post-communist average by almost one percent, while Russia’s economy contracted. Table 1. Transition in Post-Communist Countries, 1991-2004

Post-Communist The Czech Republic Russia Indicators Mean Mean Mean Infrastructure reform score 1.86 2.55 1.95 Index of economic freedom 2.59 3.69 2.32 Polity IV score 13.4 19.7 14.9 Privatization score 2.90 3.85 3.23 GDP per capita 6,232 14,896 7,360 Economic growth rate .35 1.23 -.62

Note: The means of variables are different at 1% significance level.

Table 1 also shows that the Czech Republic had countries. The countries that had a more advanced a significantly higher quality of political institution infrastructure reform were Hungary, Poland, and and rule of law as compared to the post-communist Estonia. average level. In terms of political regime, the Czech Figure 5 shows a correlation between economic Republic had already been a consolidating freedom and privatization in post-communist democracy, while, on average, post-communist countries. The Czech economy also experienced countries like Russia were anocracies (i.e. incoherent much higher level of economic freedom than the rest and unstable authority regime). of post-communist countries, while Russia’s level of Moreover, the Czech Republic had one of the economic freedom was below average. We can see most advanced transition economies. EBRD’s that the Czech Republic had the highest level of transitional indicators show that the Czech Republic’s privatization and the second highest level of privatization and infrastructure reform were economic freedom among all post-communist states. significantly higher than the post-communist average. The index of economic freedom shows that Czech The Czech Republic had one of the highest scores for Republic had the second highest quality of economic infrastructure reform among post-communist institutions among post-communist countries.

Figure 5. Privatization and Economic Freedom in Post-Communist States, 1991-2004.

4 CZECZECZE SVKSVKSVKSVKSVK HUNHUNHUN POLPOLPOL ESTESTEST HRVHRVHRV SVNSVNSVN LTULTULTU RUSRUSRUSKGZKGZKGZ LVALVALVALVALVALVA 3 KAZKAZKAZGEOGEOGEOALBALBALB ROMBGRBGRBGR MDA ARM ARMMDA UZBUZBUZB UKR 2 TJKTJKTJKTJK R =.6563

2 AZEAZEAZEAZEAZEAZE

Average Privatization Score Average Privatization BLRBLRBLR TKM 1

1 2 3 4 Average Index of Economic Freedom

Sources: Heritage Foundation, 2011; EBRD, 2011.

The last but not the least indicator of the countries. In the Czech Republic life expectancy transition economy is social conditions. We use both increased from 72 to 76 years between 1990 and life expectancy at birth and infant mortality to 2004, while it increased from 71 to 74 years in estimate level of social conditions in post-communist Slovakia in the same period (WDI, 2011). Moreover,

80 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

Hungary is the only post-communist country that had some post-communist countries which we have not a larger increase in the life expectancy than had the considered yet. The evidence, however, suggests Czech Republic. The fact that life expectancy otherwise. The main economic success of the Czech increased in the Czech Republic, while it declined in Republic that was the 2008 EBRD graduation came six post-Soviet countries such as Russia and Ukraine, soon after the EU membership. is important. Furthermore, infant mortality rates per In 2008 the Czech Republic became the first 1,000 live births dropped from ten to four in the post-communist country to exit from the EBRD’s list Czech Republic between 1990 and 2004 (WDI, of transition economies. The 2008 EBRD graduation 2011). In Slovakia infant mortality rates declined that marked eighteen years of the Czech transition from fourteen to six in the same period of time. from socialism reflected “the advanced state of Overall, we can see that the Czech Republic had the transition achieved by the Czech people” (EBRD, second highest life expectancy and the lowest rate of 2007). Moreover, the Czech Republic was and infant mortality among post-communist countries. remains the only country from the EU group of post- The improvements in living conditions are also communist countries to complete transition though impressive when we look at qualitative improvements the EBRD anticipated all eight countries to graduate in consumer goods. For example, passenger car by 2010. Jean Lemierre, the president of the EBRD, ownership grew in the Czech Republic significantly. described the Czech transition as “extraordinarily There were 228 and 163 passenger cars per 1000 successful journey that the Czech authorities and the people in the Czech Republic and Slovakia in 1990, Czech people have taken to build a thriving market respectively. In 2002 there were 356 and 247 economy anchored by the democratic institutions that passenger cars per 1000 people in the former and the bolster a sustainable, strong economy” (EBRD, latter, correspondingly. Likewise, quantity of 2007). television sets increased in the Czech Republic. If Figure 6 compares the Czech Republic’s per there were 284 and 340 television sets per 1000 capita income to this benchmark in the former people in the Czech Republic and Slovakia in 1991, republics of Yugoslavia and Russia between 1990 respectively, there were 538 and 409 television sets and 2010. Leeson and Trumbull (2006) and Roland per 1000 people in the former and the later in 2001, (2000) argue the transitions of the former Yugoslav correspondingly. republics is different in some ways from the Czech Republic or the other countries discussed above 4 The 2008 EBRD Graduation because the former Yugoslavia was a market socialist economy. Nevertheless, these countries have So far we have restricted our discussion to a struggled with many of the same issues, such as comparison between the Czech Republic and other privatization, as the other transitional economies. So post-communist countries, mainly the EU group, it is interesting to see how the Czech Republic between their start of transition and the 2004 EU performs relative to them. enlargement. Perhaps the Czech Republic’s transition has not been that impressive since 2004 or versus

Figure 6. GDP Per Capita in the Czech Republic, the Former Yugoslavia, and Russia, 1990-2010 (constant prices)

Sources: Calculated from World Bank, World Development Indicators 2011. Data for Bosnia, Herzegovina, and Montenegro are not available for a whole period.

81 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

We can draw several insights from Figure 6. Furthermore, we agree with Popov (2007) that The only economy that outperformed the Czech initial conditions can be favorable towards transition. economy is Slovenia’s. Russia again had the longest To control for differences in initial incomes and and most severe contraction, and nearly ties with transition start dates, we demonstrate the same Serbia and Macedonia for last place in terms of the information from Figure 6 in the form of a GDP extent of its economic recovery since transition index in Figure 7. The base year is still the year prior began. It is also interesting that Russia’s per capita to transition. As shown in Figure 7, the transition income was higher than Croatia’s in 1991, but the economies, except Serbia, surpassed pre-transition latter still outperformed the former. levels in eighteen years of transition.

Figure 7. Index of PPP Estimates of GDP Per Capita: the Czech Republic, the Former Yugoslavia, and Russia, First 18 Years of Transition (%)

Sources: Calculated from World Bank, World Development Indicators 2011. Data for Bosnia, Herzegovina, and Montenegro are not available for a whole period.

On average, the only economy that outperforms Figure 8 presents initial per capita income and the Czech Republic’s index (108%) is Slovenia’s its value after eighteen years of transition for all post- (119%). Both the Czech Republic and Slovenia have communist countries. As shown in Figure 8, the returned to their pre-transition levels of economic Czech Republic started transition with the second performance in six years. Another former Yugoslav highest per capita income among post-communist republic that is clearly enjoying a successful countries and completed transition with the second transition is Croatia. Nonetheless, the Czech highest per capita income among all post-communist Republic’s economy still outperforms Croatia’s countries. If the initial economic conditions matter for economy. Serbia and Macedonia are clearly future economic performance, it is important fact that experiencing very troubled transitions, no doubt in none of the post-communist countries with less large part because of the Balkan war conflict that favorable initial conditions outperformed the Czech plagued the region since its breakup. While Serbia is economy. By contrast, many post-communist so far a failed transition, Macedonia has finally countries with less favorable initial conditions returned its economy to pre-transition level. On outperformed others with the more favorable ones. A average, Macedonia’s index (87%) even case in point is Russia that lost its economic position outperformed Russia’s (84%) in eighteen years of in eighteen years of transition from the sixth to the transition. In fact, the only economy that performed ninth highest per capita income among the transition worse than Russia’s economy is Serbia (per capita economies. Poland, Lithuania, and Estonia income index = 63%). outperformed Russia while their initial per capita income was much lower than Russia’s.

82 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

Figure 8. GDP Per Capita in Post-Communist Countries After 18 Years of Transition

Sources: Calculated from World Bank, World Development Indicators 2011.

Moreover, after eighteen years of transition We defend the shock therapy approach by a several post-communist countries like Serbia, juxtaposition of the Czech Republic and other post- Georgia, and Ukraine had their per capita income communist countries. We argue that the Czech below the initial levels. Of course, economic transition is a consistent success because of the Havel performance of several post-communist countries was shock therapy. We use several facts to corroborate distorted by civil wars because civil conflicts put our argument. First, the Czech Republic was one of severe constraints on post-socialist economies the eight post-communist states which were a part of (Leeson and Trumbull, 2006). For example, the 2004 EU Enlargement. The EU membership Georgia’s economy was severely devastated by the clearly indicated an advanced level of economic, civil war. As it was mentioned earlier, the Balkan war political, and social development. Second, the Czech conflict also had a devastating economic effect on Republic was the first and only post-communist several countries of the former Yugoslavia. Thus, country to graduate from the EBRD’s class of initial economic conditions can only partially explain transition economies in 2008. The EBRD graduation differences in transition economies after eighteen demonstrates an advanced level of the Czech years of transformation. Weak rule of law, civil economy. Third, unlike the Yeltsin shock therapy, the conflict, and other political economy factors have a Havel shock therapy was a successful reform because significant effect on a path of economic development. it solved the political economy problems of reform’s If the Havel shock therapy failed to solve the political credibility and state’s commitment to reform. economy problems of credibility and commitment, Our findings show that the Czech Republic is a the Czech Republic would be on the less favorable successful transition from socialism. First, the Czech path of transition. Republic performed as well as the countries in the post-communist EU group that outperformed the rest 6 Conclusion of the post-communist countries. Second, the Czech Republic performed well above average even in The collapse of communism was celebrated comparison with the former Yugoslav countries that throughout the West. The era of tremendous had already had market socialism before their optimism, however, ended shortly. Since the mid- transition experiments began. Third, the Havel shock 1990s popular accounts of the post-communist therapy transformed a socialist country into a transitions have become very negative. The debate on working market economy with a stable democracy the speed of the market reforms (shock therapy versus and a strong rule of law, while many post-communist gradualism) has become more negative for the former countries are still in the gray zone of transition and more laudable for the latter (see Popov, 2007, economy and incoherent authority regime. Our 2000; Boettke, 2001; Kolodko, 2000; Stiglitz, 1999). findings are unapologetic to further criticism of either Many economists have criticized the shock therapy the Havel shock therapy or the Czech Republic’s efforts of the early reformers like Havel and Yeltsin. transition. Popov (2007) argues that the shock therapy had a We have attempted to set the record straight on negative impact on economic performance in the the Czech transition from socialism to capitalism. The beginning of transition, but the gradual approach had genuine shock therapies succeeded in countries like a positive effect on the rest of transition period. the Czech Republic and Poland. The shock therapies

83 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

which could not solve the political economy Returned after the War”, The Independent problems of credibility and commitment failed in Review, Spring 1997. post-communist countries like Russia. It is not really 15 Hoffman, D. (2004). The Oligarchs: Wealth and an issue of gradualism versus shock therapy, but an Power in the New Russia, New York: Public Affairs. issue of reform’s credibility and commitment to 16 Kaiser, R. (1991). Why Gorbachev Happened? New York: Simon and Shuster. reform, that determines a success of transition. The 17 Klaus, V. (1997). Renaissance. Washington, DC: Cato same speed of reform will have different effects if Institute. reform has different levels of credibility and 18 Kolodko, G. (2000). From Shock to Therapy: Political commitment. Economy of Postsocialist Transformation. Oxford, It is unfortunate that the Havel shock therapy NY: Oxford University Press. has received so much criticism prematurely. Other 19 Kraus, T. (2009). “The Issue of Restitution in the transition economies could have gained valuable Czech Republic,” The CEU Jewish Studies Yearbook, insights by looking at the Czech experience when http://web.ceu.hu/jewishstudies/yb03/ 09kraus.pdf. making decisions about their reforms. Instead, a 20 Leff, C. (1996). The Czech and Slovak Republics: Nation versus State. Boulder, CO: Westview Press. crucial misunderstanding of the shock therapy 21 Leeson, P. and Trumbull, W. (2006). “Comparing political economy discredits this approach to market Apples: Normalcy, Russia, and the Remaining Post- reforms. The debate on the speed of the market Socialist World”, Post-Soviet Affairs, 22(3), pp. 225- reforms has been a disservice to both the Havel shock 248. therapy and the millions of individuals looking for 22 Polity IV Project. (2010). Polity IV Dataset. College viable reform options. Park, MD: Center for International Development and Conflict Management, University of Maryland. References 23 Popov, V. (2007). “Shock Therapy versus Gradualism Reconsidered: Lesson from Transition Economics

after 15 Years of Reforms”, Comparative Economic 1 Boettke, P. (2001). Calculation and Coordination. Studies, 49, pp. 1-31. London: Routledge. 24 Popov, V.(2000). “Shock therapy versus gradualism: 2 Boettke, P. (1993). Why Perestroika Failed: The The end of the debate”, Comparative Economic Politics and Economics of Socialist Transformation. Studies 42(1), pp. 1–57. New York: Routledge. 25 Roland, G. (2000). Transition and Economics: 3 Dlouhy, V. (2001). “The Czech Republic: Ten Years Politics, Markets, and Firms. Cambridge: MIT Press. of Transition”, In Transition: The First Decade, ed. 26 Rothbard, M. (1995). Making Economic Sense. Blejer, Mario, and Marko Skreb. Cambridge, MA: Auburn, AL: Ludwig von Mises Institute. MIT Press. 27 Shleifer, A. and Treisman, D. (2005). “A Normal 4 EBRD. (2011). Transition Report, European Bank of Country”, Journal of Economic Perspectives, 19(1), Reconstruction and Development: London. pp. 151-174. 5 EBRD. (1999). Ten Years of Transition, European 28 Sima, J. and Stastny, D. (2000). “A Laissez Faire Bank of Reconstruction and Development: London. Fable of the Czech Republic,” Journal of Libertarian 6 EBRD. (2007). “The Czech Republic Graduates from Studies, 14(2), pp. 155-78. EBRD”, http://www.ebrd.com/pages/ 29 Stiglitz, J. (2002). Globalization and Its Discontents. news/press/2007/071023.shtml. New York: W.W. Norton. 7 Ellis, M. (1996). “Purging the Past: The Current State 30 Stiglitz, J. (1999). “Whither Reform? Ten Years of of Lustration Laws in the Former Communist Bloc”, Transition”, In Annual World Bank Conference on Law and Contemporary Problems, 59 (4), pp.181-196. Development Economics, ed. Pleskovic, Boris, and 8 EU. (2004). The 2004 enlargement: the challenge of a JosephStiglitz. Washington, DC: World Bank. 25-member EU, http://europa.eu/legislation_ 31 Szamado, E. (2002). “Poverty Gap Fuels Resentment summaries/enlargement/2004_and_2007_enlargement/ in ‘Booming’ Hungary”, Agence France-Presse, April e50017_en.htm. 17. 9 Goldman, M. (2003). Piratization of Russia: Russian 32 Transparency International. (2006). Global Corruption Reform Goes Awry. London: New York Report. London: Pluto Press. 10 Goldman, M. (1991). What Went Wrong with 33 Wheaton, B. and Zdenek, K. (1992). The Velvet Perestroika, New York: Norton. Revolution: Czechoslovakia, 1988-1991. Boulder, CO: 11 Green, P. (1999). “Prague Exchange’s Failed Reform Westview Press. Effort Leaves Some Predicting Its Demise”, 34 Weingast, B. (1995). “The Economic Role of Political International Herald Tribune, March 17, p.16. Institutions: Market-Preserving Federalism and 12 Hazlett, T. (1996). “Bottom-up Privatization: the Economic Development,” Journal of Law, Economics, Czech Experience”, In The Privatization Process, ed. and Organization 2(1), pp. 1-31. Anderson, Terry, and Peter Hill. London: Rowman 35 Williams, Kieran. (1999). “A scorecard for Czech and Littlefield. Lustration,” Central Europe Review, 1(19), 13 Heritage Foundation. (2011). Index of Economic http://www.ce-review.org/99/19/williams19.html. Freedom, Washington, DC: The 36 World Development Indicators. (2011). The World 14 Higgs, R. (1997). “Regime Uncertainty: Why the Development Indicators, Washington, DC: The World Great Depression Lasted So Long and Why Prosperity Bank.

84 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

Appendix

Table 2. Description of Transitional Indicators

Description Privatization score measures level of privatization, on a scale from 1 to 5, where: 1 means little progress and little private ownership and 5 means standards and performance typical of advanced industrial economies: more than 75% of enterprise assets in private ownership with effective corporate governance, no state ownership of small enterprises and effective tradability of land. Source: EBRD.

Infrastructure reform score measures average of five infrastructure reform indicators covering electric power, railways, roads, telecommunications, water and waste water on scale from 1 to 5 with higher score meaning higher degree of decentralization, privatization, liberalization and commercialization. Source: EBRD. Index of economic freedom measures level of economic freedom, on a scale from 1 (repressed) to 5 (free). Source: Heritage Foundation.

Polity score measures quality of political institutions and rule of law, on a scale from 0 to 20, with higher score meaning higher degree of democratization. The regime categories are the following: autocracy (0-4), anocracy (5-15), and democracy (16-20). Anocracy is a mixed or incoherent authority regime. Polity score for Bosnia and Herzegovina is mainly unavailable. Source: Polity IV project.

85 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

BANKS’ INCENTIVES TO OVER-HERD

Marcela Giraldo*

Abstract

Key Words: Banks, Herding, Banks’ Incentives

* One Bear Place #98003, Waco, TX 76798 Tel.: (254)710 – 3340. E-mail: [email protected]

1 Introduction and Literature Review bank, but also includes recessionary spillovers (changes in costs during a recession); principal-agent Economists and policy makers have spent much of problems, in particular the concern of managers to last couple years debating ways to avoid another protect their reputation; and other more general financial crisis. The Dodd-Frank act that passed in the papers, such as empirical papers that search for US congress is a framework under which new rules evidence of herding under different circumstances. and regulations must be written to re-shape the The papers that look at institutional incentives, financial sector. Most economist and politicians including the model set forth later in this paper, are accept that the rules must be changed, however the concerned mostly with the following problem: If one possible consequences of most of the proposals are small bank fails, the deposit insurance can guarantee subject of debate. Some even argue that in the attempt all the depositors receive their savings back and the to reduce risk, regulators may actually be increasing economy will not suffer from the bankruptcy. If many it. small banks fail at the same time, the consequences One of the most recognized problems that are different. The banking sector is crucial for the increase risk in the banking sector is the incentive to existence of any modern economy, so the failure of be big. A bank that is considered “too big to fail” will several banks can disrupt the normal functioning of most certainly be rescued and will bear a small cost the real economy. As a consequence, the government for excessive risk taking. Another less discussed is compelled to rescue the banks. In other words, if incentive problem is the fact that banks tend to many small banks fail together, it is equivalent to “herd,” or behave very similar to each other, so as to when one big bank fails. This is the main reason why be “too many to fail.” This article intends to explore banks have incentives to copy the actions of its the causes of this second problem, namely, bank competitors. If a bank is going to fail, it prefers to do herding. so with company so the chances of being rescued are There is a vast academic literature that studies higher. herding in the financial sector. Of that literature, only There are several papers that explore this a few papers study this problem as it concerns the explanation. Acharya and Yorulmazer (2007) banking sector. This overview focuses on those carefully evaluate the government’s time papers that study why banks herd, and that were inconsistency problem that leads it to bail out banks published in the last fifteen years 13. The literature that fail. The paper’s main argument is that when that evaluates herding in banks can be separated into many banks fail, it is ex-post optimal to bail out the four main categories: banks may herd as a response to banks, whereas when few banks fail, the banks that institutional incentives, such as expectations of survive acquire those that fail. This leads to herding government intervention or the structure of the behavior for the reasons explained above. This paper market itself; spillover fears, these refer mostly to the also shows that small banks have stronger incentives effects of the information revealed by the failure of a to herd than large banks. When a small bank imitates a large bank, the probability of being bailed out is 13 For a study of the literature before this period see higher when there is failure than if it differentiated Devenow and Welch (1996).

86 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

itself. However, big banks have little extra benefit In the principal-agent-problem category we find from copying a small bank’s behavior. papers concerned with the fact that managers want to This paper presents an extension to Acharya and protect their reputation. In other words, bank Yorulmazer (2007). It allows for many banks and managers are not only interested in the performance infinite time periods, as well as an optimal herding of their employer, but in their professional careers. In level. In other words, herding is not always trying to protect their image, bank managers are undesirable, but excess of such behavior is. The affected by group psychology: It is safer to err model in this paper is also simulated so as to explore together than individually. This occurs because in more detail the relevance of each of its parameters. judgment is less harsh when most other managers in Acharya and Yorulmazer (2008a) proposes a the market made the same mistake. A classic example solution to the time-inconsistency dilemma: When of this line of thought is Scharfstein and Stein (1990). too many banks fail, surviving banks may not have They call this behavior the “sharing the blame” the liquidity necessary to purchase the failed banks at effect14. a price high enough that avoids less efficient In the empirical literature managers and investors from entering the market (assuming there financial analysts have been proven to herd, even are such willing investors). Hence, some action is when doing so contradicts the private information required. Providing liquidity to surviving banks so known by some. One such example is Sias (2004). they can purchase the failed banks is equivalent, ex- Sias shows some evidence that implies institutional post, to a bail-out policy. However, it offers a reward investors herd as a result of deducing information to surviving banks, providing incentives to from each other’s trades. His paper measures differentiate rather than to herd. As a consequence, correlations between institutional demands across aggregate banking crises become less likely. adjacent quarters. He concludes that institutional An example of literature discussing the demand is more strongly related to lag institutional spillovers cause is Acharya and Yorulmazer (2008b). demand than to lag results. That paper presents the issue as an information A study more related to the banking sector can contagion problem. The incentives to herd are now be found in Uchida and Nakagawa (2007). This paper caused by the adverse information that emerges about evaluates data from 1975 to 2000 in search for one’s investments when another bank with similar evidence of herding by city banks in Japan. Its main exposures fails. Given the higher costs that result conclusion is that banks in Japan do herd, but for from this adverse information, a bank owner with most of the years studied the behavior could be limited liability prefers failing together with other explained by macroeconomic factors. Given the banks to doing so alone. Consider the following information at the time, all banks were acting as it example. Two banks have access to two different was optimal. The authors call this “rational herding.” projects. Each bank must choose one project to invest Only during the bubble period in Japan in the late in. Also assume that each project has a probability of 1980s, banks seemed to exercise “irrational herding.” failure p, and they are subject to the same Also empirical, but belonging to the first macroeconomic conditions. If the banks invest in the category, is Nicolo and Kwast (2002). This paper same project, when this one project succeeds both looks at how systemic risk can change according to the market structure. The paper shows that increases banks have high returns, . If it fails, both banks get zero. Now, if the banks invest in different projects, in consolidation contribute to increases in when one project fails and the other succeeds, the interdependencies among large and complex bank bank that invested in the good project will get lower organizations. returns, R, because the costs increased as a consequence of the failure of the other project. The 2 Model set up expected returns in the first case are (1-p) , whereas This model generalizes those in the previous in the second case they are (1-p)[ (1-p) +pR]. literature that belong to the first category; it looks at Hence, bank returns are highly correlated. the institutional incentives that may promote herding. The higher costs that result from a bank failure It allows for w banks and infinite periods, there are K do not only occur as a consequence of information possible projects looking for loans. For simplicity, all spillovers, but also because of “recessionary projects have the same probability function of spillovers,” such as a lower overall level of deposits. success, which depends on the amount of capital This would happen when not all deposits from the invested in it. Let be such function. It is failing bank find their way back to the banking continuous and system. This negative externality on surviving banks’ health is evaluated in Acharya (2009). Just as in Acharya and Yorulmazer (2008b), higher costs from other banks failing and limited liability provide the incentives for bank managers to prefer high bank 14 For a more detailed study on psychological factors that returns’ correlations. may explain behavior in the financial sector see Gärling et al (2009).

87 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

Each bank observes what other banks did in previous periods, then they simultaneously choose in which . project to invest in the current period. Next, projects either fail or pay back. In case of failure, the This means that there is an optimal level of government decides for each bank whether to bail it capital that maximizes the probability of success. out or not. Less capital than the optimal is not enough to take all the necessary precautions, and too much capital leads 3 Solution to the Benchmark Problem to overinvestment (such as in bubbles). It is through this function that one bank’s investment decision has If there was no government intervention when banks an externality on the payoffs of other banks. fail but the banks where instead liquidated, then the Although not common in the literature, the maximization problem for one bank would look as reason for this probability of success function is that follows: initially, ceteris paribus, more investment funds means there is more liquidity and access to capital. In Let if the bank invests in project j at period addition, more assets loaned send a signal to t, and 0 otherwise; represents the total units of customers of the project that many banks, or big capital invested in project j at period t. is the total institutions, believe the project is good (this may turn investment units in project j by all other banks at time into even more lines of credit in the future if needed). t. The value function for bank i at time t is Too much credit may lead to overinvestment and overheating of a sector in the economy caused by too big of a project and low returns. Too much debt may also lead to misuse of resources. In sum, the assumption states that levels of investment above or for representing the vector of decisions by i under the optimal level increase the probability of regarding each project16, is the discount factor, and default.

For ease in the analysis, although not a determinant of the results, f(x) is assumed to be . symmetric around . Initially let all banks have one is a vector of zeros if the bank has failed in the unit of funds available for investment. So, there are w past. If a project fails, all the banks that gave loans to units of investment available in the economy. Each the project also fail. bank must put its unit of investment in only one Let w stay constant over the long run, otherwise project (the units are not divisible). In addition, there the sector would disappear. In other words, each is deposit insurance, so that depositors are risk period new banks and projects may join the economy. neutral. Banks must pay a unique interest rate of r. This is not an assumption required for the results The relevance of this assumption and its effects on presented in this section. It is not relevant to specify the model will be explained below. when and how many banks fail or appear. An For simplicity, assume all projects have the assumption regarding this process will be made for same rate of return. Each project either pays back R the numerical example. This is however an important in case of success or zero in case of failure. This logical statement, otherwise this model would be assumption is common in related literature. describing a sector bound to disappear in the long Upon a bank’s failure, the government can run. either liquidate a bank (pays depositors and closes the Since the bank’s decision at period t affects the bank) or bail it out 15. In a bailout, the government future only in the probability that the bank will exist intervenes when the amount of failing financial assets tomorrow, during each period the banks’ decision is crosses certain threshold. There is some evidence that reduced to maximizing the probability of success, such a threshold does exist. Governments tend to seek . bailin solutions to bank failures, or allow liquidations, when only a small amount of (small) banks fail. Proposition 1: When w= *K (the total amount However, when enough banks fail so as to put the of investment funds available in the economy equals entire banking system at risk, governments all over the amount of investment that minimizes the risk of the world have consistently stepped in. The ongoing failure times the amount of projects available)[17] it is financial crisis is one testimony of this. Other a Nash Equilibrium, and a first best, to invest in empirical evidence includes Hoggarth, Reidhill and projects such that . Sinclair (2004).

15 The bailin option, when another bank or institution buys the failing bank, if performed as suggested by Acharya and 16 Yorulmazer (2008a), does not present herding incentives. In This vector has only one entry of 1 (the project chosen to fact, it is an alternative that would reduce such behavior as give the loan to), and the other entries are zeros. 17 explained in the literature review. The case of is evaluated in section 6.

88 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

Proof: already been reached. Hence, there will be over- i. maximizes the probability of success, and herding. hence maximizes the possible profits from an Proof: investment. If all banks invest in projects such that each ii. By i., once banks have reached for all project receives loans for a total amount of n* and j, any deviation (both withdrawing funds from a there is a bailout policy, each bank generates profits project and investing in a different project) increases equal to the probability of default in both projects in which the amount of funds was altered, and hence, decreases the payoff for all banks. (1) . iii. By ii., no bank wants to deviate (it is a Nash Equilibrium) and any deviation decreases the payoff Deviating and investing additional capital in a of all banks (it is also a first best). project that already has funding of n* gives profits equal to 4 Nash equilibrium with Bailout

Now assume that if a bank fails it will be bailed out with probability p. Starting on the period after the (2) , bailout, the bank only keeps a fraction s of the profits. Since in practice governments often let small banks (2) is greater than (1) when fail, but intervene when the amount of assets failed exceed a “comfortable” level that may vary with time, in this model the government bails out banks if the level of assets that are compromised in the (3) financial sector exceed a pre-established threshold. This threshold depends on many factors such as current and expected growth of the overall economy, . government deficit, connectedness of the banking system, political sentiment, etc. The left hand side of (3) is the loss in expected Let M be the threshold. If m bank-owned assets gains from overinvestment, while the right hand side fail in one period and m>M, then a failed bank is the increase in expected profits that comes from a (owning 1 unit of investment) will the bailed out with higher probability of being bailed out because of the deviation. probability . In other words, all banks are Note that the interest rate paid on deposits is equally likely to be bailed out, given that only m-M independent of the probability of the bank’s failure. assets will be rescued. This can only be true with deposit insurance. If such In this case, each bank’s objective function is: insurance were not available, the benefits of taking more risk would be lower, since the cost of attracting capital is higher. As a consequence, (3) is less likely to be true (there would be less incentives to herd). In other words, it is not only the bail out policy that increases the incentives of banks to herd, but any

policy that reduces the costs of overinvestment. Where

5 Numerical Example

In this section the functions in the model described above take specific forms in order to better and is the probability of being bailed out understand the dynamics implied, and to show how given failure18. easy it is to make over-herding optimal. Proposition 2: Under a bail-out policy, there are Assume there are 4 industries/projects and 12 conditions under which the maximum expected banks. The number of banks and projects is the same profits attainable for each bank occur when the bank in every period. If a bank fails in one period, it is invests in a project where a total investment of has either bought and recapitalized, or a new bank appears. In addition, an alternative to the failed project is found. This assumption is made to keep the 18 let be the vector that indicates the number of assets state at each period equal and hence, make the invested in each project. The probability of failing and being problem easier to solve. In addition, if new banks and bailed out is projects did not appear frequently, this model would

89 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

describe a disappearing economy. Each bank has 1 (including the bank in question), given that banks unit of capital to invest. Let [R-r]=1 and . invested in each of the four projects as vector indicates. Since the probability of being bailed out so that given that x units failed is , a) No bailout and

so, if all the banks and always play , .

b) Bailout. Assume it is expected that the government will For M=3, the value of not deviating from bailout some of the banks that fail. I evaluate whether when there is a bailout policy is: 3.1199 one-time deviations from the above equilibrium are ( ). If one firm deviates to profitable for any one bank. In order to do that I a project that is already funded by three other banks, assume s=0.5. (the only for one period, the present value of the deviation probability of being bailed out and failing in one strategy is 3.1578. See table for more simulations: period) is evaluated as follows: Define as the probability of being bailed out and that x assets (or banks in this case) fail

Table 1. Benefit of deviation with bailout

Non-deviation 1 Deviation 1

Prob fail & bail Value prob fail & bail Value

M=3 s=0.5 =0.9 0.0524 3.1199 0.0824 3.1578

M=4 s=0.5 =0.9 0.0373 3.0395 0.0428 3.0429

M=5 s=0.5 =0.9 0.0223 2.9632 0.0289 2.968

M=6 s=0.5 =0.9 0.0072 2.8906 0.015 2.8968

M=3 s=0 =0.9 0.0524 2.8571 0.0824 2.8532

M=3 s=0.25 =0.9 0.0524 2.9827 0.0824 2.9988

M=3 s=0.75 =0.9 0.0524 3.2702 0.0824 3.3321

M=3 s=1 =0.9 0.0524 3.4358 0.0824 3.524 non-deviation 1= (3 3 3 3), or three assets in each on the number of situations in which x (>M) project; deviation 1= (2 4 3 3), or bank one shifts to banks can fail. For example, in (3 3 3 3) 4 banks project 2 cannot fail since when a project fails, all the banks Effects of M: Recall that M is the maximum that invested in it must also fail. But in (2 4 3 3) it is level of assets (or banks) that the government is possible to see 4 banks fail. So, as M increases, the willing to allow to fail. Then, greater M implies lower probabilities of failure change at different rates for probabilities of bailout and hence, a lower value to both cases. any investment. Note that the benefits from deviating Effects of s: Recall that s is the fraction of the are not monotonic. The difference in values between bank that the owners keep after the bank has been non-deviation and deviation is the greatest when rescued. When s=0, the owners loose all the assets M=3. Then it decreases, but increases again between (but do not have negative profits due to limited M=4 and M=6. Non-monotonicity occurs in this liability). In this case the policy of bailout is example because the probability of bailout depends irrelevant for the bank owners, and there is no benefit from over-herding. However, increases in s increase the benefit from over-herding. In other words, the higher the fraction of the profits bank-owners get to

90 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

keep after a bailout, the greater the incentives to herd herd and avoid one of the projects. The table below today. This result is consistent with Acharya and shows the value of playing this strategy in every Yorulmazer (2008a). However, the authors in that period, and the value of the two relevant one-time paper warn that in practice very small values of s deviations. These numbers show that there is no provide incentives to the bank managers to invest in profitable deviation and hence over-herding this way very risky projects after the bailout. So that warning is a Nash equilibrium of this game. to rescue banks at a very high cost to the owners and Consider the following parameters: M=3, s=0.5 managers is not a credible threat. and =0.9. Now consider (0 4 4 4) as an equilibrium, and call it “bail equilibrium.” In this case all banks over-

Table 2. Nash Equilibrium with bailout

Bail equilibrium Deviation 2 Deviation 3

Prob. fail & bail Value Prob. fail & bail Value Prob. fail & bail Value

0.0786 3.2525 0.1332 3.0609 0.0547 2.9461

Bail equilibrium= (0 4 4 4), or putting four assets on 3 out of 4 projects in every period; Deviation 2 and 3 are the two possible one-time deviations from bail equilibrium: deviation 2= (0 3 5 4) and deviation 3= (1 3 4 4). This is considering that, for purposes of this paper (0 3 5 4) is equivalent to (0 3 4 5), and to any permutation of these four digits.

6 Excess Capital References

If there is not enough capital, it is straightforward to 1. Acharya, Viral V. (2009) “A Theory of Systemic Risk see that at least some projects will be under- and Design of Prudential Bank Regulation.” Journal of capitalized. But, what if there is more capital in the Financial Stability 5: 224-255. 2. Acharya, Viral V. and Tanju Yorulmazer (2007) “Too market than is optimally needed for the existing Many to Fail –An Analysis of Time-Inconsistency in projects/ industries? An extreme case of this situation Bank Closure Policies.” Journal of Financial is Greece and Ireland right after they joined the Intermediation 16: 1-31. European union. The adoption of the Euro and the 3. Acharya, Viral V. and Tanju Yorulmazer (2008a). implicit support by the stronger countries in the union “Cash-in-the-Market pricing and Optimal Resolution brought cheap capital to many of the smaller of Bank Failures.” The Review of Financial Studies countries that joined. 21(6): 2705-2742. Say bank comes into the market and in period 4. Acharya, Viral V. and Tanju Yorulmazer (2008b). “Information Contagion and Bank Herding.” Journal t, holds true. If the best alternative of Money, Credit and Banking 40(1): 215-231. investment has a return of z,, as long as 5. Devenow, Andrea and Ivo Welch (1996). “Rational , the bank will invest in one of the Herding in Financial Economics.” European Economic existing projects i.e. there will be overinvestment. Review 40(3-5): 603-615. Hence, extra capital leads to overinvestment and 6. Gärling, Tommy, Erich Kirchler, Alan Lewis and Fred increased risk, even as gets very small. Van Raaij (2009) “Psychology, Financial Decision Making, and Financial Crises.” Psychological Science Conclusion in the Public Interest 10(1): 1-47. 7. Hoggarth, Glenn, Jack Reidhill and Peter Sinclair This paper explores previous literature that presents (2004) “On the Resolution of Banking Crises: Theory theories on why banks tend to herd. Then, it proposes and Evidence.” Bank of England, Working Paper No. 229. a model that explains how government policies, 8. Nicolo, Gianni De and Myron L. Kwast (2002) although sometimes necessary, are important in “Systemic Risk and Financial Consolidation: Are They explaining this behavior. The model is innovative in Related?” Journal of Banking and Finance 26(5): 861- that it allows for several banks and time periods, as 880. well as in its inclusion of an optimal level of herding 9. Scharfstein, David S. and Jeremy C. Stein (1990) and a flexible government policy. The model shows “Herd Behavior and Investment” The American that it is not only the bail out policy that increases the Economic Review 80(3): 465-479. incentives of banks to herd, but any policy that 10. Sias, Richard W. (2004) “Institutional Herding.” The reduces the costs of overinvestment. Review of Financial Studies 17(1): 165-206.

The model is simulated numerically to make further explore the model and prove the existence of functions that generate herding behavior.

91 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

THE PERFORMANCE OF NEWLY PRIVATIZED FIRMS: THE CASE OF PORTUGAL

José Vaz Ferreira*

Abstract

The aim of this study is to investigate the pre and post privatization financial, social and operational performance of forty two Portuguese companies in most of sectors of economic activity that experience full or partial privatization through public share offering, direct sale or public contest, for the period from 1989 to 2009. That is, this work investigates, whether or not, the privatization of sate-owned enterprises (SOE’s) had caused improvements on the economic and financial health of those privatized companies, as it is suggested by the literature of property rights, public choice and agency theory. First, we document significant improvements on profitability, operating efficiency, capital investment, real output, dividend payout, treasury applications, activity levels and capital structure. Secondly, we experience significant decreases in employment after privatization. Third, we observe that, following privatization, the financial equilibrium (short and long) of firms was negatively affected. Lastly, our results are generally robust surviving the partition of the dataset into various sub-samples.

JEL classification: G3; G32, L33

Keywords: Initial Public Offerings; Privatization; Ownership Structure; Corporate Governance; Economic, Social, Financial And Dividend Performance Of New Privatized Firms

* University of Coimbra, School of Economics, Av. Dias da Silva, nº 165, 3004-512 Coimbra, Portugal Tel.: +351 239 790 500; +351 917 229 375 E-mail: [email protected]

1 Introduction firm and keeps the discussion of this problem. To explain that a privatized firm may produce more Privatization is the sale by a government of state efficiently than a nationalized one, the analysis owned enterprises (SOEs) to private investors. usually falls back on assuming that there are some Privatization, in essence, relates to the transfer of exogenously given differences in the abilities of the responsibility for the performance of a specific government and the private owner. Although there is service from the public to the private spheres. When a certainly a lot of casual empirical evidence to support firm is privatized, that is, when the ownership these assumptions, it would be more satisfactory to changes from the government to private hands, a lot explain the differences between the two of factors related with its functioning change, such as, organizational modes endogenously. organization, procedures, commercial and marketing The majority of the investigation on strategy, industrial technology, strategic planning, privatizations, so far, has had the following etc. Our aim is to investigate if these changes have omissions: in first place, the investigation on this consequences in the operational, social and financial area, has been oriented only to the operational and performance of the privatized firms. economic side of the firm’s performance, during the Privatization contributes to use of markets to post-privatization period; as a matter of fact, most of allocate resources, as defended by (Boycko et al., the authors with work on this area, did not consider 1996). Since the first privatizations in Germany in the on their investigation, the financial side of the firm’s early 1960s and the privatizations of the Britain's performance after privatization. Most of work done government in the early 1980s, privatization is now till now, as (Clamote, 1999) and (D’Souza and accepted as an important tool of economic policy Megginson, 1999), was oriented only to the used by governments all over the world. What kind of operational (economic) performance of privatized goods and services should be provided by companies, before and after the privatization year, government employees as opposed to private firms? and they analysed the following performance The discussion of this problem is still to economic areas: (1) Profitability, (2) Operating continue for a long period of time. The economic efficiency, (3) Capital Investment, (4) Real Output theory still finds it difficult to explain what makes the (5) Employment and (6) Leverage. difference between a privatized and a nationalized

92 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

Our work has the objective to investigate how versus firms that were not restructured before privatization in Portugal, since its beginning years, privatization. affect not only the operational (economic) Privatization typically transfers both control performance of the former SOE’s, but also their rights and cash flow rights to managers who then financial performance and, as a consequence, their show a greater interest in profitability than did the financial equilibrium and structure, intending to fill in politicians, as defended by (Jensen and Meckling, this knowledge gap. That is, the scope is not only 1976). Because of the importance of this area, the limited to economic aspects of firm after analysis conducted on this study seeks to determine privatization, but also, it is oriented to the financial whether the privatization of SOEs in Portugal is truly consequences on firms of the privatization process. In desirable and lives up to the expectations of order to achieve this objective, besides the economic governments and development agencies for the analysis, similar to the authors mentioned above, we performance of newly privatized firms. We feel that a added some other aspects to the economic analysis multi-industry sample provides a broad perspective of and we developed a financial analysis, before and share issue privatizations and offers significant after the privatization period, based on following opportunities to identify the sources of the economic financial indicators: (7) Dividend policy, (8) and financial performance in newly-privatized firms. Treasury, (9) Activity levels, (10) Short term Using univariate pre versus post privatization equilibrium and (11) Long term equilibrium. This is comparisons, we examine whether the privatization the first study developed in Portugal, covering in has or not has changed the financial and operational detail, not only the operational (economic) performance of Portuguese privatized firms so far. performance, but also the financial performance of On the economic side, we document significant those firms, what is an omission of previous studies. improvements on profitability, operating efficiency, In second place, most of the work done so far, capital investment, and real output and activity levels. considers the firms as a whole, in aggregate terms, Firms on the competitive sector of the economy, with some exceptions, such as, (Clamote, 1999) and firms with changes on more than 50% of the Board, (D’Souza and Megginson, 1999) that had divided the firms in the nonfinancial sector, firms with foreign full sample into five subsamples: (1) Non-competitive control, firms with concentrated structures after versus competitive firms, (2) Firms with more than privatization, firms with stocks listed in an Exchange 50% control versus firms with less than 50% control Market, firms with shareholders in management, by government after privatization (3) Firms with firms that changed the total control after more than 50% change on the Board of Directors privatizations and firms that restructured, have a more versus firms with less than 50% change on the Board significant operational improvement than the of Directors, (4) Firms in which a new CEO is correspondent opposite subsample. appointed after privatization versus those in which On the social side, we experience a significant the old CEO is retained and (5) Firms from decrease on employment after privatization, which industrialized countries versus from developed means that, after the privatization process, the countries. personal costs, among others, are some of the priority In fact, our work goes beyond the sub-analysis, fixed costs that are cut-off by management. The more based on sub-samples that were developed by significant decline of this type of costs is experienced previous studies filling in this knowledge gap. by the sub-samples mentioned on the last paragraph. Besides the sub-samples presented above (1), (2) and Lastly, on the financial side, we document a (3), our study investigates the performance behaviour decline on the short and long term equilibrium, which (operational and financial) of the following other sub- means that companies, after privatization, at least samples: (4) Financial Sector Firms versus Non- during the first years, have some financial instability, Financial Sector Firms. (5) Foreign Allocation of mainly due to a great effort to finance their growth Control (more than fifty percent) Versus National and capital investments, lay-offs, and to finance the Allocation of Control (more than fifty percent) (6) redirection of their commercial and marketing Concentrated ownership structure versus non- strategy. concentrated ownership structure after privatization. This article is organized as follows. Section 2.2 (7) Share Issue Privatizations (SIPs) versus Direct provides the theoretical and empirical research on Sale (DS) or Public Contest (PC) privatizations. (8) privatization. Section 2.3 describes the data and Firms that have their stock officially quoted after sample collection. Methodology, empirical proxies privatization versus firms that do not have their stock and testable predictions are described in Section 2.4. officially quoted after privatization (9) Privatizations Section 2.5 presents the empirical results. Section 2.6 in or before 1990 versus privatizations after 1990. sums and concludes. (10) Firms that have shareholders in management after privatization versus firms that do not have 2.2 Literature review shareholders in management after privatization (11) Firms that were restructured before privatization Privatization is an issue for political leaders trying to deliver services demanded by their citizenry, while

93 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

maintaining reasonable cost of operating government. 2.2.1.2 Empirical evidence of the efficiency of state It is also an issue for those who see a more restricted versus private ownership role for government and argue that the private sector should perform most tasks not specifically delegated The consequences of government ownership on firm to government, as (Parker, 1994). Others, such as, performance can be analysed using two alternative (Rees, 1988) are interested in privatization because of methodologies. Firstly, through a comparison a belief in the power of the marketplace and between the performances of the government owned competition to provide goods and services at a fair with the privately owned firms. The major difficulty price, responding to public demand. of this method is to choose the appropriate set of comparison firms and the best measures of 2.2.1 Research on the state-owned and privately- performance. Secondly, some government owned owned firm’s performance companies have political goals, instead of profitability and efficiency goals, what makes the 2.2.1.1 The theory of the efficiency of state versus comparison more difficult to be done. private ownership Prior studies, such as, (Eckel et al., 1997), have found that firms are more profitable and efficient There are some theories about privatization defending after privatization. Nevertheless, changes in both the a direct relationship between the impact of competitive environment and firm objectives usually privatization and the degree of market failure. occur simultaneously with a change in ownership. Welfare theory argues that the privatization process These studies of privatization thus measure the join has the most relevant impact for state owned effect of changes in ownership, market structure and companies integrated in competitive markets. firm objectives, but assume that the primary effect is (Sheshinski and Lopez-Calva, 1999), argue that there due to a change in ownership. should be "... important efficiency gains from changes In addition, (Vining and Boardman, 1992), to private ownership in competitive structures." The using a Canadian firm’s database had results much competition effects can be so significant that conduct similar to the results of other authors, thus is, private the state owned firm to react to pressures in order to firms perform much better than state owned improve efficiency without privatization. companies. Inefficiency in state owned firms can be explained because those firms do not have to compete 2.2.1.3 Are there policy alternatives to privatization? with private companies, as they receive funding from the government. In addition, in those companies there There are several authors that think that, more than is no discipline as we can observe in private the privatization itself, competition and deregulation, companies. are more relevant factors to improve performance of The government political and fiscal policies state owned firms, such as, (Yarrow, 1986), (Kay and have been determinant to the success or not of the Thompson, 1986), (Bishop and Kay, 1989), (Vickers efficiency of some privatizations. In addition, and Yarrow, 1991), and (Allen and Gale, 1994). governments have raised great amounts of money On the contrary, other authors are convinced through the sale of state owned enterprises. Those that privatization is absolutely necessary and the only sales contribute to reduce the fiscal deficit in many way to achieve better performance results, such as, countries, as had happened in Portugal. (Vinnig and Boardman, 1992), (Boycko et al., 1994), Privatization develops factor and product (Nellis, 1994), (Brada, 1996), and (Shleifer, 1998). markets, as well as capital markets, as defended by (Ehrlich et al., 1994). The consequences of 2.2.2 Financial performance of divested firms privatization on the industry sector are different from country to country, depending on the strength of the The research of the financial performance of existing private sector. Privatization can also develop privatized firms and of the state owned companies the growth of institutions that improve the operations raise other methodological questions: in first place, of markets. privatization has been a government tool to achieve In summary, according to the existent literature, certain political objectives. Additionally, there has there is no doubt, that state ownership has important been a problem with data availability and weaknesses. (Shleifer, 1998) sums up much of the consistency. Also, a question arises when the literature with, "... a good government that wants to researchers have to decide to choose accounting or further “social goals,” would rarely own producers to market data. In fact, in some privatized and private meet its objectives". companies may exist some accounting problems, according to the goals of the owner. Finally, there are questions related with the selection of a benchmark to compare performance and to choose the appropriate statistical tests.

94 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

For some authors, as (Boubakri and Cosset, Many authors already developed work about the 1998), the increase in profitability of privatized firms consequences of privatization on the performance of may the result of the report earnings “management” divested firms, such as, (Claessens and Djankov, by the government, for instance, decreasing reported 1999), (Lizal et al., and 2000), (Frydman et al., and earnings before the privatization to convince 1999) and many others. Some of the conclusions of employees about the benefits of privatization. these authors about the causes of post-privatization According to (Boycko et al., 1993) state-owned performance improvement are as follows: foreign and companies do not have profitability as a goal; instead, outside ownership; concentration on private those firms have different objectives, as the ownership after the IPO; restructuring of firms after maximization of employment and the development of the IPO, more frequent when outsiders take control; backward regions with a variety of indirect subsidies. when new CEOs are nominated to run the new privatized companies; post-privatization 2.2.2.1 Empirical studies employing data from non- improvements are greater for small companies, but transition economies that improvement decreases in long run, probably due to increased competition; employment decrease is The studies from non-transition economics confirm normally followed by a significant increase in labour other studies, that is, privatization is linked with productivity. improvements in the post-IPO period. (Martin and Parker, 1995) study is one exception, since he found a 2.2.3 A Summary of privatization research decrease in performance for six of eleven British firms after privatization. In these non-transition The privatization research conducts us to the economies, the general rule is a general performance following conclusions: improvement after privatization, not only in  A great part of firms that are privatized through economic and social terms but also in financial terms. an Initial Public Offering (IPO), believe they The contrary performance results in different will improve their capital structure and increase studies may be explained by different methodologies their profitability in the short term, according to used in those studies, the size of the sample, some (Febra, 2000); make-up done by he accountants and, probably, a lot  The role of the state owned enterprises in the of omitted factors. That is, there is not a “standard” economy was significantly diminished with the output. Depending upon the proxies involved, the privatization programs; outcome may be more or less relevant in relation to  Privatization is a complex process that is related the performance results after privatization with political and economic factors; therefore, (La Porta and López-de-Silanes, 1999) the technique used to privatize a state owned developed a single-country study. They executed a firm depends on that type of factors; global investigation of the majority of Mexican  The IPO underpricing is used by governments to privatizations and they compared performance favour domestic over foreign investors. By this changes to industry-matched private firms. (Dewenter way, state owned firms’ employees are and Malatesta, 1996) used different periods to favoured, since they keep preferential investigate performance of private firms and state- allocations, (Megginson and Netter, 2001); owned firms over an extended time period  The existence of “golden shares” gives governments veto power and other control 2.2.2.2 Empirical tests of privatization in transition restrictions; economies  Privatization proved to be the right mechanism to achieve results that would not be attained To investigate privatization on firm performance without it; is much more complicated in transition economies  After privatization, firms increase their than in non-transition economies. The main difficulty profitability, output, investment, dividend to test the privatization effects in transition payout, efficiency, decrease their leverage and economies is the fact that, as they move from become more efficient, according to (Clamote, Communism to a market economy, everything occurs 1999); at the same time, not only in economic but also in  In general, employment falls with the political terms. As a result, to isolate the privatization and there is always a large consequences of privatization, it is a very difficult compensating improvement; task. In addition, according to (Djankov and Murrel,

2000), there are enormous difficulties in terms of  Privatization, most of the times, brings new accounting procedures and the type of financial managers with improvements in performance; reporting. The work that is done for transition  Buying shares in a share issue privatization and economies has many problems with a relevant selling them in the short run, normally, lead to selection bias and omitted variables. abnormal returns;  Countries with large privatization programs normally have had rapid growth in the stock

95 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

market capitalization and trading volume, activity, including commercial banks, insurance according to (Almeida and Duque, 2000); companies, industrial and commercial companies.  Finally, the privatization phenomenon improved We required directly from the privatized firms: the capital market regulations, the information (1) the offering prospectus for their initial offer, disclosure rules, and contributed to create which invariably presents multiple years of modern financial systems. preprivatization financial data, as well as details As we saw, there are different theories about about the offering itself, and (2) the annual reports privatization and its effects in the post-privatization from the postprivatization periods. Approximately period. Most of them defend a direct relationship 90% of the companies approached fully or partially between the impact of privatization and the relevant complied with the requests. In several cases, we performance improvements of privatized firms. Some supplemented the financial statements sent with argue that the privatization process has the most secondary sources, namely, commercial banks, Bank relevant impact for state owned companies integrated of Portugal and Euronext Lisbon databases. We also in competitive markets. Others conclude that the most had personnel contacts with managers of some firms. relevant efficiency gains occur when firms change to In case of doubts about some aspects of the firms, we private ownership in competitive markets. For other also made direct phone calls. We did not include any theories, there is inefficiency in SEOs, because those company by relying exclusively on secondary firms do not have to compete with private companies, sources. since they get financing from government. Also, the We employ local currency data in all our inefficiency of CEOs may be explained by the lack of analyses and, whenever possible, we compute ratios discipline that we can observe in private companies using nominal data in both the numerator and and, therefore, state ownership has relevant denominator. In computing real sales and sales weaknesses. efficiency (revenue per employee), sales revenue data In order to confirm these theories, in this second was deflated by the appropriate consumer price index chapter, we define the best empirical methodologies (CPI). A similar procedure was employed to compute and the consequent work that is designed to conclude net income per employee. if our results are closer or not to those theories. With Our data includes privatizations of forty two a sample of forty two privatized companies, we firms. These transactions take place from 1987-2009. developed statistical tests (the Wilcoxon Signed Rank Therefore, our data span a larger time period than any Test and the Kruskal-Wallis Test). The first of them other privatization study made for the Portuguese is used to measure the post-operational, social and Market. Table 1 provides the sample of privatized financial performance in comparison with the same firms with the name of the company, the type of type of performance along the time before the IPO. industry, the issue date and the percentage of capital The second test is used to compare the several that was privatized at the date of the issue. The performance indicators for each pair of sub-samples sample is well diversified, exhibiting a wide temporal of the 11 sub-sampling criteria. A significant dispersion. difference between the sub-samples would indicate that the subgroup classification factor might be an 2.4 Methodology, empirical proxies and important determinant of post-privatization testable predictions performance changes. In this chapter, we use two different methodology 2.3 Data and sample collection techniques, which have different objectives: the Wilcoxon Signed Rank Test and the Kruskal-Wallis The factors behind the selection and definition the test. sample are as follows: Firstly, we limit our analysis to The Wilcoxon Signed Rank Test uses the Portuguese companies that were fully or partially magnitude and the direction of the differences to privatized through a public share offering and conclude, or not, if there are true difference pairs of through a direct sale or public contest, primarily data designed from one sample or two related because companies that are privatized by this way samples. This test should be chosen, since we want to continue to generate post-issue financial and incorporate the size and the direction of the accounting data that is directly comparable to pre- differences. With this methodology, we test if there divestiture data. Secondly, we select firms that have are significant differences in performance (analysed their initial public offering of shares, direct sales or by 12 indicators), before and after the privatization. public contest with financial information from 1987 There is a significant difference when we reject the to 2009 and have, at least, three annual observations null hypothesis (with H0 there are not significant in the years N-3 to N-1 and in the period N+1 to differences). The rejection of H0 is stronger than the 2009, where the year of privatization is defined as no rejection. year N. Finally, we define a very well diversified That is, the null hypothesis means that there are sample, including companies from all sectors of not differences among the mean values of each pair, that the pairs are identical and that the difference

96 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

magnitude is insignificant. The rejection of H0 is the the Direct Sale, depending on the privatization statically evidence that the performance of sub- regime. samples show significant differences between before Having computed pre-and post-privatization and after the IPO. In short, the objective of the means and medians, we use the Wilcoxon signed- Wilcoxon Signed Rank Test is to measure the post- rank test, as one of the methods of testing for operational, social and financial performance in significant changes in the selected variables. This comparison with the same type of performance along procedure tests whether means and median the time before the IPO. differences between pre and post privatization. In The Kruskal-Wallis test works with independent order to compare the performance improvements in sub-samples and it tests if they may be considered the privatized firms by comparing the financial similar. The variables are not correlated. The null indicators, we assumed paired data for the analysis, hypothesis (H0) is that two independent samples are which means that values in the two groups being similar. Using this test, we assume that the two compared are naturally linked or paired and usually subsamples are independent; therefore, we test H0 arise from individuals being measured more than with the perspective of rejection of it. If we do not once. For each company, we join the data before with reject the null hypothesis, we have statistical evidence data after privatization. We base our conclusions on to not consider that sub-sample pair. That is, the the standardized test statistic Z, which, for samples of rejection of H0 means there is statistical evidence of at least 10, follows approximately a standard normal differences between the sub-samples, which means distribution. Additionally, we assume that the that such sub-samples show different performance distribution of the differences between pairs of behaviour. observations is symmetric. This methodology was In short, the utilization of both methodologies also used by (Boubakri and Cosset, 1998), using the software SPSS, is justified by the need to (Megginson et al., 1994) or (D’Souza and test, simultaneously, the existence or not, of Megginson, 1999). statistically significant differences between variables defined in a different way (correlated in Wilcoxon 2.4.2 The Kruskal-Wallis Test for testing the Signed Rank Test and independent in Kruskal-Wallis significant differences between the sub-samples test). That is, as we want to test two types of different variables, we use the more appropriated tests to each In the second stage of empirical testing, we test for kind of analysis. significant differences between each dichotomous sub-sample pair using Kruskal-Wallis (KW) tests; 2.4.1 The Wilcoxon Signed Rank Test for this procedure, is the non-parametric equivalent of the measuring post-operational and financial one-way ANOVA. All observations are ranked performance regardless of the treatment group; ranks are handled as in the rank sum test. The mean rank sum is Governments expect that privatization will be the calculated for each group and the overall mean rank solution for most of the problems state owned sum is calculated. enterprises face today, such as, low operating The KW test was used to compare the several efficiency, low profitability, low output, high levels performance indicators for each pair of sub-samples of employment, weak capital structure, etc. They of the 11 sub-sampling criteria. A significant expect that privatized firms will become financial difference between the sub-samples would indicate healthier. that the subgroup classification factor might be an We use a complete set of indicators to reflect the important determinant of post-privatization operational and financial health of the privatized performance changes. With the null hypothesis that companies before and after privatization. On one the pair of sub-samples from possibly different hand, we use the indicators as (Megginson et al., populations actually originates similar results, it was 1994). On the other hand, we added other interpreted that the rejection of the null hypothesis ratios/indicators in order to have a more complete implies that statistic evidence exists for difference analysis of firms on the short term equilibrium, long between the sub-samples, therefore the assumption of term equilibrium, treasury, capital structure, etc. the sub-sample criterion; this methodology, the non- We compute empirical proxies for each parametric Kruskal-Wallis test, was used by company for the three years before privatization and (D’Souza et al., 2001). for all available years after privatization to 2009. We then compute median and means for each variable for 2.4.3 Sub-Samples, Empirical Proxies and Testable the pre-privatization [years N–3 to N–1] and post Predictions privatization [years +1 to 2009] periods. The year of the privatization, year N, is excluded from the The primary objective of this study is to investigate analysis because it includes both the public and how privatization in Portugal affects the financial and private ownership phases of the enterprise. The date operational performance of the former SOE’s. of privatization is the date of the IPO or the date of Governments expect that privatized firms will

97 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

improve the financial and operational performance of achieve that objective, we present the empirical firms, and most of the times, virtually all proxies for each determinant predicted to affect post- governments launching privatization programs have privatization performance. We will describe how we specific and generally very optimistic, expectations expect each variable to impact the newly-privatized about what these programs yield. firm’s financial and operating performance. In the first stage of the empirical test, we study the past privatization period observing the following 2.4.3.1 Competitive versus non-competitive analysis performance areas: first, those areas and its proxies, as studied by (Boubakri and Cosset, 1998) and by One believes that the financial performance of firms (Megginson et al., 1994). These include profitability in competitive markets is well different than those (return on sales), operating efficiency (sales firms that are not included in non-competitive efficiency), capital investment (real capital markets. The reason for the splitting the sample into expenditure to sales), output (real sales), employment competitive versus non-competitive industries is then (total employment), and dividend policy (dividend to straightforward. According to (D’ Souza and sales) and leverage (capital structure) (total debt to Megginson, 1999), competitive firms are defined as total assets). “those that are subject to international product market In addition to these performance areas, we use, competition, and non-competitive firms as those that in a second set of other areas and proxies as predicted are relatively free of product market competition”. relationship with great importance to understand the For the sale of enterprises in non-competitive post privatization performance of firms, as follows: sectors, the steps are more numerous and the process treasury (treasury applications), activity levels (sales is more complex. Successful privatization of natural to total assets), short term financial equilibrium (cash monopolies requires a regulatory framework that and banks to short term debt) and long term financial clarifies service goals, and develops cost equilibrium (net cash flow to long term debt). These minimization targets. For example, firms from the variables/indicators are very important because, on telecommunications and utilities industries are one hand, they give us a more complete economic included in the non-competitive sample. performance analysis than done before by other Several studies, such as, (Clamote, 1999) and authors, such as, (Boubakri and Cosset, 1998) and by (D’ Souza and Megginson, 1999), concluded that (Megginson et al., 1994) and, on the other hand, they privatization of enterprises in competitive industries, developed a financial performance analysis before such as, airlines, retail operations, or manufacturing, and after privatization never done before by any work yielded more robust and rapid performance of this area. improvements, than in non-competitive industries, as Having computed pre-and post privatization long as there are no economy distortions that means and medians, the Wilcoxon signed-rank test constrain competition. was then used in order to test for significant changes Industries in a competitive environment must in those variables. Therefore, we did test the restructure, reduce costs and, additionally, they are following hypotheses that privatization: (1) increases obliged to manage their resources in a very a firm’s profitability, (2) increases its operating professional way. When a firm is in a competitive efficiency, (3) increases its capital investment environment it is expected, that more pressure is put spending, (4) increases its output, (5) decreases on the firm, and this pressure is responsible for employment, (6) increases its payout ratio, (7) greater efficiency and profitability. This competition increases its treasury applications, (8) improves its force drives firms to higher levels of profitability and activity levels, (9) improves its short run financial efficiency. Thus, competition is expected to be a equilibrium, (10) improves its long run financial determinant of post-privatization performance equilibrium, (11) improves its capital structure. In improvements; therefore, firms in a competitive addition to analyse the full sample of privatized environment are expected to have greater financial companies, we also cut our full sample into several performance improvements than those firms in a non- dichotomous subsamples that are presented in competitive environment. Table 2. Thus, based on existing theory and findings, on Table 3 presents a summary of testable one hand, we expect that both types of firms predictions, including: in first place, the performance (competitive and non competitive) experience areas. We examine for changes resulting from improvements with the privatization process. privatization. In second place, we include the However, on the other hand, we expect that financial indicators used for each performance area. operational and financial gains are greater for firms in In order to establish the predicted relationship, we competitive markets during the postprivatization choose the best variable and its predicted changes for period. each area of performance. We divide the sample into firms included in In a second stage of empirical testing (Kruskal- competitive markets and firms within the non- Wallis), we are interested on why the performance competitive markets to determine the effect of improves, using sub-sampling criterions. In order to competition on post-privatization performance. Of the

98 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

forty two companies for which we have data, twenty consequence to answer some questions, as follows: six firms (62 percent) are operating in competitive which are the favoured approaches to financial industries. Firms included in the competitive sample privatizations? What evidence is there on the pre- and firms included in the non-competitive sample, are versus post-privatization operating and financial presented in Table II. performance of privatized banks and as compared to non-financial companies? 2.4.3.2 Change in the composition of the new Board During the past twenty years, a large number of of Directors after privatization financial institutions in Portugal have been privatized by the Portuguese government through public Before or immediately after privatization, turnover offerings of shares. This represented a fundamental among members of the Board of Directors is very break with the national past that emphasized the frequent, most of times, due to political reasons; strategic role of financial firms institutions and, in therefore, there is no stability inside the Board and we particular, commercial banks, in funding the nation’s have seen along the privatization period, various economic development and the national Directors going in and going out of the Board. When government’s key role in planning and directing that a firm goes public, both the ownership structure and development. the management structure of the firm, change. In contrast with the non-financial sector, where According to (Crutchley et al., 2002) significant there is now a well-established body of empirical changes in the Board of Directors after the IPO may literature on the effects of privatization (D’ Souza et be necessary to ensure that the wealth of the new al., 2000), the evidence on financial privatizations is shareholders is maximized. only beginning to emerge. (Verbrugge et al., 1999) A large turnover (fifty percent or more) in a empirically examined how the financial performance privatized firm's Board, represents two things: a) a of banks changes after being privatized via public powerful signal of a desire coming from the new share offering. They found some improvement in owners to change firm direction, with new bank operational and financial performance following management and ideas; b) a willingness to remove privatization. potential human resources constraints on the Thus, we expect that improvement changes in development process, with positive consequences on the financial performance of privatized financial firms performance. institutions will be much less pronounced than in the (Clamote, 1999), (Boubakri and Cosset, 1998), case of non-financial institutions. Our sample (Macquieira and Zurita, 1996) and (D’ Souza and includes 16 financial institutions (38%) and 26 non- Megginson, 1999), concluded on their investigation, financial institutions (see Table 2). that companies with greater than fifty percent changes in Board of Directors got better economic 2.4.3.4 Foreign Versus National Allocation of and financial results after privatization, than Control Analysis companies with less than fifty percent changes. We will also test these results. In addition to changes in post privatization We divide the sample into firms with less than performance between financial versus non-financial fifty percent turnover in Board of Directors and firms firms; it is relevant to investigate the influence of with greater than fifty percent changes. We expect the foreign investment and management know-how on high-board-change sub-sample to yield greater post privatization performance, as compared to the performance improvements than the sub-sample with post privatization performance of national less than fifty percent board change; as a matter of investments. fact, based on the existing findings, we expect that Any government that intends to privatize SOEs changes in Board of Directors will positively impact using public share offerings faces a very crucial the degree of post-privatization performance decision: how to allocate shares. The share allocation improvement. decision is very relevant and requires the government Twenty eight firms (66%) out of the forty two to choose whether to benefit one group of potential companies from the sample changed the majority of investors over another (i.e., domestic investors, SOE their board of directors (more than 50%) after employees, or both, over foreign and institutional privatization and the new board stayed on functions investors). (Anderson et al., 1997), in their study, for at least three years on the job. Firms included in identify 41 firms with direct foreign investment and each sample are presented in Table 2. 947 firms with no foreign investment. They found that profitability as measured either by return on 2.4.3.3 Financial versus Non-Financial firms equity or revenue per employee is significantly higher for the firms with foreign investors. It is clear that we know very little about financial (Smith et al., 1987), document a significantly privatizations as compared with non-financial positive relationship between profitability and foreign privatizations. Having investigated the non-financial ownership and a significantly negative relationship sector, the study of the financial sector is a logic between leverage and foreign ownership. As a result,

99 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

this would mean that foreign investment would bring other words, we expect that improvement changes in new management culture, new technology, new the financial performance will be much more products and marketing orientations procedures and pronounced in concentrated structures than in the new financial support with impact on financial ratios case of non-concentrated structures. and economic and financial performance. Twenty two firms (52%) out of the 42 We have reasons to believe that foreign companies from the sample have a concentrated investment would have a different behaviour in structure after privatization and 20 firms (48%) have Portuguese economies. Then we expect that the a non-concentrated structure after privatization; firms greatest performance improvements will result from included in each sample are presented in Table 2. privatization in which foreign private owners gain control of the firm. We expect that foreign allocation 2.4.3.6 Share Issue Privatizations (SIPs) versus of control will lead to improvement changes in the Direct Sale (DS) financial performance of privatized firms much more pronounced than in the case of national allocation of The share issue privatization and the direct sale are control; accordingly, we expect that the greatest the most known methods of privatization. According performance improvements will result from to (Megginson and Netter, 2001), sales of shares privatization in which foreign private owners gain through public capital markets (SIPs), are more likely control of the firm. Our sample includes 39 firms in less developed capital markets and for larger and (92%) with national allocation of control and 3 firms more profitable state owned enterprises. The with foreign allocation of control. countries decision to use IPO’s more frequently result from the governments’ need and desire to use IPO to 2.4.3.5 Concentrated versus non-concentrated develop the national market’s liquidity. IPO’s are structure after privatization more likely when income is more equal through the country, providing more potential investors and In addition to changes between foreign and national avoiding the need for extensive underpricing of the allocation, changes between firms whose capital is offerings (Megginson and Netter, 2001). concentrated in a few shareholders (it may be a Direct asset sales (sales to a small group of family) after privatization and firms with capital investors using private capital markets) and public whose dispersion after privatization is very high, is contests (there are various investors with specific also investigated. We are interested in exploring the proposals), are more likely to occur where relationship between corporate performance and the governments respect property rights and are not concentration degree of ownership. A concentrated expected to expropriate the privatized assets structure will happen when the privatized capital is (Megginson and Netter, 2001). concentrated in a few shareholders and when, at least, Simultaneously, (Pinkerton, 1982) concluded, one owner, has more than 50% of the privatized on his study, that, when raising new capital, managers capital. have historically rejected the direct sales method The reason for the splitting the sample into favouring, instead, the seemingly more expensive concentrated versus non-concentrated property underwritten public issue. They demonstrated structures is explained by the basic principle of the empirically that firms which engage in direct offers property rights theory. According to this theory, the enjoy a comparative economic and financial greatest incentive for maximization profits and performance improvement, which is more than efficiency exists on companies with concentrated sufficient to account for some issuing reported cost structures, where the owners are few and well known. differences between the two methods of equity The theoretical principle is understandable, since financing. Consequently, we expect that improvement agency problems identified in companies with changes in the economic and financial performance disperse ownership are not common within will be much more pronounced in the case of SIPs companies with one owner or with concentrated than in the case of direct sales or public contest. structures. As a result, this would mean that Our sample includes forty two companies, concentrated structures would bring higher where 23 firms (55 %) were privatized by IPO and 19 management incentives, a direct presence and firms were privatized by direct sale. Firms included influence of the owner on the main strategic in each sample are presented in Table 2. decisions, marketing orientations for the future of the companies with positive results on the financial and 2.4.3.7 Firms that have their stock listed versus operational performance of the firm. firms that do not have their stock listed on a stock Therefore, we expect that a shareholder exchange concentrated structure will produce positive results on the performance behaviour of Portuguese companies. We test the effect of firms that are listed after Hence, privatizations that generate the largest privatization and firms that are not listed on the post- concentrated amount of private ownership will privatization period. Starting a stock exchange listing generate the greatest performance improvements. In enhances capital and management monitoring that

100 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

will likely trigger post-privatization performance subsamples: firms that were privatized in, or before improvements. Thus, we expect that improvement 1990, and firms that were privatized after 1990. changes in the financial and operational performance Consequently, we expect that improvement will be much more pronounced when firms were changes in the financial performance will be much listed than in the case of firms that not listed. less pronounced for firms privatized in, or before, (Dewenter and Malatesta, 1997) and (Anderson 1990, than firms privatized after 1990. Since, after et al., 1997) argue that state-owned firms are less 1990, there were no special fiscal incentives, the efficient, because they are immune to capital market financial performance of privatized firms would be scrutiny. As a result, managerial performance is more homogeneous, without relevant changes (ceteris inadequately monitored. The public trading of shares paribus). establish the possibility of takeover by outsiders Our sample includes 42 companies, where 12 introduces the discipline of the managerial labour firms (29 %) were privatized in, or before, 1990 and market and provides the ability to link compensation 30 firms (71 %) were privatized after 1990. Firms to performance; as a result, when shares trade in the included in each sample are presented in Table 2. public equity markets, owners and managers have enhanced capacity to spur greater managerial effort 2.4.3.9 Firms that have shareholders in and accountability. The mere fact that contests management versus firms that do not have involving privatized companies have occurred, shareholders in management suggests that the introduction of capital market monitoring may trigger post-privatization The rationale for splitting up the sample into firms performance improvements. Therefore, we seek to that face the agency problem and firms with further establish a linkage between the fact that the shareholders in management is straightforward. The firm had or did not have its stock listed and nature of decisions that maximize the wealth of the performance of firms following privatization. firm’s shareholders should be different in both Our sample includes 42 companies, where 25 situations with consequences in post-privatization firms (60 %) are listed after privatization and 17 performance. An initial public offering of common firms (40 %) not listed after privatization. Firms stock typically leads to significant changes in the included in each sample are presented in Table 2. ownership of a company’s stock and reflects the dilution of an owner/manager’s stake as depicted in 2.4.3.8 Privatization in or before 1990 versus the analysis of agency costs by (Jensen and Meckling, privatization after 1990 1976). Their argument implies that a company’s performance suffers after going public. According to (Mello, 1996), the existence of fiscal Our research is motivated by the evidence that benefits is a condition to the definition of the management ownership appears to play a significant companies capital structure and the use of capital role in the performance of many companies. (Jensen markets as a permanent financial method; on his and Meckling, 1976), argued that agents or managers study about the Portuguese capital market, between have incentives to serve their own interests, while 1986 and 1990, Mello [1996], concludes that the may be to the detriment of the principal. After the fiscal benefits effect is well observable on the high IPO, as the management ownership rises, managers number of companies that were officially quoted in pay a larger share of the cost of the deviation from 1986, 1987 and 1998. the value-maximization and are less to squander In addition, (Mello, 1996), concludes that, in corporate wealth. This implies that corporate relative terms, the percentage of firms financing performance is expected to increase with the level of through the Portuguese capital market before 1990, insider ownership (Jensen and Meckling, 1976). was substantially higher than the percentage of firms On the contrary, the offsetting costs of that did the same after 1990. However, Mello management ownership have been raised up by clarified that the fiscal incentives might not be the (Fama and Jensen, 1983). According to them, when a only cause of the capital market development during manager owns a small stake, market discipline, those years. He adds other reasons for such evolution, product market and the market for corporate control as the development of the Portuguese economy and may force him toward value maximization. On the its capital market dynamic. other hand, if a manager controls a substantial Additionally, (Mello, 1996), concludes, on his fraction of the firm’s equity, he may have enough investigation, that only 30 percent of total firms listed voting power or influence to guarantee his from 1996 to 1990, stayed listed after the admission; employment in the firm and make policies without the other 70 percent, either quit from the market or market control. That is, corporate assets can be less changed from the official to the non-official market. valuable when managed by an individual free from Our study extends the Mello’s study and includes checks on his control. firm privatizations from 1987 to 2002. For the Our sample includes forty two companies, reasons explained above, we separate data into two where 26 firms (62 %) have shareholders in management after privatization and 16 firms (38 %)

101 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

do not have shareholders in management after SOE prior to sale. As a matter of fact, a related privatization. Firms included in each sample are practical question about privatization is whether presented in Table 2. governments should restructure SOEs (e.g., lay off redundant workers) prior to selling or leave this to the 2.4.3.10 Total versus Partial Privatization Analysis new owners. This is related to the question of whether reform and privatization should proceed quickly or Privatization is a process, not an event. In very few slowly. Early advice from the World Bank (Nellis cases is state ownership is eliminated with a single and Kikefi, 1989), was that governments should share offering. In many cases, there are one or more restructure SOEs prior to divestment, since seasoned offerings that follow an IPO and these governments are better able than private owners to transactions often occur a year or more after the cushion the financial blow to any displaced workers initial offer. The numerous cases in which by using unemployment or pension payments. privatization occurs in stages over time suggests that Two empirical papers that examined SOE revenue maximization is one of the forces driving reform prior to privatization are (Lopez-de-Silanes, privatizations, along the government’s wish to retain 1997) and (Malatesta, 2000). The first author its influence over the company. If the objective was examined whether prior government restructuring of to eliminate state ownership, one transaction could SOEs improved the net price received for the accomplish that goal. company, and finds evidence that it does not. He The models illustrate the importance of the shows that prices received by the government would sequencing and staging to build reputational capital have increased by 71 cents per dollar of assets if the with investors by the governments, building domestic only restructuring step taken by the government had support for the privatization program, as well as been to fire the CEO and if the assets had been identifying bidders that maximize the efficiency of divested, on average, one year earlier. He argues that the firm in the future. Some articles that empirically other restructuring steps slow down the process and examine sequencing or staging are the follows: (Jones consume too many resources to be worthwhile. et al., 1999) and (Megginson et al., 2000). (Malatesta, 2000) finds some evidence that the Several papers empirically examine the choices improvements brought about by privatization occur governments actually make in designing SIP before the SOE is privatized. programs. (Menyah and Paudyal, 1996) investigate Some industries, just prior to privatization, restructure the way in which the objectives of privatization through organizational changes and/or acquisitions influence the procedures and incentives used in the and divestures and/or financial restructurings (i.e., sale of state-owned shares, on the London Stock debt write-offs). Admitting that firms restructure in Exchange by the U.K. government. In control order to improve profitability and efficiency, we privatizations, the government sells voting control (it predict that restructuring should increase performance sells enough shares to bring its holdings below 50 improvements. We expect that changes in the percent); in revenue privatizations, the government financial performance will be much more pronounced retains a majority stake. Additionally, outside when firms were restructured prior to sale than in the investors (unlike managers), need not be cash case of firms that were not restructured prior to sale, constrained, and hence can, in aggregate, afford a since a well restructured firm is better prepared to larger ownership stake. face the marketplace and, thus, to improve more its According to (D’ Souza and Megginson, 1999), operational and financial performance than firms that this logic suggests that selling voting control to did not restructured before the sale. outside investors is most conducive to efficiency Our sample includes forty two companies, improvements; thus, we expect that improvement where 16 firms (38 %) had restructured before changes in the financial performance, will be much privatization and 26 firms had not restructured before more pronounced when firms were totally privatized privatization. Firms included in each sample are than in the case of firms that were partially privatized, presented in Table 2. since in control privatizations, the new owners have conditions to make structural management decisions 2.5 Empirical results in order to improve performance. Our sample includes forty two companies, In this section, we present and discuss our empirical where 21 firms (50 %) were privatized partially and findings for the full sample of all privatized firms and 21 firms (50 %) were privatized totally. Firms for the eleven subsamples. Our empirical work included in each sample are presented in Table 2. measures the post-privatization operational and financial performance. In global terms, our 2.4.3.11 Restructuring SOEs Prior to Sale versus investigation confirms, on one hand, that, following not restructuring analysis privatization, as it was expected by us, firms experience improvements (increase), significantly, in One of the more complex issues in this area involves average (median) levels of profitability, operating the interrelated questions of whether to restructure a efficiency, capital investment, output, treasury

102 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

applications, dividend policy, activity levels and (92 %) during the preprivatization period (-3 to -1 decrease in employment levels, when compared to the years), to 175 % (178 %) afterward. The Wilcoxon average (median) values from the pre-privatization tests show that SALEFF increases significantly (at period. the 5 % level) after privatization. In addition, 73 % On the other hand, our findings, in opposite to it (71 %) of all companies experienced an improvement was expected, show that firms do not experience on efficiency after privatization. This highly improvements in average (median) levels of short significant increase shows that in the post- term equilibrium and long term equilibrium when privatization firms use their resources on a much compared to the average (median) values from the more efficiency way. So, all the figures for the all pre-privatization period. sample show us very relevant operating efficiency improvements. 2.5.1 Profitability Changes All of the subsamples show relevant postprivatization improvements in operating We measure profitability by the return on sales efficiency. We observe that eighteen out of twenty indicator (ROS - net income to sales) as in (D’Souza two mean increases, and seventeen out of twenty two and Megginson, 1999). Table 4 summarizes the median increases. The results are significant, based results for the full sample of all privatized firms. As on the Wilcoxon test and the majority of the we expected, profitability increases significantly after proportion test statistics (mean and median) are privatization, when testing ROS for the full sample of significantly positive at the 5 % level. Nevertheless, 42 companies. The mean (median) increase in ROS not all the subsamples experience identical efficiency after divestiture is 6.0 percentage points (6.0 points), improvements. The Kruskal-Wallis tests show that from 10 to 16 percent of sales (21 to 27 percent) and privatizations by IPO experience greater efficiency 67 % of the companies, experienced an improvement gains than from. We also observed that Total on the average ROS after privatization and 63% Privatizations tend to show a more significant observed an increase on its median. Wilcoxon tests improvement on efficiency than do Partial show that ROS increases significantly (at the 5 Privatizations. percent level) after privatization. As we expected, either for the sample or for the All the subsamples also present significant subsamples, divested companies improve their postprivatization improvements in profitability. operating efficiency and reach their objective more Eleven put of the twenty two firms observed a frequently as by governments launch their statistically significant increase in the mean and the privatization programs. Our results agreed with median based on the Wilcoxon test. The majority of (D’Souza and Megginson, 1999), (Boubakri and the proportion test statistics for mean and median are Cosset, 1998), (Boardman et al., 2000) and (Clamote, also significantly positive. The Kruskal-Wallis test 2000), who concluded that by the privatization also a shows significant difference in average levels mechanism, governments clearly hope that these between most of subgroups, at five percent firms will employ their human, financial and significance level of significance. These results mean technological resources more efficiently; the that by partitioning our total data into these sub- shareholders (including employees) in a private samples, we find that there are significant differences. company capture most of the benefits of efficiency In spite of some performance improvements in improvements, but they also suffer most if, in some non-competitive markets, as we expected, exceptions, efficiency is not improved. performance improvements in competitive markets are more robust, as they have more significant 2.5.3 Capital investment performance changes than in non-competitive markets. With the exception of (Jain and Kini, 1994), We compute capital investment spending using the who found a significant decrease in profitability, our real capital expenditure to sales indicator (RCESA - results confirm most of the expectations about the real capital expenditure to sales) as in (D’Souza and competitive markets, with stronger performance Megginson, 1999). The capital investment spending improvements after privatization than in non- indicator increases significantly after privatization for competitive markets, in line with (Megginson et al., the full sample of 42 companies. The mean (median) 1994), (Boubakri and Cosset, 1998), (Verbrugge et increase in RCESA after divestiture is 8 percentage al., 2000), (Clamote, 2000), (Boardman et al., 2000) points (9 points), from 53 percent to 61 percent of and (Dewenter and Malatesta, 2000).. sales (31 to 40 percent). All of the subsamples also present significant 2.5.2 Operating Efficiency postprivatization improvements in profitability. Five out of twenty two mean increases and six out of We measure operating efficiency with the sales twenty two median increases are significant, based on efficiency (SALEFF - sales to total employment) as the Wilcoxon test and the majority of the proportion in (D’Souza and Megginson, 1999). Sales per test statistics (mean and median) are significantly employee increased from an average (median) 92 % positive. The proportion test also reflects greater

103 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

investment spending since 54 % (52 %) of the firms the majority of the proportion test statistics (mean and in our sample report higher values of this ratio after median) are significantly positive. privatization. Wilcoxon tests show that RCESA Our results confirm most of the expectations increases significantly (at the 5 % level) after about this financial indicator, showing a very privatization. However, the investment intensity is significant increase, as in (Jain and Kini, 1994), not the same for all subsamples. The Kruskal-Wallis (Macquieira and Zurita, 1996), (D’Souza and tests indicate that privatizations by IPO experience Megginson, 1999), (D’Souza and Megginson, 2000), greater investment spending than do firms privatized (Clamote, 2000) and (Megginson et al., 1994). These by Direct Sale. Listed companies also perform better last authors concluded that governments hope and than unlisted companies and firms that restructured expect that real sales will increase after privatization before privatization also perform better than that the due to better incentives, more flexible financing others that did not restructure. opportunities, increased competition and greater As we expected, capital spending increases scope for entrepreneurial initiative. Contrary to our significantly after privatization; our results are closed results are (Boubakri and Cosset, 1999) who did not to (Macquieira and Zurita, 1996), (D’Souza and find a significant increase, and (Boycko et al., 1993) Megginson, 2000), (Boardman et al., 2000) and who argued that effective privatization will lead to a (Megginson et al., 1994). These last authors reduction in output, since the government can no suggested a list of reasons why expecting that longer entice managers (through subsidies) to privatized companies would increase capital spending maintain inefficiently high output levels. after divesture. First, their initial public offering companies have greater access to private debt and to 2.5.5 Employment the equity market than most SOE. Secondly, if privatization is sided by deregulation, the former SOE We compute employment using as proxy the total will face very large investment needs in order to employment (EMPL – total number of employees) as become competitive with other private companies. In in (D’Souza and Megginson, 1999). Overall, we find third place, SOE tend to stress labour over capital that employment decreases significantly after inputs in their production processes, and the power of privatization; we compare the average (median) for politicians, labour unions and other interest groups the pre-privatization period to the average (median) tend to leave SOE employees rich and capital poor. In level for the post-privatization period. The decrease fourth place, removal of government control of the in average (median) levels following privatization is SOE also reduces the government’s ability to force significant at the 5% significance level. The SOE managers to overproduce politically attractive Wilcoxon test shows a significant average (median) but economically wasteful goods. Finally, as decrease in employment from 2173 employees (2095 privatizations promote entrepreneurship, former SOE employees) to 1827 employees (1837 employees). will have the incentive and the means to invest in Our proportion test statistic also reflects less growth options both at home and abroad. employment mean levels since 63 percent (69 percent) of our sample report lower values of 2.5.4 Real Output employment in the years after privatization. Nevertheless, the percentage of decline is not We test for changes in output by calculating the real the same for all subsamples. The Kruskal-Wallis test sales indicator (SAL – nominal sales to consumer shows significant differences in mean (median) levels price index) as in (D’Souza and Megginson, 1999) between the majorities of the subgroups, all at five for the preprivatization period and comparing it to the percent significance level of significance. These three-year average level for the postprivatization results means that by partitioning our total data into period. We found that real output increases these sub-samples, we find that there are significant significantly after privatization. Average (median) differences. SAL, rises from 94 % (97 %) to 218 % (242 %); the Our results confirm most of the expectations increase in mean and median levels following about employment, that is, a very significant privatization is significant at the 5 % significance decrease, as (D’Souza and Megginson, 1999), level. The proportion test also indicates that real (Dewenter and Malatesta, 2000), (Boardman et al., output significantly rises in the years following 2000) and (Clamote, 2000). This differs from the privatization since 69 % (71 %) of our sample report results of (Megginson et al., 1994), who found higher values of this indicator in the years after insignificant decreases on employment, and, for privatization. them, the great fear of all governments contemplating Most of the subsamples also present significant privatization programs is that efficiency and postprivatization improvements in profitability. We profitability will be achieved only at a cost: observe eleven out of twenty two mean increases and unemployment; that is, governments expect large twelve out of the twenty two median increases. All of declines in employment levels after privatization. them are significant based on the Wilcoxon test and Also, prior to privatization, most SOE tend to be overstaffed. Thus, in order to increase efficiency,

104 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

extensive layoffs would be expected following (mean and median) are significantly positive. The privatization and the reduction of subsidies. In Kruskal-Wallis test does not show significant contrast, the results of (Macquieira and Zurita, 1996) differences in mean (median) levels for most of and (Boubakri and Cosset, 1998) all showed subgroups, at 5% significance level. These results employment rises after privatization. mean that by partitioning our total data into these sub-samples, we find that there are no significant 2.5.6 Dividend Policy differences. Our results confirm our expectations, since, for We measure dividend policy with the dividend to most of the sample, if capital expenditures and sales (DIVSAL – dividend to sales) as in (D’Souza dividends in absolute terms increased, treasury and Megginson, 1999) in order to examine whether applications should increase too. The only component dividend payments increases following privatization. of treasury applications that had decreased after The average (median) dividend payment increases privatization was cash and banks. from 1 percent (1 percent) to 7 percent (4 percent). The proportion test also shows that dividend to sales 2.5.8 Activity Levels significantly rises in the years following privatization; both the Wilcoxon and proportion tests show that We compute activity levels using the sales to total dividend to sales increase significantly after assets indicator (STA - Sales divided by Total privatization. Whatever the subsamples, our Assets). The activity levels indicator increases proportion test statistic reflects a higher rate of significantly after privatization for the full sample of dividend to sales, mean (median) levels, since 64 42 companies. The mean (median) increase in STA percent (54 percent) of our sample report higher after divestiture is 40 percentage points (17 points), payout ratio after privatization. from 52 percent to 91 percent of sales (39 to 56 The Kruskal-Wallis tests show that dividend percent). payments increase significantly more for Some subsamples also present significant privatizations by IPO than for privatizations by Direct postprivatization improvements in activity levels. Sale, significantly more for firms with listed stocks Eight out of twenty two mean (median) increases are than for non-listed stocks, significantly more for significant based on the Wilcoxon test and most of firms with shareholders in the firms’ management the proportion test statistics (for mean and median) after privatization, than for firms without are significantly positive. The proportion test also shareholders in management after privatization. reflects relevant improvements in activity levels, Our results confirm most of the expectations since 58 % (55 %) of our sample report higher values about dividend policy, that is, a very significant of this ratio following privatization. Wilcoxon tests increase after privatization, as (Megginson et al., show that the STA ratio increases significantly (at the 1994), (Macquieira and Zurita, 1996), (Clamote, 5 % level) after privatization. 2000) and (Boubakri and Cosset, 1998). They The Kruskal-Wallis tests shows that concluded that, following privatization, dividend privatizations in competitive markets experience payments increase in result of the pressure of private greater activity levels than do firms in investors for dividends and dividend payments, as noncompetitive markets, the non financial sector opposed to governments. The classic response to the shows higher activity levels than the financial sector. atomized ownership structure to which most Firms with foreign allocation of control present privatization programs lead is, therefore, the increase higher activity levels than firms with national on dividends payments. allocation of control, privatizations by IPO show higher activity levels than privatizations by Direct 2.5.7 Treasury Sale, listed firms perform better than firms not listed, firms with shareholders in management show higher We compute treasury with the treasury applications levels than do firms without shareholders in indicator (TA – cash and banks plus dividends plus management, total privatizations show higher activity capital expenditure). We find that that treasury levels than partial privatizations and firms that applications increase after privatization. For the full restructured before privatization, perform better than sample of all privatized firms, the mean (median) TA do firms that did not restructure before privatization. rises after privatization, mainly due to the increase of So far, our empirical results show several capital expenditures. The proportion test also indicators (operational and financial) which are indicates that treasury applications significantly rise improved after privatization. One of reasons, among in the years following privatization, since 54 % (52 others, is the approach that comes from new owners %) of our sample report higher values of this and managers. With new and possibly better indicator in the years after privatization. motivated managers, we certainly have better Most of the subsamples also present inventory management (raw materials, finish goods, postprivatization improvements in treasury etc.) and, consequently, higher activity levels and applications and most of the proportion test statistics profitability.

105 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

their investments, and tending to attain the long term 2.5.9 Short Term Equilibrium equilibrium.

We compute short term equilibrium with the cash and 2.5.11 Capital Structure banks to short term debt (CBTSTD – cash and banks divided by short term debt). We find that the short We measure capital structure with total debt to total term equilibrium is worst after privatization. Average assets (TDTA – total debt divided by total assets, as (median) CBTSTD decreases from 20 % (16 %) to 16 in (Boubakri and Cosset, 1998). We find that the % (13 %). The proportion test indicates that the short capital structure is improved after privatization. Mean term equilibrium declines in the years following (median) TDTA decreases from 74 % (79 %) to 62 % privatization. Only 23 % (19 %) of our sample report (63 %). The proportion test shows that the capital improved short term equilibrium in the years after structure is significantly improved following privatization. The subsamples also present privatization. In addition, 65 % (61 %) of all postprivatization decrease in short term equilibrium. companies experienced an improved capital structure The Kruskal-Wallis test shows differences in mean after privatization. This highly significant increase (median) levels between most of subgroups (seven shows that, the increase on total assets was relatively out of 11), all at five percent significance level higher than the increase on total debt after We may have expected that, after privatization, privatization. The Kruskal-Wallis test shows the financial equilibrium of the firms would improve significant differences in mean (median) levels slightly in the short term, during the first years of between most of subgroups (seven out of eleven), all private management. However, either for the total at 5% significance level. samples either for most of the subsamples, we find a Our results confirm most of the expectations decrease on short term financial equilibrium after about improvements on capital structure after privatization. This may be the result of an increase in privatization. (Macquieira and Zurita, 1996), investment spending, restructure and develop in result (Boubakri and Cosset, 1998), (Clamote, 2000) and of management decisions that, sometimes, require a (Megginson, et al., 1994) concluded that, while most great amount of money, such as, lay-off workers (as governments do not place great priority on improving we observed earlier in this chapter. In order to finance financial soundness of the newly privatized firms, these restructuring decisions, companies are forced to most do expect capital structure to improve dropping get financing (most of times, short term financing) the leverage ratios after privatization, because SOEs, and, in consequence, their financial structure may be traditionally, have very high debt levels, the only affected in short term. ways of equity available to those firms are capital injections from governments and retained earnings if 2.5.10 Long Term Equilibrium they exist. On the other hand, the switch from public to We measure long term equilibrium with net cash flow private ownership should lead to more consistent to long term debt (NCFTLTD – net cash flow divided capital structure because the government’s removal of by long term debt). We find that the long term debt guarantees will increase the firms’ cost of equilibrium is worst after privatization. Average borrowing and because companies will have (median) NCFTLTD decreases from 26 % (7 %) to increased access to public equity markets. 20 % (3 %). The proportion test shows that the long term equilibrium declines following privatization. 2.6 Summary and conclusions Only 37 % (31 %) of our sample report improved performance following privatization. Over the last fifteen years, the privatization process The subsamples also present a postprivatization has transformed the Portuguese economic landscape decrease in long term equilibrium. The Kruskal- throughout a sweeping reduction of the role of the Wallis test shows significant difference in mean state in the economy. An economic event such (median) levels between most of subgroups, all profound raises many important questions – most of significant at 5% significance level (five on total). which are, as yet, not completely answered. We These results mean that by partitioning our total data investigate whether or not the privatization of SOE into these sub-samples, there are significant had caused improvements on the economic and differences. financial performance of those privatized companies, We expected that, after privatization, the as it is suggested by the literature of property rights, financial equilibrium of the firms would improve public choice and agency theory. slightly in the long term. However, either the total Abroad, so far, most of the work done on sample either most of the subsamples show a privatizations and their effects on the divesture decrease on long term financial equilibrium after companies, had been oriented just to the economic privatization. These results are puzzling since, we and social side of the firm (profitability, operating expect that, in the long run, after the restructuring efficiency and output), ignoring its financial side, as decisions, firms tend to consolidate, to profit from happen with (Megginson et al., 1994), (Macquieira

106 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

and Zurita, 1996), (Boubakri and Cosset, 1998), (D’ References Souza and Megginson, 1999), (Verbrugge et al., 2000), (Dewenter and Malatesta, 2000) and 1. Almeida, J. M. and João Duque, (2000), ‘Ownership (Boardman et al., 2000), among others. In Portugal, structure and Initial Public Offerings in Small the only existing work on the performance of newly Economies – The case of Portugal’, working paper, privatized firms, (Clamote, 2000), has the same ISEG – Instituto Superior de Economia e Gestão, Universidade Técnica de Lisboa. economic orientation. Our work, filling in this gap, 2. Alves, Paulo and João Duque, (1997), ‘Novos adds the financial to the economic and social Resultados Empíricos sobre Estratégias de perspective of the divesture firm performance. Investimento Contrárias no Mercado de Capitais In addition, the same authors have considered, Português’, Cadernos do Mercado de Valores on their study, just a full, aggregate sample, with a Mobiliários, CMVM – Comissão do Mercado dos few exceptions, as, (Clamote, 1999) and (D’Souza & Valores Mobiliários, 1. Megginson, 1999), whose work only considered a 3. Alpalhão, R., 1988, ‘Ofertas Públicas Iniciais: O caso few criteria to study the sample separately: Português’, working paper nº 100, Universidade Nova competition, total (control) privatization, new CEO de Lisboa, Faculdade de Economia. 4. Anderson, Seth C., T. Randolph Beard and Jeffrey A. and new Board of Directors. Our work excludes the Born, (1995), ‘Initial Public Offerings – Findings and new CEO criteria but adds others: the listing process, Theories’, Kluwer Academic Publishers. the pre-restructurings, the timing of the privatization, 5. Anderson, S., A de Palma and J. Thisse, (1997) the existence of shareholders in management, the ‘Privatization and Efficiency in a differentiated concentration of capital, the financial and non- Industry.’ European Economic Review, 41, pp. 1635- financial sector, the foreign and national control and 1671. the method of privatization. We are sure that these 6. Ang, James S., Rebel A. Cole and James Wuh Lin, multi-industry samples improve the quality of the (2000), ‘Agency Costs and Ownership Structure’ The empirical results, in a more analytical way, and Journal of Finance”, 55 (1), pp. 81-106. provides additional perspectives to a better 7. Aussenegg, Wolfgang, (2000), ‘Privatization versus Private Sector Initial Public Offerings in Poland”, understanding of the postprivatization firm Working Paper, Vienna University of Technology. performance. We work with a sample composed of 8. Barberis, N., M. Boycko, A. Shleifer and N. 42 companies from all activity sectors, with analysis Tsukanova, (1996) ‘How does Privatization Work: for the period 1989 to 2003. This goes far beyond Evidence from Russian Shops. “Journal of Political another work with Portuguese data, (Clamote, 2000) Economy, 104, pp. 764-790. that uses only 20 companies for the period 1989 to 9. Barclay, Michael and Clifford Holderness, (1989), 1995 period. ‘Private Benefits from Control of Public On the economic side, we document significant Corporations’, Journal of Financial Economics 58 (1), pp. 371-396. improvements on profitability, operating efficiency, 10. Bennell, Paul, (1997), ‘Privatization in Sub-Saharan capital investment, and real output and activity levels. Africa: Progress and Prospects during the 1990s’ Firms on the competitive sector of the economy, World Development, 25 (11) pp. 1785-1803. firms with changes on more than 50% of the Board, 11. Bhagat, Sanjai and Peter A. Frost, (2000), ‘Issuing firms in the nonfinancial sector, firms with foreign Costs to existing shareholders in competitive and control, firms with concentrated structures after negotiated underwritten public utility equity offerings’, privatization, firms with stocks listed in an Exchange, Journal of Financial Economics, 19, pp. 23.39. firms with shareholders in management, firms that 12. Brennan, M.J. and J. Franks, (1997),’Underpricing, changed the total control after privatizations and Ownership and Control in Initial Public Offerings of Equity Securities in the U.K.”, Journal of Financial firms that restructured, have a more significant Economics, 45 (1), pp. 391-413. operational improvement than the correspondent 13. Bishop, Matthew R. and John A. Kay, (1989), opposite subsample. ‘Privatization in the United Kingdom: Lessons from On the social side, we experience a significant Experience’ World Development, 17 (5), pp. 643-657. decrease on employment after privatization, which 14. Bolsa de Derivados do Porto, (1996), ‘Processos de means that, after the privatization process, the (RE) Privatização – Sociedades Cotadas (1989/1996) personal costs, among others, are some of the priority ”, Porto: Bolsa de Derivados do Porto, Instituto do fixed costs that are cut-off by management. The more Mercado de capitais. significant decline of this type of costs is experienced 15. Bortolotti, Bernardo, Juliet D’Souza, Marcella Fantini and William Megginson, (2001), ‘Sources of by the sub-samples mentioned on the last paragraph. performance improvement in privatized firms: a Lastly, on the financial side, we document a clinical study of the global telecommunications decline on the short and long term equilibrium, which industry’, Working Paper, University of Oklahoma. means that companies, after privatization, at least 16. Boubakri, N. and J. Cosset, (1998), ‘The Financial and during the first years, have some financial instability, Operating Performance of Newly-privatized Firms: mainly due to a great effort to finance their growth Evidence from Developing Countries”. Journal of and capital investments, lay-offs, and to finance the Finance, 53 (3), pp. 1081-1110. redirection of their commercial and marketing 17. Boutchkova, Maria K. and W. Megginson, (2000), strategy. ‘Privatization and the Rise of Global Capital Markets” Financial Management.

107 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

18. Boyco, M.; A. Shleifer and R. W. Vishny, (1994), firms during the 1990s’, The journal of Finance, 54 ‘Voucher Privatization’, Journal of Financial (4), pp. 1397-1438. Economics, 35, pp. 249-266. 37. Duque, João, (1999), ‘A Regulação do Sistema de 19. Boyco, M.; A. Shleifer and R. W. Vishny, (1996), ‘A Valores Mobiliários – Uma Abordagem pela Teoria Theory of Privatization”, The Economic Journal, 106, Financeira”, Cadernos do Mercado de Valores pp. 309-319. Mobiliários, CMVM – Comissão do Mercado dos 20. Brada, Josef C, (1996), ‘Privatization is Transition – Valores Mobiliários, 6. Or is it?’ Journal of Economic Perspectives, 10, (2), 38. Duque, João and Florbela Razina, (1998), ‘A pp. 67-86. interdependência entre os Mercados Financeiros 21. Brown, J. David, John S. Earle and Álmos Telegdy, Internacionais” Cadernos do Mercado de Valores (2004), ‘Does Privatization Raise Productivity? Mobiliários, CMVM - Comissão do Mercado de Evidence from Comprehensive Panel Data on Valores Mobiliários, 3. Manufacturing Firms in Hungary, Romania, Russia 39. Eckel, C., D. Eckel and V. Singal, (1997), and Ukraine’, Upjohn Institute Staff Working Paper, ‘Privatization and Efficiency: Industry Effects of the pp. 54-107. Sale of British Airways’, Journal of Financial 22. Caves, Douglas W. and L. R. Christensen, (1980), Economics, 43, pp. 275-298. ‘The Relative Efficiency of Public and Private Firms 40. Ehrlich, Isaac, G. Gallais-Hamonno, Z. Liu, and R. in a Competitive Environment: The Case of Canadian Lutter, (2000), ‘Productivity Growth and Firm Railroads” Journal of Political Economy, 88 (5), pp. Ownership: An Analytical and Empirical 958-976. Investigation” Journal of Political Economy, 102, (5), 23. Chemmanur, Thomas J., and Paolo Fulghieri, (1999) pp. 1006-1038. ‘A Theory of the Going-Public Decision’, The Review 41. Estrin, Saul and V. Pérotin, (1991), ‘Does Ownership of Financial Studies, 12 (2), pp. 249-279. always Matter?’ International Journal of Industrial 24. Chemmanur, Thomas J., (2001), ‘The pricing of Organization, 9, pp. 55-72. IPO’s: a dynamic model with information production´, 42. Fama, Eugene, (1980), ‘Agency Problems and the The Journal of Finance, 48 (1), pp. 285-319. Theory of the Firm’, Journal of Political Economy, 88, 25. Claessens Stijn, Simeon Djankov and Larry H.P. Lang, (2), pp. 288-30. (2000), ‘The Separation of Ownership and Control in 43. Fama, Eugene and Michael Jensen, (1983), East Asian Corporations’, Journal of Financial ‘Separation of Ownership and Control’, Journal of Economics, 58, pp. 81-112. Law & Economics, 26, pp. 375-393. 26. Clamote, C. A., (1999), ‘O Impacto da Privatização no 44. Febra, Lígia Catarina Marques, (2000), ‘Ofertas Desempenho Financeiro e Operativo das Empresas Públicas Iniciais. Razões para o seu Lançamento em privatizadas em Portugal’, Tese de Mestrado em Portugal’, Tese de Mestrado em Ciências Gestão e Estratégia Industrial, ISEG, Lisboa. Empresariais, Faculdade de Economia da Universidade 27. Corwin, Sahne A., and Jeffrey H. Harris, (2001), “The do Porto, 2001. Initial Listing Decisions of Firms that go Public” 45. Ferling, Rhona L., (1994), ‘Should You go Public?’ Financial Management, 30 (1), pp. 35-55. Financial Executive. 28. Cragg, M and Dyck, A., (1999) “Management Control 46. Field, Laura Casares, (1998), ‘The IPO as the First and Privatization in the United Kingdom” Rand Stage in the Sale of the Firm’ Working Paper, Penn Journal of Economics, 30 (3), pp. 475-497. State University. 29. Cragg, M and Dyck, A., (1998), ‘Privatization, 47. Fonseca, Paulo, (2004), ‘Monopólios Públicos vs Compensation and Management Incentives: Evidence Monopólios Privados: eficiência produtiva em public from the United Kingdom’, Working Paper, Harvard utilities’ – Estudos de Gestão, Portuguese Journal of Business School. Management Studies. 30. Cunha, R.C., (1999), ‘Mudanças nas Empresas 48. Garner, (1989), ‘Has Public Enterprise Failed?’ in Privatizadas: A Cultura Empresarial, in Ministério das Ramanadham, V.V. (eds) - Privatization in the United Finanças (eds), Privatizações e Regulação – A Kingdom, London: Routledge. Experiência Portuguesa’, Lisboa, M. F. – Direcção 49. Goergen, Marc G. J., (1995), “The Evolution of Geral de Estudos e Previsão. Ownership and Control in German IPOs: A Dynamic 31. Demsetz, Harold, (1983), ‘The structure of Ownership Analysis’, Part of the PHD Thesis at University of and the Theory of the Firm’, Journal of Law and Oxford, Oxford, UK. Economics 25, pp. 375-390. 50. Gonçalves, Victor F. C., António Palma dos Reis e 32. Demsetz, Harold and Kenneth Lehn, (1985), ‘The João Duque, (1997), ‘Strategic Performance Analysis structure of Corporate Ownership: causes and of the Portuguese Financial Services Industry’. consequences”, Journal of Political Economy, 93, pp. Working Paper, Cadernos de Económicas, Documento 1155-1177. de Trabalho nº 4/97, Departamento de Gestão, ISEG, 33. Dewenter, Kathryn L. and Paul H. Malatesta, (2001), Universidade Técnica de Lisboa. ‘State-Owned and Privately Owned Firms: An 51. Gonçalves, Victor F. C., António Palma dos Reis e Empirical Analysis of Profitability, Leverage and João Duque, (1997), ‘The Impact of the Corporate Labor Intensity’, The American Economic Review. Strategy on the Information System Strategy: A 34. Dewenter, Kathryn L. and Paul H. Malatesta, (1996), Survey of Portuguese Financial Corporations’, ‘Public Offerings of State and Privately-Owned Working Paper, Cadernos de Económicas, Documento Enterprises’ The Journal of Finance, pp. 1660-1679. de Trabalho nº 2/97, Departamento de Gestão, ISEG, 35. Donaldson, David, (1995), ‘Privatization: Principles Universidade Técnica de Lisboa. and Practice’. Washington, D.C. The World Bank. 52. Gugler Klaus, Susanne Kalss, Alex Stomper and Josef 36. D’Souza, Juliet, and William L. Megginson, (1999), Zechner, (1997), ‘The Separation of Ownership and ‘The financial and operating performance of privatized

108 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

Control: An Austrian Perspective”, Journal of Performance of Newly Privatised Firms: An Financial Economics. International Empirical Analysis” The Journal of 53. Gupta, Nandini, (2002) “Partial Privatization and Firm Finance, 49 (2), pp. 403-452. Performance” Working Paper, Michigan Business 72. Megginson, William L. and Jeffry M. Netter, (1997), School. ‘Equity to the People: the record on share issue 54. Harper, Joel T., (2002), ‘The Performance of privatization programs’, Privatisational (International) Privatized Firms in the Czech Republic’, Journal of yearbook, pp. 27-34. Banking and Finance, 26, pp. 621-649. 73. Megginson, William L. Robert C. Nash, Jeffry M. 55. Havrylyshyn, and McGettigan, D., (1999), Netter and Adam Schwartz, (2000), “The Long-Run ‘Privatization in Transition Countries: A Sampling of Return to Investors in Share Issue Privatization’, the Literature” IMF Working Paper. Financial Management, 29 (1). 56. Holmen, Martin and Peter Högfeldt, (1999) ‘Motives, 74. Megginson, William L. Robert C. Nash, Jeffry M. Underpricing and Ownership structure in Swedish Netter and Annette B. Poulsen, (2000), ‘The Choice of IPO’s’, Journal of Financial Economics. Private versus Public Capital Markets: Evidence from 57. Ibbotson, F.G., J.L.Sindelar, and J.R. Ritter, (1988), Privatizations” Working Paper, University of Georgia. "Initial Public Offerings’, Journal of Applied 75. Mello, Antonio Sampaio and John E. Parsons, (1998), Corporate Finance, pp. 37-45. ‘Going Public and the Ownership Structure of the 58. Jain, Bharat A., and Omesh Kini, (1994), ‘The post- Firm’, Journal of Financial Economics, 49, pp. 79-109. issue operating performance of IPO firms’, Journal of 76. Mikkelson, Wayne H., M Megan Partch and Kshitij, Finance, 49, pp. 1699-1726. (1997), ‘Ownership and Operating of Companies that 59. Jelic, Ranko and Richard Briston, (1999), ‘Hungarian go Public’, Financial Economics, 44, pp. 281-307. Privatisation Strategy and Financial Performance of 77. Pagano, Marco, Fabio Panetta and Luigi Zingales, Privatised Companies’, Journal of Business Finance & (1998), ‘Why Do Companies Go Public? An empirical Accounting 26 (9), pp. 1319-1357. Analysis’, Journal of Finance, 53, pp. 27-64. 60. Jensen, Michael C. and William H. Meckling, (1976), 78. Pagano, Marco, and Ailsa Röell, (1998), ‘The Choice ‘Theory of the Firm: Managerial Behaviour, Agency of Stock Ownership Structure: Agency Costs, Costs and Ownership Structure’, Journal of Financial Monitoring, and The Decision to Go Public’, Quarterly Economics, 3, pp. 305-360. Journal of Economics, pp. 23-49. 61. Jones, Steven L., W. Megginson, R. Nash and Jeffry 79. Ministério das Finanças, (1995), “Privatizações em Netter, (2000), ‘Share Issue Privatizations as Financial Portugal – Uma Reforma Estrutural”, Lisboa, M. F. Means to Political and Economic Ends’ Journal of GAFEEP – Gabinete para a Análise do Financiamento Financial Economics. do Estado e das Empresas Públicas. 62. Kim, Jeong-Bon, Itzhak Krinsky e Jason Lee, (1993), 80. Ministério das Finanças, (1999), ‘Privatizações e ‘Motives for going public and underpricing: new Regulação: A Experiência Portuguesa”, Lisboa, findings from Korea’, Journal of Business Finance & Ministério das Finanças – Direcção Geral de Estudos e Accounting, pp. 195-211. Previsão. 63. Kunz, Roger M; Aggarwal, Reena, (1994), ‘Why 81. Moore, John, (1992), ‘British Privatization – Taking initial public offerings are underpriced: Evidence from Capitalism to the People” Harvard Business Review. Switzerland’, Journal of Banking and Finance, 18 (4) 82. Nellis, John, (1994), ‘Is Privatization Necessary?’ pp. 705-723. Private Sector – The World Bank, FPD, Note Nº 7. 64. La Porta, Rafael, Francisco Lopez-de-Silanes and 83. Nellis, John and S. Kikeri, (1989), “Public Enterprise Andrei Shleifer, (1999), ‘Corporate Ownership Reform: Privatization and the World Bank” World Around the World’, Journal of Finance 54, pp. 471- Development, 17 (5) pp. 659-672. 517. 84. Osório, A. e Santos, V., (1999), ‘ A Eficiência 65. La Porta, Rafael and Florencio López-de-Silanes, Económica das Empresas Públicas e Privadas: Uma (1999), ‘The Benefits of Privatization: Evidence from Análise Comparada’, in Ministério das Finanças (eds), Mexico”, The Quarterly Journal of Economics. Privatizações e Regulação – A Experiência 66. Levis, M., (1996), ‘Empirical Issues in Raising Equity Portuguesa, Lisboa, M. F. Direcção Geral de Estudos e Capital – Advances in Finance, Investment and Previsão. Banking’, Elsevier Science Publishers. 85. Parker, David, (1994), ‘Privatization and Business 67. López-de-Silanes, Andrei Schleifer and Robert W. Restructuring: Change and Continuity in the Privatised Vishny, (1997), ‘Privatization in the United States’ Industries’ The Review of Policy Issues, 1, pp. 3-27 Rand Journal of Economics, 28, (3), pp. 447-471. 86. Parker, David, (1995), ‘Privatization and Agency 68. McConnell, John and Henri Servaes, (1990), Status: Identifying the Critical Factors for Performance ‘Additional Evidence on Equity Ownership Structure Improvement, British Journal of Management, 6, pp. and Firm Performance’, Journal of Financial 29-43. Economics, 27, pp. 595-612. 87. Perotti, Enrico C., (1995), ‘Credible Privatization’ The 69. Megginson, William L., Robert C. Nash and Mathias American Economic Review, September, 85 (4). Van Randenborgh, (1994), ‘The Financial and 88. Pryke, Richard, (1982), ‘The Comparative Operating Performance of Newly Privatized Firms: an Performance of Public and Private Enterprise’ Fiscal International Empirical Analysis, The Journal of Studies, 3, pp. 68-81. Finance, 49 (2). 89. Ramamurti, Ravi, (1997), ‘Testing the Limits of 70. Megginson, William L. and Jeffry M. Netter, (2001), Privatization: Argentine Railroads’, World ‘From State to Market: A survey of empirical studies Development, 25 (12), pp. 1973-1993. on privatization’, Journal of Economic Literature. 90. Ramos, Luís Moura, (2002), ‘Privatization in Portugal: 71. Megginson, William L. Robert C. Nash, and M. Employee Owners or Just Happy Employees?’ Randenborgh, (1994), “The Financial and Operating Estudos do Grupo de Estudos Monetários e

109 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

Financeiros (GEMF) -Faculdade de Economia da português”, Dissertação para a obtenção do grau de Universidade de Coimbra, Nº 7. Mestre, INDEG/ISCTE. 91. Rees, Ray, (1988), Inefficiency, Public Enterprise and 109. Tomas, Carlos, (1993), ‘The Portuguese Market", Privatization’ European Economic Review, 32, pp. Global Investor - Portuguese Handbook, pp. 15-25. 422-431. 110. Verbrugge, James A., William L. Megginson and 92. Ritter, Jay R., (1991), ‘The Long-run Performance of Wanda L. Owens, (1999), “State Ownership and the Initial Public Offerings’, Journal of Finance 46, pp. 3- Financial Performance of Privatized Banks: an 27. Empirical Analysis” World Bank/Federal Reserve 93. Ritter, Jay R., (1987), ‘The Costs of Going Public’, Bank of Dallas Conference on Banking Privatization, Journal of Financial Economics, 17, pp. 269-281. Washington. 94. Röel, A., 1996, ‘The decision to go public: an 111. Vickers, John and G. Yarrow “Privatization: Not the overview’, European Economic Review, 40, pp. 1071- Only Answer”, World Development, 17, (5), pp. 633- 1081. 641. 95. Ross, Stephens A., (1973), ‘The Economic Theory of 112. Vining, Aidan R. and A. Boardman, (1992), Agency: The Principal’s Problem’, American ‘Ownership versus Competition: Efficiency in Public Economic Review, 63, pp. 134-139. Enterprise”, Public Choice, 73, pp. 205-239. 96. Santos, Anibal, (1989), ‘Models of Privatization and 113. Walle, Nicolas van de, (1989), ‘Privatization in Market Structure’ Economia, pp. 56-81. Development Countries: A Review of the Issues’, 97. Sappington, David E. M., (1991), ‘Incentives in World Development, 17 (5), pp. 601-615. Principal – Agent Relationships’, Journal of Economic 114. Walking, Ralph and Michaek Long, (1984), Agency Perspectives, 5, (2), pp. 45-66. Theory, Managerial welfare and Takeover Bid 98. Schultz, P., (1993), ‘Unit initial public offerings: a Resistance’, Rand Journal of Economics 15, pp. 54-68. form of staged financing’, Journal of Financial 115. Weisbach, M., (1988), ‘Outside Directors and CEO Economics, 34, pp. 199-229. turnover’, Journal of Financial Economics 20, pp. 431- 99. Shafic, Nemat, (1995), ‘Making a Market: Mass 460. Privatization in the Czech and Slovak Republics” 116. Willig, Robert D., (1993) “Public versus Regulate World Development, 23 (7), pp. 1143-1156. Private Enterprise” Proceedings of the World Bank 100. Shleifer, Andrei. (1998), ‘State versus Private Annual Conference on Development Economics, pp. Ownership” Journal of Economic Perspectives, 12 (4), 155-180. pp. 133-150. 117. Witmer, Richard H; Molner, John P, (1996), ‘What 101. Shleifer, A. and Vishny, R., (1996) “A Theory of you need to know before an IPO", Corporate Finance, Privatization” Economic Journal, 106, pp. 309-319. 140, pp. 53-54. 102. Shleifer, Andrei and Robert Vishny, (1997), ‘A Survey 118. World Bank, (1997), ‘Between State and Market: of Corporate Governance’, Journal of Finance 52, pp. Mass Privatization in Transition Economies’, 737-783. Washington, D.C. The World Bank. 103. Schmidt, Klaus M., (1996), ‘The Costs and Benefits of 119. Wu, Shuo and Hong Cui, (1999), ‘Ownership Privatization: An Incomplete Contracts Approach’ The Structure and Corporate Performance: Evidence from Journal of Law, Economics & Organization, 12 (1). the Mainland China”, Economics & Finance 104. Smith, Stephen C., Beom-Cheol Cin and Milan Department, City University of Hong Kong, 65-86. Vodopivec, (1997), ‘Privatization Incidence, 120. Zhou, Xianming, (2001), ‘Understanding the Ownership Forms and Firm Performance: Evidence Determinants of Managerial Ownership and the Link from Slovenia’, Journal of Comparative Economics, Between Ownership and Performance: Comment’, 25, pp. 158-179. Journal of Financial Economics, 62, pp. 559-571. 105. Singh, Manohar and Wallace N. Davidson III, (2003), 121. Zingales, L., (1995), ‘Insiders ownership and the ‘Agency Costs, Ownership Structure and Corporate decision to go public’, Review of Economic Studies, Governance Mechanisms’, Journal of Banking and 62, pp. 425-448. Finance 27, pp. 793-816. 122. Yarrow George, (1986), ‘Privatization in Theory and 106. Sousa, F. and R. Cruz, (1995), ‘O Processo de Practice’, Economic Policy, pp. 324-377. Privatizações em Portugal’, Porto, Associação 123. Yarrow, George and J. Vickers, (1991), ‘Economic Industrial Portuense. Perspectives on Privatization’ Journal of Economic 107. Stoughton, Neal M., Kit Pong Wong and Josef Perspectives, 5 (2), pp. 111-132. Zechner, (2001), ‘IPOs and Product Quality’, Journal 124. Yotopoulos, Pan A., (1989) ‘The (Rip) Tide of of Business, 74 (3). Privatization: Lessons from Chile”, World 108. Teixeira, Fernando J. C. e Silva Nunes, (2002), Development, 17 (5), pp. 683-702. “Análise das Ofertas Públicas Iniciais (IPOs) – O caso

110 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

Appendix

Table 1. Sample of privatized firms from 1989 to 2009

Percentage of capital that was Company Industry privatized at the date of the issue Alcool Generos Alimentares, SA Galenic Products Industry 100% Aliança Seguradora Insurance 49% Banco de Fomento & Exterior Banking 20% Banco Espírito Santo Banking 40% Banco Fonsecas & Burnay Banking 80% Banco Internacional do Funchal (Banif) Banking 16% Banco Pinto & Sotto Mayor Banking 80% Banco Português do Atlântico Banking 33% Banco Totta & Açores Banking 49% Beralt & Tin Mining and Minerals 100% Bonança Insurance 60% Brisa Services 35% Celbi Cellulose and Paper 100% Centralcer Commerce 100% Cimpor Exploration of Minerals 20% Cive Industry Glassmaker 100% Companhia de Seguros de Crédito, S.A. Insurance 100% Crédito Predial Português Banking 100% Diário de Notícias Media 100% Efacec Services 87% Fisipe Textile 49% EDP Electricity 30% Jornal de Notícias Media 100% Império Insurance 100% Ipetex Textile 100% Lisnave Shipyard 100% Metalsines Metallomechanics 100% Mundial Confiança Insurance 100% Portucel Cellulose and Paper 44% Petrogal Petroleum 25% Portugal Telecom Telecommunication 14% Quimigal Chemicals 90% Rádio Comercial Media 100% Secil Cement Industry 51% Soc. Fin. Portuguesa Banking 100% Sociedade de Cargas e Descargas Marítimas, Transportation 100% S.A. (Socamar) SN-Longos Metallurgic Industry 80% Tabaqueira Tobacco Industry 65% Tranquilidade Insurance 49% Transinsular Transportation 100% União de Bancos Portugueses Banking 61% Unicer Commerce 49%

111 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

Table 2. Sub-Sample Table

COMPETITI DIRECTO CONTRO STRUCTUR METH MARKET MANAGEME RESTRUCTI SECTOR YEAR ON RS L E OD S NT TYPE ON FIRM NoC F NoF NoC SI D L NoL <199 >199 NoS P T C NC CH NaC FC CS SM R NoR H S S S P S x x 0 0 M P P A S X X X X X X X X X X X

Aga X X X X X X X X X X X

BFE X X X X X X X X X X X

BFB X X X X X X X X X X X

Banif X X X X X X X X X X X

BP & SM X X X X X X X X X X X

BPA X X X X X X X X X X X

BTA X X X X X X X X X X X

BES X X X X X X X X X X X

B & T X X X X X X X X X X X

Bonança X X X X X X X X X X X

Brisa X X X X X X X X X X

Celbi X X X X X X X X X X X

Centralcer X X X X X X X X X X X

Cimpor X X X X X X X X X X X

Cive X X X X X X X X X X X

Cosec X X X X X X X X X X X

CPP X X X X X X X X X X X

DN X X X X X X X X X X X

EDP X X X X X X X X X X

Efacec X X X X X X X X X X X

Fisipe X X X X X X X X X X X

Império X X X X X X X X X X X

Ipetex X X X X X X X X X X X

JN X X X X X X X X X X X

Lisnave X X X X X X X X X X X

Metalsines X X X X X X X X X X X

M C X X X X X X X X X X X

Petrogal X X X X X X X X X X X

Portucel X X X X X X X X X X X

P T X X X X X X X X X X X

Quimigal X X X X X X X X X X X R. X X X X X X X X X X X Comercial Secil X X X X X X X X X X X

SN- Longos X X X X X X X X X X X

SFP X X X X X X X X X X X

Socarmar X X X X X X X X X X X

Tabaqueira X X X X X X X X X X X

Tranquilida X X X X X X X X X X X de Transinsula X X X X X X X X X X X r UBP X X X X X X X X X X X

Unicer X X X X X X X X X X

TOTAL 26 16 28 14 16 26 39 3 22 20 23 19 25 17 12 30 26 16 24 18 16 26

NOTES: [C - Competitive market; NC - No Competitive Market]; [CH - Change in the composition in the Board of Directors; NoCH - No change in the composition in the Board of Directors]; [FS - Financial sector; NoFS - No financial sector]; [NaC - National control; FC - Foreign control]; [CS - Concentrated structure; NoCS - No concentrated structure]; [SIP - Share Issue Privatization; DS - Direct sale]; [Lx - Listed in a stock exchange; NoLx - No listed in a stock exchenge]; [<1990 - Privatization before 1990; <1990 - Privatization before 1990]; [SM - With shareholders in management; NoSM - Without shareholders in management]; [PP - Partilal privatization; TP - Total privatixation]; [R - Reestrutured prior to the sale; NoR - Not reestructured prior to the sale].

112 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

Table 3. Summary of Testable Predictions

PREDICTED CARACTERISTICS PROXIES RELATIONSHIP

PROFITABILITY Return on Sales (ROS) = Net Income / Sales ROSA > ROSB

SALEFFA > OPERATING EFFICIENCY Sales Efficiency (SALEFF) = Sales / Total Employment SALEFFB Real Capital Expenditure to Sales (RCESA) = Real Capital RCESAA > CAPITAL INVESTMENT Expenditure / Sales RCESAB

REAL OUTPUT Real Sales (SAL)=Nominal sales/Consumer price index SALA > SALB

TOTAL EMPLOYMENT Total Employment (EMPL)=Total Number of Employees EMPLA < EMPLB

DIVSALA > DIVIDEND POLICY Dividend to Sales (DIVSAL)=Dividend/Sales DIVSALB Treasury Applications (TA)=Cash and Banks + Dividends + TREASURY TAA > TAB Capital Expenditures

ACTIVITY LEVELS Sales to Total Assets (STA) = Sales/Total Assets STAA > STAB

SHORT TERM (ST) Cash and Banks to ST Debt (CBTSTD) = Cash and CBTSTDA > EQUILIBRIUM Banks/ST Debt CBTSTDB Net Cash Flow to LT Debt (NCFTLTD)=Net Cash Flow/LT NCFTLTDA > LONG TERM (LT) EQUILIBRIUM Debt NCFTLTDB

CAPITAL STRUCTURE Total Debt to Total Assets (TDTA)=Total Debt/Total Assets TDTAA < TDTAB

113 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

Table 4. Summary of results from tests of predictions of the full sample of all Privatized Firms

This table presents empirical results for our full sample of privatized firms. The table presents the results of the Wilcoxon rank sum test (with its z-statistic) - that is employed as a test for significance for change in mean and median values - for each empirical proxy; presenting the number of useable observations, the mean and the median values of the proxy before and after privatization and their change in the proxy’s value after versus before privatization and the test of significance of the mean and median change. The three final columns elements are the percentage of firms whose values of empirical proxy change as predicted and respective test of significance of this change.

Z- Z-statistics Percentage of Z-statistics Z-statistics Mean Mean Mean statistics for firms with for for VARIABLES N Before After Change for difference improved significant significant (Median) (Median) (Median) difference in Medians performance: performance performance in Means (After - Mean(Median) (mean) (median) PROFITABILITY 0.10 0.16 0.06 2.014* 2.216* 67.35% 4.392* 4.124* Return on Sales (ROS) 42 0.21 0.27 0.06 63.21% OPERATING EFFICENCY 0.92 1.75 0.83 3.276* 3.253* 73.81% 4.860* 4.782* Sales Efficiency (SALEFF) 42 0.92 1.78 0.86 71.43% CAPITAL INVESTMENT 0.53 0.61 0.08 1.580 1.652* 54.32% 3.792* 3.712* Real Capital Exp. to Sales (RCESA) 42 0.31 0.40 0.09 52.21% REAL OUTPUT 0.94 2.18 1.24 2.313* 2.532* 69.05% 4.703* 4.782* Real Sales (SAL) 42 0.97 2.42 1.45 71.43% TOTAL EMPLOYMENT 2173.37 1827.74 -345.63 2.394* 2.570* 63.33% 4.421* 4.703* Total Employment (EMPL) 42 2095.43 1837.63 -257.80 69.05% DIVIDEND POLICY 0.01 0.07 0.06 1.553 1.571 64.29% 4.514* 4.197* Dividend to Sales (DIVSAL) 42 0.01 0.04 0.02 54.76% TREASURY 100793.73 105120.93 4327.20 1.585 1.525 54.92% 4.625* 3.924* Treasury Aplications (TA) 42 101398.60 111950.90 10552.30 52.73% ACTIVITY LEVELS 0.52 0.91 0.40 1.712* 1.844* 58.10% 3.516* 3.408* Sales to Total Assets (STA) 42 0.39 0.56 0.17 55.71% SHORT TERM (ST)

EQUILIBRIUM 0.20 0.16 -0.040 1.102 1.019 23.46% 2.824* 2.521* Cash/ Banks to ST Debt (CBTSTD) 42 0.16 0.13 -0.028 19.05% LONG TERM (LT) EQUILIBRIUM

Net Cash Flow to LT Debt 26.95 20.71 -6.245 1.196 1.419 37.14% 4.286* 4.457* 42 (NCFTLTD) 7.78 3.38 -4.399 31.90% CAPITAL STRUCTURE 0.74 0.62 -0.126 1.763* 1.809* 65.24% 3.823* 4.457* Total Debt to Assets (TDTA) 42 0.79 0.63 -0.162 61.90%

* rejection of H0 at five percent level of significance

114 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

Table 5. Comparisons of performance changes following privatization of companies operating in competitive industries versus companies operating in non-competitive industries

This table presents comparisons of performance changes for companies operating in competitive industries and companies operating in non- competitive industries. The table presents the results of the Wilcoxon rank sum test (with its z-statistic) - that is employed as a test for significance for change in mean and median values between before and after privatization - and of the Kruskal-Wallis test between competitive and not competitive firms - in mean terms and in median terms respectively (statistic mentions the 'p' value using the chi-squared approximation) - for each empirical proxy and each subsample of the pair. The table presents the number of useable observations, the mean and the median values of the proxy before and after privatization, their change in the proxy’s value after versus before privatization, the respective test of significance for the mean and median change, the mean rank of the KW test between competitive and non-competitive subsample and the respectively statistic 'p' value for mean and median comparison.

Z- Z- KW Results for differences statistics statistics Percentage Z-statistics Z-statistics between subsamples for mean of firms for for Mean Before Mean After Mean Change for for VARIABLES N with improved significance significance Mean Rank KW test (Median) (Median) (Median) difference difference performance performance performance in Means in C NC 'p' value (After- Medians Mean (Median) (Mean) (Median) PROFITABILITY Return on Sales (ROS) Before) (After- 0.16 0.19 0.03 58.62% Competitive 26 1.678* 1.732* 2.883* 2.911* 0.24 0.29 0.05 53.24% 23.22 18.78 0.051* 0.06 0.11 0.05 54.22% Non competitive 16 1.221 1.412 2.713* 2.891* 0.15 0.26 0.11 51.23% OPERATING EFFICENCY Sales Efficiency (SALEFF) 1.34 1.94 0.60 79.23% Competitive 26 2.012* 2.163* 3.724* 3.724* 0.97 1.91 0.94 69.23% 22.23 20.83 0.54 0.88 1.41 0.53 81.25% Non competitive 16 2.213* 2.302* 3.18* 3.059* 0.79 1.15 0.36 75.00% CAPITAL INVESTMENT Real Cap. Exp. to Sales (RCESA) 0.68 0.71 0.02 58.46% Competitive 26 0.579 0.834 2.733* 2.884* 0.32 0.48 0.16 52.31% 21.77 21.46 0.21 0.49 0.60 0.11 53.75% Noncompetitive 16 0.397 1.322 2.446* 2.451* 0.18 0.34 0.17 47.50% REAL OUTPUT Real Sales (SAL) 1.98 2.67 0.69 61.54% Competitive 26 1.659* 1.343* 3.516* 3.621* 1.11 2.65 1.54 65.38% 21.15 20.81 0.27 0.47 1.98 1.51 81.25% Noncompetitive 16 1.772* 1.771* 3.18* 3.28* 0.91 2.11 1.20 81.25% EMPLOYMENT Total Employment (EMPL) 2709.67 2006.06 -703.61 46.92% Competitive 26 2.549* 2.427* 2.126* 3.654* 2299.50 1983.50 -316.00 59.23% 21.62 19.69 0.045* 2020.24 1759.60 -260.64 43.75% Noncompetitive 16 0.534 0.879 2.256* 2.864* 1917.00 1703.25 -213.75 68.75% DIVIDEND POLICY Dividend to Sales (DIVSAL) 0.04 0.11 0.08 65.38% Competitive 26 1.576 1.113 3.621* 3.408* 0.02 0.09 0.07 57.69% 21.65 21.25 0.91 0.01 0.06 0.05 62.50% Noncompetitive 16 1.256 1.427 2.803* 2.521* 0.01 0.02 0.01 50.00% TREASURY Treasury Aplications (TA) 120228.80 124987.23 4758.43 56.15% Competitive 26 1.523 1.563 3.231* 2.666* 114562.88 119220.32 4657.44 44.62% 22.06 20.44 0.33 87977.65 97232.56 9254.91 56.25% Noncompetitive 16 1.157 1.623 2.456* 2.366* 91221.92 101090.33 9868.41 43.75% ACTIVITY LEVELS Sales to Total Assets (STA) 0.56 1.49 0.92 43.08% Competitive 26 1.683* 1.75* 2.171* 2.201* 0.55 0.56 0.01 41.18% 22.55 19.95 0.032* 0.46 0.73 0.27 62.50% Noncompetitive 16 0.103 0.102 2.773* 2.656* 0.43 0.41 -0.01 56.25% SHORT TERM (ST) EQUILIBRIUM Cash/Banks to ST Debt (CBTSTD) 0.12 0.10 -0.02 21.54% Competitive 26 1.099 1.589 1.652* 1.342 0.13 0.08 -0.04 17.69% 22.58 18.13 0.10 0.27 0.19 -0.09 37.50% Noncompetitive 16 0.150 0.123 2.010* 2.121* 0.19 0.16 -0.04 27.50% LONG TERM (LT) EQUILIBRIUM Net cash flow to LT Debt (NCFTLTD) 16.40 13.66 -2.74 30.49% Competitive 26 0.56 0.357 3.018* 3.096* 4.23 1.12 -3.11 23.85% 20.22 22.31 0.11 33.43 29.77 -3.66 38.75% Noncompetitive 16 1.079 1.523 2.834* 3.129* 9.88 5.34 -4.54 45.00% CAPITAL STRUCTURE Total Debt to Total Assets (TDTA) 0.79 0.69 -0.10 30.00% Competitive 26 0.235 0.245 3.048* 3.378* 0.85 0.71 -0.14 37.69% 21.44 19.78 0.09 0.69 0.60 -0.09 37.50% Noncompetitive 16 0.229 1.02 2.0401* 2.434* 0.66 0.59 -0.07 38.75%

* rejection of H0 at five percent level of significance

115 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

Table 6. Comparisons of performance changes following privatization of companies with less than fifty percent change in Board of Directors versus companies with greater than or equal to fifty percent change in Board of Directors

This table presents comparisons of performance changes for companies with less than fifty percent change in Board of Director and companies with greater than or equal to fifty percent change in Board of Directors. The table presents the results of the Wilcoxon rank sum test (with its z-statistic) - that is employed as a test for significance for change in mean and median values between before and after privatization - and of the Kruskal-Wallis test companies with less than fifty percent change in Board of Director and companies with greater than or equal to fifty percent change in Board of Directors - in mean terms and in median terms respectively (statistic mentions the 'p' value using the chi-squared approximation) - for each empirical proxy and each subsample of the pair. The table presents the number of useable observations, the mean and the median values of the proxy before and after privatization, their change in the proxy’s value after versus before privatization, the respective test of significance for the mean and median change, the mean rank of the KW test between change -50% and no change +50% subsample and the respectively statistic 'p' value for mean and median comparison.

KW Results for diferences Z- between subsamples for Z-statistics Z- statistics statistics Mean Rankmean KW test Percentage Z-statistics for for for Mean Mean of firms for Chang Mean Before difference difference significan e in VARIABLES After Change with improved significance Change in N (Median) in Means in Medians ce Board (Median) (Median) performance performance Board of 'p' (After- (After- performan of Mean (Median) (Median) Director value Before) Before) ce Direct -50% (Mean) or +50% PROFITABILITY Return on Sales (ROS) Change in Board of Director - 0.12 0.09 -0.03 37.14% 14 0.135 0.280 2.5021* 2.0366* 50% 0.26 0.19 -0.07 30.00% 17.25 25.83 0.067* Change in Board of Director 0.08 0.26 0.18 63.57% 28 1.653* 1.678* 3.398* 3.0296* +50% 0.19 0.29 0.10 60.00% OPERATING EFFICENCY Sales Efficiency (SALEFF)

Change in Board of Director - 0.99 0.89 -0.10 35.14% 14 0.754 0.094 2.321* 2.336* 50% 0.97 0.93 -0.04 29.00% 16.29 25.93 0.055* Change in Board of Director 0.87 2.12 1.25 82.14% 28 3.188* 3.412* 4.097* 4.097* +50% 0.85 2.31 1.46 82.14% CAPITAL INVESTMENT Real Capital Exp. to Sales (RCESA)

Change in Board of Director - 0.33 0.39 0.06 42.86% 14 0.377 1.297 2.201* 2.023* 50% 0.24 0.24 0.00 35.71% 21.32 21.86 0.894 Change in Board of Director 0.63 0.72 0.09 59.29% 28 0.683 1.202 2.934* 3.059* +50% 0.38 0.49 0.11 52.86% REAL OUTPUT Real Sales (SAL)

Change in Board of Director - 0.85 1.55 0.70 51.23% 14 0.157 0.126 2.436* 2.456* 50% 0.86 1.90 1.05 52.77% 18.71 23.37 0.073* Change in Board of Director 0.98 2.99 2.01 78.57% 28 2.232* 2.978* 3.917* 4.007* +50% 0.99 3.17 2.18 82.14% EMPLOYMENT Total Employment (EMPL)

Change in Board of Director - 2681.18 1788.33 -892.85 38.57% 14 0.602 0.580 1.786* 2.666* 50% 2229.00 1953.25 -275.75 54.29% 21.86 22.66 0.061* Change in Board of Director 1969.46 1917.45 -52.01 35.71% 28 2.229* 2.569* 2.753* 3.760* +50% 1924.00 1666.50 -257.50 69.43% DIVIDEND POLICY Dividend to Sales (DIVSAL)

Change in Board of Director - 0.01 0.10 0.09 69.43% 14 1.238* 1.117 2.812* 2.546* 50% 0.01 0.01 0.00 64.29% 21.86 21.79 0.147 Change in Board of Director 0.02 0.03 0.02 60.71% 28 1.397 1.523 3.771* 3.326* +50% 0.01 0.06 0.05 51.25% TREASURY Treasury Aplications (TA) Change in Board of Director - 94900.82 105987.88 11087.06 30.00% 14 1.148 1.116 2.786* 2.113* 50% 89828.37 101222.29 11393.92 25.71% 21.71 22.07 0.557 Change in Board of Director 117292.22 126989.99 9697.77 40.00% 28 1.125 1.443 3.396* 2.894* +50% 116722.80 132454.88 15732.08 29.29% ACTIVITY LEVELS Sales to Total Assets (STA) Change in Board of Director - 0.37 0.72 0.35 45.71% 14 0.491 0.874 2.143* 2.113* 50% 0.35 0.39 0.04 45.71% 21.46 20.57 0.142 Change in Board of Director 0.58 1.51 0.93 59.29% 28 1.002 1.071 2.764* 2.773* +50% 0.55 0.62 0.06 55.71% SHORT TERM (ST) EQUILIBRIUM Cash/Banks to ST Debt (CBTSTD) Change in Board of Director - 0.19 0.11 -0.07 35.71% 14 1.363 1.099 2.123* 2.143* 50% 0.12 0.09 -0.04 35.71% 20.04 22.43 0.078* Change in Board of Director 0.21 0.18 -0.03 24.29% 28 1.834* 2.892* 1.796* 1.574 +50% 0.20 0.18 -0.02 20.71% LONG TERM (LT) EQUILIBRIUM Net cash flow to LT Debt (NCFTLTD) Change in Board of Director - 19.64 11.21 -8.43 34.29% 14 0.865 0.295 2.556* 2.471* 50% 1.60 1.15 -0.45 27.14% 21.79 20.93 0.831 Change in Board of Director 37.11 30.46 -6.65 43.57% 28 0.766 1.595 3.338* 3.684* +50% 12.17 4.18 -7.99 44.29% CAPITAL STRUCTURE Total Debt to Total Assets (TDTA) Change in Board of Director - 0.63 0.57 -0.06 32.86% 14 0.555 0.812 2.191* 2.793* 50% 0.68 0.57 -0.11 41.43% 21.50 21.05 0.455 Change in Board of Director 0.85 0.74 -0.11 56.43% 28 0.054 0.866 3.150* 3.496* +50% 0.92 0.69 -0.23 67.14%

* rejection of H0 at five percent level of significance

116 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

Table 7. Comparisons of performance changes following privatization of firms that belong to the financial sector versus firms that do not belong to the financial sector

This table presents comparisons of performance changes for firms that belong in the financial sector versus firms that do not belong in the financial sector. The table presents the results of the Wilcoxon rank sum test (with its z-statistic) - that is employed as a test for significance for change in mean and median values between before and after privatization - and of the Kruskal-Wallis test firms from financial sector versus firms that aren’t from the financial sector - in mean terms and in median terms respectively (statistic mentions the 'p' value using the chi-squared approximation) - for each empirical proxy and each subsample of the pair. The table presents the number of useable observations. the mean and the median values of the proxy before and after privatization, their change in the proxy’s value after versus before privatization, the respective test of significance for the mean and median change, the mean rank of the KW test between Financial Firms and Not Financial Firms subsample and the respectively statistic 'p' value for mean and median comparison.

KW Results for differences between Z-statistics Z-statistics Percentage of Z-statistics subsamples for mean for for Z-statistics for Mean Mean Mean firms for difference difference significance Mean Rank KW test VARIABLES N Before After Change with improved significance in Means in Medians performance (Median) (Median) (Median) performance performance (After- (After- (Median) Mean (Median) (Mean) FS NoFS 'p' value Before) Before)

PROFITABILITY Return on Sales (ROS) 0.12 0.15 0.03 60.00% Financial 15 0.346 0.910 2.666* 2.023* 0.23 0.25 0.02 43.33% 18.47 23.92 0.068* No 0.08 0.23 0.15 50.00% 26 1.792* 1.845* 3.180* 3.408* Financial 0.17 0.34 0.17 57.69% OPERATING EFFICENCY Sales Efficiency (SALEFF) 1.04 1.42 0.38 66.67% Financial 15 1.449 1.534 2.803* 2.803* 0.97 1.49 0.52 66.67% 19.13 22.42 0.052* No 0.72 1.89 1.17 76.92% 26 2.794* 2.718* 3.920* 3.823* Financial 0.81 1.99 1.18 73.08% CAPITAL INVESTMENT Real Capital Exp. to Sales (RCESA) 0.58 0.59 0.01 53.33% Financial 15 0.126 0.421 2.023* 2.201* 0.39 0.41 0.02 50.00% 18.80 21.52 0.106 No 0.41 0.88 0.47 56.15% 26 1.553 2.191* 3.059* 2.934* Financial 0.19 0.69 0.50 52.31% REAL OUTPUT Real Sales (SAL) 0.78 1.59 0.81 66.67% Financial 15 1.477 1.088 2.803* 2.934* 0.81 1.78 0.97 73.33% 18.33 23.80 0.09 No 1.19 3.55 2.36 69.23% 26 1.332 1.834* 3.724* 3.724* Financial 1.29 3.78 2.49 69.23% EMPLOYMENT Total Employment (EMPL) 2795.90 2277.64 -518.26 56.67% Financial 15 1.590 1.477 1.826* 2.934* 2372.00 2022.50 -349.50 73.33% 17.40 24.36 0.04* No 1517.05 1488.16 -28.89 48.46% 26 1.613* 1.816* 2.803* 3.621* Financial 1539.50 1464.50 -75.00 65.38% DIVIDEND POLICY Dividend to Sales (DIVSAL) 0.01 0.12 0.10 80.00% Financial 15 1.675* 1.943* 3.059* 3.180* 0.00 0.01 0.01 86.67% 21.47 21.12 0.295 No 0.02 0.05 0.03 53.85% 26 1.327 1.566 3.296* 2.666* Financial 0.00 0.00 0.00 34.62% TREASURY Treasury Aplications (TA) 119878.00 123909.00 4031.00 53.33% Financial 15 1.625 1.517 2.521* 2.201* 118777.00 120982.00 2205.00 40.00% 20.27 20.64 0.922 No 97222.00 98787.00 1565.00 50.00% 26 1.200 1.629 3.186* 2.803* Financial 89767.00 93450.00 3683.00 38.46% ACTIVITY LEVELS Sales to Total Assets (STA) 0.47 0.87 0.40 66.67% Financial 15 1.204 1.188 2.323* 2.989* 0.47 0.59 0.12 66.67% 17.47 25.92 0.03* No 0.59 1.94 1.35 57.69% 26 3.048* 3.185* 3.408* 3.296* Financial 0.33 1.66 1.33 53.85% SHORT TERM (ST) EQUILIBRIUM Cash/Banks to ST Debt (CBTSTD) 0.36 0.19 -0.17 26.67% Financial 15 1.989* 1.712* 1.826* 1.604 0.25 0.19 -0.06 20.00% 25.00 18.60 0.06* No 0.11 0.09 -0.02 19.23% 26 1.248 1.451 2.023* 2.023* Financial 0.12 0.09 -0.03 19.23% LONG TERM (LT) EQUILIBRIUM Net cash flow to LT Debt (NCFTLTD) 28.34 24.55 -3.79 43.33% Financial 15 0.802 0.708 2.521* 2.666* 8.99 7.44 -1.55 40.00% 21.07 23.16 0.63 No 22.77 21.23 -1.54 47.69% 26 0.486 1.157 3.408* 3.516* Financial 4.84 4.11 -0.73 41.54% CAPITAL STRUCTURE Total Debt to Total Assets (TDTA) 0.90 0.89 -0.01 53.33% Financial 15 0.057 0.628 2.521* 2.366* 0.86 0.90 0.03 56.67% 20.60 24.44 0.02* No 0.74 0.55 -0.19 52.31% 26 1.677* 1.683* 2.934* 3.724* Financial 0.60 0.46 -0.14 69.23%

* rejection of H0 at five percent level of significance

117 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

Table 8. Comparisons of Performance Changes Following Privatization of Companies with Greater or Equal to 50% of Capital in National Hands versus Companies with Less than 50% of Capital in National Hand

This table presents comparisons of performance changes for companies with greater or equal to fifty percent of capital in national hands (National Allocation) versus companies with less than fifty percent of capital in national hands (Foreign Allocation). The table presents the results of the Wilcoxon rank sum test (with its z-statistic) - that is employed as a test for significance for change in mean and median values between before and after privatization - and of the Kruskal- Wallis test companies with National Allocation and companies with Foreign Allocation - in mean terms and in median terms respectively (statistic mentions the 'p' value using the chi-squared approximation) - for each empirical proxy and each subsample of the pair. The table presents the number of useable observations, the mean and the median values of the proxy before and after privatization, their change in the proxy’s value after versus before privatization, the respective test of significance for the mean and median change, the mean rank of the KW test between National and Foreign subsample and the respectively statistic 'p' value for mean and median comparison.

Z-statistics Z- statistics KW Results for differences Percentage Z-statistics Z- statistics for for between subsamples for mean of firms for for Mean Before Mean After Mean Change difference difference Mean Rank KW test N with improved significance significance VARIABLES (Median) (Median) (Median) in Means in Medians performance performance performance (After- (After- Mean (Median) (Mean) (Median) NaC FC 'p' value Before) Before) PROFITABILITY Return on Sales (ROS) 0.12 0.16 0.04 53.66% National Allocation 39 0.180 0.141 4.107* 3.823* 0.23 0.26 0.03 48.72% 18.85 23.88 0.051* 0.11 0.39 0.28 63.33% Foreign Allocation 3 1.664* 1.669* 2.209* 2.342* 0.23 0.49 0.26 66.67% OPERATING EFFICENCY Sales Efficiency (SALEFF) 0.93 1.71 0.78 70.73% National Allocation 39 2.107* 2.770* 4.683* 4.583* 0.91 1.72 0.81 71.79% 21.38 23.00 0.826 0.69 1.97 1.28 76.67% Foreign Allocation 3 1.729* 1.839* 1.722* 1.892* 0.71 1.97 1.26 76.67% CAPITAL INVESTMENT Real Capital Exp. to Sales (RCESA) 0.55 0.57 0.02 51.46% National Allocation 39 0.184 1.374 3.581* 3.561* 0.36 0.37 0.01 53.59% 20.36 21.66 0.788 0.26 0.79 0.53 73.88% Foreign Allocation 3 0.594 1.593 1.098 1.213 0.28 0.89 0.61 65.77% REAL OUTPUT Real Sales (SAL) 0.95 2.16 1.21 65.85% National Allocation 39 0.397 0.475 4.041* 4.623* 0.99 2.39 1.40 71.79% 17.87 24.76 0.075* 0.77 2.89 2.12 76.67% Foreign Allocation 3 2.402* 2.680* 1.292 1.342 0.69 2.91 2.22 76.67% EMPLOYMENT Total Employment (EMPL) 2241.79 1821.19 -420.60 54.15% National Allocation 39 2.193* 2.255* 3.316* 4.397* 2365.00 1876.00 -489.00 66.67% 21.77 18.00 0.06* 1283.89 1127.88 -156.01 43.55% Foreign Allocation 3 1.584 1.594 1.476 1.604 1519.00 1361.00 -158.00 53.33% DIVIDEND POLICY Dividend to Sales (DIVSAL) 0.02 0.07 0.06 60.98% National Allocation 39 0.594 0.468 4.412* 4.117* 0.01 0.06 0.05 56.41% 21.41 22.67 0.864 0.01 0.03 0.02 66.67% Foreign Allocation 3 0.497 0.447 1.332 1.456 0.01 0.02 0.01 43.33% TREASURY Treasury Aplications (TA) 98222.00 103332.00 5110.00 36.34% National Allocation 39 1.579 1.578 3.823* 3.408* 102887.00 109878.00 6991.00 38.46% 21.85 21.00 0.510 102110.62 109656.00 7545.38 36.67% Foreign Allocation 3 1.119 1.102 1.342 1.432 109888.00 117989.00 8101.00 33.33% ACTIVITY LEVELS Sales to Total Assets (STA) 0.51 0.79 0.28 51.71% National Allocation 39 1.594 1.119 3.150* 3.450* 0.42 0.45 0.03 53.33% 20.62 23.98 0.052* 0.67 0.99 0.32 78.55% Foreign Allocation 3 1.651* 1.695* 1.694* 1.732* 0.71 0.79 0.09 66.67% SHORT TERM (ST) EQUILIBRIUM Cash/Banks to ST Debt (CBTSTD) 0.19 0.15 -0.03 19.51% National Allocation 39 1.495 1.236 2.491* 2.336* 0.14 0.12 -0.02 17.95% 22.15 22.00 0.592 0.23 0.20 -0.03 33.33% Foreign Allocation 3 1.027 1.201 1.712* 1.832* 0.18 0.14 -0.04 33.33% LONG TERM (LT) EQUILIBRIUM Net cash flow to LT Debt (NCFTLTD) 24.90 18.96 -5.94 33.66% National Allocation 39 1.660* 1.753* 3.907* 4.326* 6.89 2.29 -4.60 41.54% 25.72 17.67 0.036* 28.69 27.46 -1.24 36.67% Foreign Allocation 3 0.535 0.535 1.692* 1.794* 9.55 7.34 -2.21 36.67% CAPITAL STRUCTURE Total Debt to Total Assets (TDTA) 0.79 0.63 -0.16 53.90% National Allocation 39 0.218 0.838 3.724* 4.197* 0.76 0.62 -0.14 58.97% 22.38 21.88 0.092 0.80 0.74 -0.05 53.33% Foreign Allocation 3 0.535 1.604 1.701* 1.694* 0.85 0.79 -0.06 63.90%

* rejection of H0 at five percent level of significance

118 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

Table 9. Comparisons of performance changes following privatization of firms that have concentrated structure versus firms that have a more flexible structure

This table presents comparisons of performance changes for firms that have concentrated structure versus firms that have a more flexible structure (No Concentrated Structure). The table presents the results of the Wilcoxon rank sum test (with its z-statistic) - that is employed as a test for significance for change in mean and median values between before and after privatization - and of the Kruskal-Wallis test firms that have concentrated structure versus firms that have a more flexible structure - in mean terms and in median terms respectively (statistic mentions the 'p' value using the chi-squared approximation) - for each empirical proxy and each subsample of the pair. The table presents the number of useable observations, the mean and the median values of the proxy before and after privatization, their change in the proxy’s value after versus before privatization, the respective test of significance for the mean and median change, the mean rank of the KW test between Concentrated Structure and No Concentrated Structure subsample and the respectively statistic 'p' value for mean and median comparison.

Z-statistics Z - statistics KW Results for diferences Percentage Z-statistics Z-statistics for for between subsamples for mean of firms for for Mean Before Mean After Mean Change difference difference VARIABLES N with improved significance significance Mean Rank KW test (Median) (Median) (Median) in Means in Medians performance performance performance (After- (After- Mean (Median) (Mean) (Median) CS NoCS 'p' value Before) Before) PROFITABILITY Return on Sales (ROS) 0.08 0.20 0.12 60.00% Concentrated Structure 22 1.661* 1.732* 2.934* 3.059* 0.18 0.31 0.13 64.55% 25.34 18.69 0.025* 0.14 0.15 0.01 55.00% No Concentrated Structure 20 0.785 1.047 2.934* 2.666* 0.25 0.25 0.00 45.00% OPERATING EFFICENCY Sales Efficiency (SALEFF) 0.94 1.81 0.86 72.73% Concentrated Structure 22 2.549* 2.833* 3.621* 3.621* 0.93 1.87 0.93 77.27% 21.48 22.60 0.416 0.89 1.55 0.66 70.00% No Concentrated Structure 20 2.115* 1.842* 3.296* 3.296* 0.89 1.49 0.60 70.00% CAPITAL INVESTMENT Real Capital Exp. to Sales (RCESA) 0.58 0.75 0.17 62.73% Concentrated Structure 22 0.896 1.308 2.023* 2.666* 0.32 0.47 0.16 60.91% 21.80 22.31 0.239 0.48 0.57 0.10 50.00% No Concentrated Structure 20 0.817 1.112 2.803* 2.521* 0.27 0.35 0.08 50.00% REAL OUTPUT Real Sales (SAL) 1.01 2.47 1.46 72.73% Concentrated Structure 22 2.173* 2.354* 3.516* 3.621* 1.22 2.72 1.50 77.27% 23.93 19.12 0.050* 0.86 1.78 0.91 65.00% No Concentrated Structure 20 1.043 1.269 3.180* 3.296* 0.93 2.03 1.10 70.00% EMPLOYMENT Total Employment (EMPL) 2476.00 1789.00 -687.00 31.82% Concentrated Structure 22 2.711* 2.419* 2.366* 3.408* 2232.00 1698.00 -534.00 68.18% 24.61 19.50 0.022* 1898.00 1867.00 -31.00 35.00% No Concentrated Structure 20 1.060 1.408 2.366* 3.408* 1886.00 1872.00 -14.00 75.00% DIVIDEND POLICY Dividend to Sales (DIVSAL) 0.01 0.04 0.03 50.00% Concentrated Structure 22 1.016 1.544 2.934* 2.666* 0.01 0.02 0.01 40.91% 21.61 19.60 0.19 0.02 0.10 0.09 70.00% No Concentrated Structure 20 1.482 1.738 3.516* 3.296* 0.00 0.07 0.07 60.00% TREASURY Treasury Aplications (TA) 111232.00 119789.00 8557.00 36.36% Concentrated Structure 22 1.464 1.485 2.521* 2.023* 111282.00 127899.00 16617.00 22.73% 21.09 22.95 0.627 96787.00 100222.00 3435.00 35.00% No Concentrated Structure 20 1.269 1.435 3.180* 2.934* 92787.00 102009.00 9222.00 35.00% ACTIVITY LEVELS Sales to Total Assets (STA) 0.48 0.99 0.51 36.36% Concentrated Structure 22 1.266 1.348 2.521* 2.360* 0.41 0.57 0.16 31.82% 22.66 20.26 0.375 0.57 0.80 0.23 40.00% No Concentrated Structure 20 0.691 0.840 2.521* 2.521* 0.67 0.49 -0.18 40.00% SHORT TERM (ST) EQUILIBRIUM Cash/Banks to ST Debt (CBTSTD) 0.28 0.23 -0.05 13.64% Concentrated Structure 22 1.739* 1.895* 1.604 1.604 0.19 0.18 -0.01 13.64% 21.02 23.02 0.601 0.17 0.11 -0.06 25.00% No Concentrated Structure 20 2.053* 2.636* 2.023* 2.023* 0.10 0.09 -0.01 25.00% LONG TERM (LT) EQUILIBRIUM Net cash flow to LT Debt (NCFTLTD) 22.34 20.01 -2.33 34.55% Concentrated Structure 22 0.776 1.460 3.059* 3.296* 6.92 4.28 -2.64 33.64% 20.75 23.31 0.504 28.88 22.76 -6.12 35.00% No Concentrated Structure 20 1.102 1.326 3.180* 3.180* 7.88 2.33 -5.55 35.00% CAPITAL STRUCTURE Total Debt to Total Assets (TDTA) 0.78 0.49 -0.29 60.91% Concentrated Structure 22 1.780* 1.687* 2.666* 3.296* 0.83 0.61 -0.22 63.64% 24.89 18.07 0.053* 0.69 0.65 -0.04 60.00% No Concentrated Structure 20 1.037 1.486 2.803* 3.180* 0.75 0.69 -0.06 65.00%

* rejection of H0 at five percent level of significance

119 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

Table 10. Comparisons of performance changes following privatization of firms that were privatized by Share Issue Privatization (SIP) versus firms that were privatized by Direct Sale (DS)

This table presents comparisons of performance changes for firms that were privatized by SIP versus firms that were privatized by Direct Sale. The table presents the results of the Wilcoxon rank sum test (with its z-statistic) - that is employed as a test for significance for change in mean and median values between before and after privatization - and of the Kruskal-Wallis test firms that were privatized by IPO versus firms that were privatized by Direct Sale - in mean terms and in median terms respectively (statistic mentions the 'p' value using the chi- squared approximation) - for each empirical proxy and each subsample of the pair. The table presents the number of useable observations, the mean and the median values of the proxy before and after privatization, their change in the proxy’s value after versus before privatization, the respective test of significance for the mean and median change, the mean rank of the KW test between IPO and Direct Sale subsample and the respectively statistic 'p' value for mean and median comparison.

Z-statistics Z- statistics KW Results for diferences Percentage Z-statistics Z- statistics for for between subsamples for mean of firms for for Mean Before Mean After Mean Change difference difference Mean Rank VARIABLES with improved significance significance KW test N (Median) (Median) (Median) in Means in Medians performance performance performance (After- (After- Mean (Median) (Mean) (Median) SPI DS 'p' value Before) Before) PROFITABILITY Return on Sales (ROS) 0.08 0.29 0.21 60.00% Privatization by IPO 19 1.908* 1.739* 2.803* 2.366* 0.17 0.38 0.21 55.00% 24.30 19.61 0.064* 0.13 0.14 0.01 60.00% Privatization by Direct Sale 18 0.175 1.218 2.666* 2.934* 0.22 0.25 0.03 61.11% OPERATING EFFICENCY Sales Efficiency (SALEFF) 0.95 1.99 1.04 75.00% Privatization by IPO 19 2.697* 2.677* 3.408* 3.408* 0.97 2.17 1.20 75.00% 22.90 15.72 0.047* 0.71 1.69 0.98 61.11% Privatization by Direct Sale 18 1.198 1.207 2.934* 2.934* 0.77 1.65 0.88 61.11% CAPITAL INVESTMENT Real Capital Exp. to Sales (RCESA) 0.44 0.52 0.08 50.00% Privatization by IPO 19 1.752* 1.650* 2.201* 2.366* 0.24 0.33 0.09 55.00% 21.05 19.78 0.045* 0.57 0.78 0.21 58.89% Privatization by Direct Sale 18 0.803 0.645 2.366* 2.666* 0.37 0.69 0.32 50.00% REAL OUTPUT Real Sales (SAL) 0.84 3.77 2.93 70.00% Privatization by IPO 19 1.871* 2.614* 3.296* 3.516* 0.88 3.75 2.87 80.00% 25.25 13.11 0.001* 1.10 1.25 0.15 55.56% Privatization by Direct Sale 18 0.370 0.131 2.803* 2.803* 0.99 1.12 0.13 55.56% EMPLOYMENT Total Employment (EMPL) 2570.25 1958.13 -612.12 65.00% Privatization by IPO 19 2.389* 2.352* 2.023* 3.516* 2572.00 1423.50 -1148.50 70.00% 25.60 12.72 0.03* 1960.68 1766.33 -194.34 53.33% Privatization by Direct Sale 18 1.607 1.459 2.201* 2.934* 1529.00 1383.50 -145.50 51.11% DIVIDEND POLICY Dividend to Sales (DIVSAL) 0.02 0.12 0.10 70.00% Privatization by IPO 19 1.121 1.208 3.516* 3.408* 0.01 0.07 0.06 75.00% 22.90 15.72 0.046* 0.01 0.03 0.02 48.89% Privatization by Direct Sale 18 1.367 1.485 2.366* 2.023* 0.00 0.03 0.03 47.78% TREASURY Treasury Aplications (TA) 94920.00 95677.00 757.00 40.00% Privatization by IPO 19 1.583 1.465 3.059* 2.666* 99232.00 102989.00 3757.00 35.00% 22.95 20.11 0.748 104009.00 117826.00 13817.00 33.33% Privatization by Direct Sale 18 0.299 0.784 2.201* 2.023* 111272.00 121345.00 10073.00 27.78% ACTIVITY LEVELS Sales to Total Assets (STA) 0.35 1.33 0.98 15.00% Privatization by IPO 19 2.148* 2.614* 1.604 1.342 0.33 0.43 0.10 10.00% 23.75 15.67 0.028* 0.65 0.71 0.06 55.56% Privatization by Direct Sale 18 0.142 0.047 2.803* 2.803* 0.57 0.59 0.02 55.56% SHORT TERM (ST) EQUILIBRIUM Cash/Banks to ST Debt (CBTSTD) 0.27 0.14 -0.13 20.00% Privatization by IPO 19 2.334* 3.181* 1.826* 1.604 0.19 0.08 -0.10 15.00% 23.30 15.28 0.026* 0.20 0.18 -0.02 16.67% Privatization by Direct Sale 18 1.217 1.179 1.604 1.604 0.09 0.08 -0.01 16.67% LONG TERM (LT) EQUILIBRIUM Net cash flow to LT Debt (NCFTLTD) 41.99 25.99 -16.00 45.00% Privatization by IPO 19 1.647* 1.966* 2.934* 3.180* 3.88 3.44 -0.44 45.00% 22.80 20.28 0.682 16.18 15.12 -1.06 40.00% Privatization by Direct Sale 18 0.491 0.155 2.666* 2.803* 8.55 3.32 -5.23 45.56% CAPITAL STRUCTURE Total Debt to Total Assets (TDTA) 0.72 0.51 -0.21 65.00% Privatization by IPO 19 1.492* 1.574* 2.666* 2.934* 0.52 0.49 -0.02 65.00% 23.55 19.77 0.018* 0.82 0.79 -0.02 64.44% Privatization by Direct Sale 18 0.430 0.081 2.521* 3.180* 0.84 0.81 -0.03 62.22%

* rejection of H0 at five percent level of significance

120 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

Table 11. Comparisons of performance changes following privatization of firms that are listed versus firms that are not listed in a stock exchange

This table presents comparisons of performance changes for firms that have their stock officially quoted versus firms that do not have their stock officially quoted. The table presents the results of the Wilcoxon rank sum test (with its z-statistic) - that is employed as a test for significance for change in mean and median values between before and after privatization - and of the Kruskal-Wallis test firms that have their stock officially quoted versus firms that do not haven’t - in mean terms and in median terms respectively (statistic mentions the 'p' value using the chi-squared approximation) - for each empirical proxy and each subsample of the pair. The table presents the number of useable observations, the mean and the median values of the proxy before and after privatization, their change in the proxy’s value after versus before privatization, the respective test of significance for the mean and median change, the mean rank of the KW test between Firms with stock officially quoted and Firms with stock not officially quoted subsample and the respectively statistic 'p' value for mean and median comparison.

Z- KW Results for differences Z- statistics between subsamples for mean statistics Percentage Z-statistics Z- statistics for for of firms for for Mean Rank KW test Mean Before Mean After Mean Change difference VARIABLES N difference with improved significance significance (Median) (Median) (Median) in in Means performance performance performance Medians (After- Mean (Median) (Mean) (Median) Lx NoLx 'p' value (After- Before) Before) PROFITABILITY Return on Sales (ROS) 0.08 0.29 0.21 63.85% Firms with stock officially quoted 26 1.944* 2.259* 3.296* 3.059* 0.15 0.34 0.19 58.00% 22.96 20.82 0.076* 0.14 0.15 0.01 60.00% Firms with stock not officially quoted 16 0.095 0.854 2.521* 2.666* 0.20 0.23 0.03 62.94% OPERATING EFFICENCY Sales Efficiency (SALEFF) 0.90 1.98 1.08 73.08% Firms with stock officially quoted 26 2.987* 2.785* 3.823* 3.724* 0.86 1.89 1.03 72.00% 24.28 17.41 0.075* 0.94 1.57 0.63 68.75% Firms with stock not officially quoted 16 1.587 1.810* 2.934* 3.059* 0.93 1.49 0.56 70.59% CAPITAL INVESTMENT Real Capital Exp. to Sales (RCESA) 0.34 0.94 0.60 52.31% Firms with stock officially quoted 26 1.843* 1.651* 2.934* 2.666* 0.21 0.62 0.41 56.00% 24.32 17.35 0.071* 0.60 0.65 0.05 57.50% Firms with stock not officially quoted 16 0.908 1.278 2.201* 2.521* 0.39 0.40 0.01 57.06% REAL OUTPUT Real Sales (SAL) 0.87 2.56 1.69 65.38% Firms with stock officially quoted 26 1.729* 2.301* 3.621* 3.724* 0.86 2.78 1.92 72.00% 26.56 16.06 0.01* 1.04 1.78 0.74 68.75% Firms with stock not officially quoted 16 1.634 1.373 2.934* 3.059* 1.09 1.19 0.10 70.59% EMPLOYMENT Total Employment (EMPL) 3076.59 2626.11 -450.48 50.77% Firms with stock officially quoted 26 1.682* 1.978* 2.521* 3.823* 2670.00 2441.00 -229.00 66.00% 26.72 16.35 0.05* 1845.10 1653.67 -191.44 47.50% Firms with stock not officially quoted 16 1.501 1.160 2.201* 2.803* 1423.00 1304.00 -119.00 58.82% DIVIDEND POLICY Dividend to Sales (DIVSAL) 0.02 0.10 0.08 76.92% Firms with stock officially quoted 26 1.249 1.469 3.920* 3.724* 0.00 0.02 0.02 72.00% 25.20 17.06 0.017* 0.01 0.03 0.02 37.50% Firms with stock not officially quoted 16 1.542 1.363 2.201* 2.023* 0.00 0.00 0.00 29.41% TREASURY Treasury Aplications (TA) 101909.00 107898.00 5989.00 57.69% Firms with stock officially quoted 26 1.857* 1.772* 3.408* 2.934* 105640.00 112323.00 6683.00 44.00% 20.74 22.62 0.062 89878.00 103454.00 13576.00 57.50% Firms with stock not officially quoted 16 1.610 1.562 2.201* 2.023* 91234.00 108704.00 17470.00 59.41% ACTIVITY LEVELS Sales to Total Assets (STA) 0.40 1.14 0.75 66.92% Firms with stock officially quoted 26 1.872* 2.087* 2.366* 2.201* 0.32 0.59 0.27 64.00% 18.08 24.53 0.028* 0.69 0.77 0.08 50.00% Firms with stock not officially quoted 16 0.233 0.078 2.521* 2.666* 0.57 0.68 0.11 52.94% SHORT TERM (ST) EQUILIBRIUM Cash/Banks to ST Debt (CBTSTD) 0.25 0.19 -0.06 53.08% Firms with stock officially quoted 26 2.258* 3.119* 2.201* 2.023* 0.19 0.18 -0.01 50.00% 24.24 19.47 0.079 0.13 0.12 -0.01 48.75% Firms with stock not officially quoted 16 1.023 0.980 1.604 1.604 0.12 0.12 0.00 47.65% LONG TERM (LT) EQUILIBRIUM Net cash flow to LT Debt (NCFTLTD) 31.87 29.55 -2.32 53.85% Firms with stock officially quoted 26 1.043 1.217 3.296* 3.408* 7.98 4.99 -2.99 60.00% 20.72 22.65 0.617 18.79 14.55 -4.24 56.25% Firms with stock not officially quoted 16 0.440 1.224 2.666* 2.934* 6.77 2.66 -4.11 64.71% CAPITAL STRUCTURE Total Debt to Total Assets (TDTA) 0.73 0.58 -0.15 56.15% Firms with stock officially quoted 26 0.027 0.043 3.059* 3.296* 0.60 0.49 -0.11 56.00% 24.28 20.41 0.075 0.76 0.71 -0.05 57.50% Firms with stock not officially quoted 16 0.517 1.500 2.201* 3.059* 0.77 0.72 -0.05 60.59%

* rejection of H0 at five percent level of significance

121 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

Table 12. Comparisons of performance changes following privatization for firms that were privatized before or in 1990 versus firms that were privatized after 1990

This table presents comparisons of performance changes for firms that were privatized before 1990 versus firms that were privatized after 1990. The table presents the results of the Wilcoxon rank sum test (with its z-statistic) - that is employed as a test for significance for change in mean and median values between before and after privatization - and of the Kruskal-Wallis test firms that were privatized before 1990 versus firms that were privatized after 1990 - in mean terms and in median terms respectively (statistic mentions the 'p' value using the chi- squared approximation) - for each empirical proxy and each subsample of the pair. The table presents the number of useable observations, the mean and the median values of the proxy before and after privatization, their change in the proxy’s value after versus before privatization, the respective test of significance for the mean and median change, the mean rank of the KW test between Privatization before 1990 and Privatization after 1990 subsample and the respectively statistic 'p' value for mean and median comparison.

Z-statistics KW Results for differences Z-statistics Percentage Z-statistics Z- statistics for between subsamples for mean for of firms for for Mean Before Mean After Mean Change difference VARIABLES N difference with improved significance significance Mean Rank KW test (Median) (Median) (Median) in Means in Medians performance performance performance (After- Before After (After-Before) Mean (Median) (Mean) (Median) 'p' value Before) 1997 1997 PROFITABILITY Return on Sales (ROS) 0.13 0.13 0.00 70.00% Privatization in or Before 1990 12 0.314 0.368 2.201* 2.366* 0.23 0.26 0.03 68.33% 16.25 22.97 0.042* 0.04 0.29 0.25 56.67% Privatization After 1990 30 1.642* 1.923* 3.724* 3.180* 0.12 0.38 0.26 53.33% OPERATING EFFICENCY Sales Efficiency (SALEFF) 0.77 1.89 1.12 81.67% Privatization in or Before 1990 12 2.903* 2.746* 2.934* 2.666* 0.79 1.83 1.04 75.00% 22.25 20.48 0.667 1.09 1.13 0.04 66.67% Privatization After 1990 30 1.872* 2.139* 4.015* 3.920* 1.13 1.15 0.02 66.67% CAPITAL INVESTMENT Real Capital Exp. to Sales (RCESA) 0.32 0.79 0.47 68.33% Privatization in or Before 1990 12 2.864* 2.944* 2.898* 3.121* 0.21 0.59 0.38 68.33% 25.42 19.48 0.028* 0.27 0.28 0.01 53.33% Privatization After 1990 30 0.962 0.228 2.516* 2.588* 0.25 0.27 0.02 53.33% REAL OUTPUT Real Sales (SAL) 0.74 2.89 2.15 83.33% Privatization in or Before 1990 12 1.561* 1.883* 2.803* 2.666* 0.15 3.13 2.98 75.00% 22.42 20.41 0.626 1.19 1.88 0.69 63.33% Privatization After 1990 30 1.800* 2.117* 3.902* 3.920* 0.53 1.17 0.64 66.67% EMPLOYMENT Total Employment (EMPL) 2236.00 1811.00 -425.00 68.33% Privatization in or Before 1990 12 2.981* 3.059* 2.934* 2144.00 1899.00 -245.00 71.67% 23.33 20.03 0.422 2098.00 1845.00 -253.00 53.33% Privatization After 1990 30 1.018 1.097 3.297* 3.621* 1992.00 1808.00 -184.00 56.67% DIVIDEND POLICY Dividend to Sales (DIVSAL) 0.01 0.11 0.10 65.00% Privatization in or Before 1990 12 1.160 1.033 2.666* 2.366* 0.00 0.01 0.01 58.33% 20.92 21.62 0.751 0.02 0.05 0.04 50.00% Privatization After 1990 30 1.496 1.199 3.724* 3.408* 0.00 0.01 0.01 50.00% TREASURY Treasury Aplications (TA) 102398.00 109980.00 7582.00 40.00% Privatization in or Before 1990 12 1.628 1.504 2.201* 1.604 108090.00 116778.00 8688.00 25.00% 21.75 20.69 0.796 98767.00 103442.00 4675.00 40.00% Privatization After 1990 30 1.410 1.434 3.408* 3.180* 99454.00 104767.00 5313.00 43.33% ACTIVITY LEVELS Sales to Total Assets (STA) 0.55 0.89 0.34 61.67% Privatization in or Before 1990 12 0.118 0.746 2.023* 1.826* 0.44 0.55 0.11 63.33% 23.00 21.34 0.169 0.48 0.98 0.50 56.67% Privatization After 1990 30 1.352 0.864 3.059* 2.934* 0.46 0.59 0.13 56.67% SHORT TERM (ST) EQUILIBRIUM Cash/Banks to ST Debt (CBTSTD) 0.23 0.18 -0.05 16.67% Privatization in or Before 1990 12 1.487 1.600 1.342 1.342 0.18 0.15 -0.03 16.67% 22.58 20.94 0.585 0.17 0.13 -0.04 23.33% Privatization After 1990 30 1.111 1.557 2.366* 2.201* 0.14 0.12 -0.02 20.00% LONG TERM (LT) EQUILIBRIUM Net cash flow to LT Debt (NCFTLTD) 0.16 0.13 -0.03 36.67% Privatization in or Before 1990 12 0.415 0.051 2.521* 2.366* 0.13 0.12 -0.01 38.33% 20.58 21.17 0.886 0.22 0.19 -0.03 33.33% Privatization After 1990 30 1.143 1.119 3.479* 3.823* 0.19 0.14 -0.05 33.33% CAPITAL STRUCTURE Total Debt to Total Assets (TDTA) 0.65 0.62 -0.03 60.00% Privatization in or Before 1990 12 0.413 0.317 2.201* 2.521* 0.68 0.58 -0.10 66.67% 23.08 20.14 0.474 0.79 0.67 -0.12 53.33% Privatization After 1990 30 0.701 1.277 2.982* 3.724* 0.82 0.77 -0.05 60.00%

* rejection of H0 at five percent level of significance

122 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

Table 13. Comparisons of performance changes following privatization for firms that have shareholders in management versus firms that do not have shareholders in management

This table presents comparisons of performance changes for firms that have shareholders in management versus firms that do not have shareholders in management. The table presents the results of the Wilcoxon rank sum test (with its z-statistic) - that is employed as a test for significance for change in mean and median values between before and after privatization - and of the Kruskal-Wallis test for firms that have shareholders in management versus firms that do not have shareholders in management- in mean terms and in median terms respectively (statistic mentions the 'p' value using the chi-squared approximation) - for each empirical proxy and each subsample of the pair. The table presents the number of useable observations, the mean and the median values of the proxy before and after privatization, their change in the proxy’s value after versus before privatization, the respective test of significance for the mean and median change, the mean rank of the KW test between shareholders in management and non-shareholders in management subsample and the respectively statistic 'p' value for mean and median comparison.

Z-statistics Z -statistics KW Results for differences Percentage Z-statistics Z- statistics for for between subsamples for mean of firms for for Mean Before Mean After Mean Change difference difference N with improved significance significance Mean Rank KW test VARIABLES (Median) (Median) (Median) in Means in Medians performance performance performance (After- (After- Mean (Median) (Mean) (Median) SM NoSM 'p' value Before) Before) PROFITABILITY Return on Sales (ROS) 0.06 0.27 0.21 63.85% Shareholders in Management 26 1.830* 1.829* 3.180* 3.296* 0.26 0.41 0.15 83.85% 23.12 18.88 0.027* 0.13 0.15 0.02 46.25% Non-Shareholders in Management 16 0.259 0.492 2.666* 2.366* 0.14 0.19 0.05 43.75% OPERATING EFFICENCY Sales Efficiency (SALEFF) 0.81 2.48 1.67 69.23% Shareholders in Management 26 2.299* 2.408* 3.823* 3.724* 0.83 2.59 1.76 69.23% 22.65 19.63 0.437 1.13 1.21 0.08 51.25% Non-Shareholders in Management 16 2.457* 2.224* 3.180* 3.059* 1.19 1.22 0.03 55.00% CAPITAL INVESTMENT Real Capital Exp. to Sales (RCESA) 0.39 0.79 0.40 58.46% Shareholders in Management 26 1.579 1.609 2.803* 3.059* 0.19 0.53 0.34 56.15% 24.50 19.13 0.501 0.61 0.69 0.08 50.00% Non-Shareholders in Management 16 0.517 1.447 2.521* 2.023* 0.39 0.43 0.04 41.25% REAL OUTPUT Real Sales (SAL) 0.69 3.49 2.80 69.23% Shareholders in Management 26 2.328* 2.540* 3.724* 3.823* 0.71 2.89 2.18 73.08% 24.65 19.25 0.059* 1.23 2.01 0.78 68.75% Non-Shareholders in Management 16 0.931 1.008 2.934* 2.934* 1.39 2.17 0.78 68.75% EMPLOYMENT Total Employment (EMPL) 2122.00 1675.00 -447.00 62.31% Shareholders in Management 26 1.562* 1.733* 3.724* 3.516* 2134.00 1841.25 -292.75 61.54% 21.12 22.13 0.796 1844.00 1799.00 -45.00 55.00% Non-Shareholders in Management 16 1.862* 2.430* 1.826* 3.180* 1983.00 1899.50 -83.50 51.25% DIVIDEND POLICY Dividend to Sales (DIVSAL) 0.00 1.00 1.00 73.08% Shareholders in Management 26 1.013 1.261 3.408* 3.621* 0.00 1.01 1.01 65.38% 24.54 16.56 0.040* 0.01 0.03 0.01 46.25% Non-Shareholders in Management 16 1.519 1.647 2.666* 2.201* 0.01 0.02 0.01 47.50% TREASURY Treasury Aplications (TA) 108220.00 109675.00 1455.00 36.15% Shareholders in Management 26 1.586 1.609 2.934* 2.666* 116520.00 119230.00 2710.00 34.62% 22.29 20.72 0.754 95428.00 101091.00 5663.00 36.25% Non-Shareholders in Management 16 1.602 1.454 2.666* 2.366* 93450.00 103229.00 9779.00 33.75% ACTIVITY LEVELS Sales to Total Assets (STA) 0.39 1.39 1.00 64.62% Shareholders in Management 26 1.627* 1.750* 2.803* 2.521* 0.35 1.36 1.01 60.77% 25.35 21.75 0.049* 0.61 0.81 0.20 53.75% Non-Shareholders in Management 16 0.052 0.103 2.366* 2.366* 0.38 0.41 0.03 53.75% SHORT TERM (ST) EQUILIBRIUM Cash/Banks to ST Debt (CBTSTD) 0.23 0.12 -0.11 19.23% Shareholders in Management 26 2.335* 2.475* 2.023* 1.826* 0.19 0.09 -0.10 15.38% 23.00 20.94 0.311 0.19 0.17 -0.02 25.00% Non-Shareholders in Management 16 0.795 1.503 1.826* 1.826* 0.17 0.15 -0.02 25.00% LONG TERM (LT) EQUILIBRIUM Net cash flow to LT Debt (NCFTLTD) 29.55 29.21 -0.34 67.69% Shareholders in Management 26 0.608 0.761 3.516* 3.408* 9.44 8.89 -0.55 67.69% 20.85 21.94 0.816 19.88 13.23 -6.65 62.50% Non-Shareholders in Management 16 1.568* 1.916* 2.803* 2.934* 5.73 1.99 -3.74 68.75% CAPITAL STRUCTURE Total Debt to Total Assets (TDTA) 0.88 0.33 -0.55 62.31% Shareholders in Management 26 0.591* 1.660* 3.724* 3.724* 0.99 0.69 -0.30 69.23% 25.27 18.88 0.087* 0.69 0.65 -0.04 50.00% Non-Shareholders in Management 16 0.078 1.223 2.521* 2.521* 0.65 0.61 -0.04 50.00%

* rejection of H0 at five percent level of significance

123 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

Table 14. Comparisons of performance changes following privatization for firms that were privatized partially versus firms that were privatized totally

This table presents comparisons of performance changes for firms that were privatized partially versus firms that were privatized totally. The table presents the results of the Wilcoxon rank sum test (with its z-statistic) - that is employed as a test for significance for change in mean and median values between before and after privatization - and of the Kruskal-Wallis test firms that were privatized partially versus firms that were privatized totally - in mean terms and in median terms respectively (statistic mentions the 'p' value using the chi-squared approximation) - for each empirical proxy and each subsample of the pair. The table presents the number of useable observations, the mean and the median values of the proxy before and after privatization, their change in the proxy’s value after versus before privatization, the respective test of significance for the mean and median change, the mean rank of the KW test between Partial Privatization and Total Privatization subsample and the respectively statistic 'p' value for mean and median comparison.

Z-statistics Z- statistics Percentage Z-statistics Z- statistics KW Results for differences for for of firms for for between subsamples for mean Mean Before Mean After Mean Change VARIABLES N difference difference with improved significance significance Mean Rank KW test (Median) (Median) (Median) in Means in Medians performance performance performance (After- (After-Before) Mean (Median) (Mean) (Median) PP TP 'p' value Before) PROFITABILITY Return on Sales (ROS) 0.13 0.14 0.01 56.67% Partial (Revenue) Privatization 20 0.807 0.393 3.316* 2.894* 0.25 0.26 0.01 52.38% 19.14 23.86 0.05* 0.07 0.23 0.16 65.45% Total (Control ) Privatization 22 1.671* 1.787* 2.793* 2.793* 0.14 0.29 0.15 65.45% OPERATING EFFICENCY Sales Efficiency (SALEFF) 0.87 1.55 0.68 66.19% Partial (Revenue) Privatization 20 2.773* 3.126* 3.496* 3.476* 0.82 1.59 0.77 66.19% 21.90 23.10 0.02* 0.97 1.99 1.02 72.73% Total (Control ) Privatization 22 2.119* 1.719* 3.496* 3.316* 0.97 2.12 1.15 73.64% CAPITAL INVESTMENT Real Capital Exp. to Sales (RCESA) 0.49 0.51 0.02 52.38% Partial (Revenue) Privatization 20 1.309 0.160 2.934* 2.934* 0.22 0.23 0.01 52.38% 18.38 24.62 0.142 0.66 0.98 0.32 67.27% Total (Control ) Privatization 22 1.912* 2.190* 2.201* 2.201* 0.43 0.69 0.26 67.27% REAL OUTPUT Real Sales (SAL) 1.12 2.01 0.89 61.43% Partial (Revenue) Privatization 20 1.417 1.432 2.708* 2.416* 1.24 2.22 0.98 66.19% 14.48 28.52 0.05* 0.71 2.98 2.27 78.18% Total (Control ) Privatization 22 1.693* 1.712* 3.408* 3.296* 0.75 2.89 2.14 73.64% EMPLOYMENT Total Employment (EMPL) 2234.00 2111.00 -123.00 58.57% Partial (Revenue) Privatization 20 1.216 1.511 2.201* 3.496* 2213.00 2031.00 -182.00 56.19% 15.86 27.14 0.01* 1977.00 1622.00 -355.00 76.36% Total (Control ) Privatization 22 1.602* 1.799* 2.521* 3.210* 1912.00 1655.00 -257.00 69.09% DIVIDEND POLICY Dividend to Sales (DIVSAL) 0.02 0.03 0.01 56.67% Partial (Revenue) Privatization 20 1.613 1.622 3.296* 3.296* 0.01 0.02 0.01 56.67% 22.67 19.33 0.093 0.01 0.12 0.11 53.64% Total (Control ) Privatization 22 1.229 1.240 3.296* 2.666* 0.01 0.07 0.06 50.91% TREASURY Treasury Aplications (TA) 97890.00 99767.00 1877.00 42.86% Partial (Revenue) Privatization 20 0.845 1.605 2.666* 2.521* 101667.00 109456.00 7789.00 38.10% 19.60 23.40 0.268 89710.00 121345.00 31635.00 44.55% Total (Control ) Privatization 22 1.618* 1.711* 3.059* 2.521* 97655.00 128769.00 31114.00 36.36% ACTIVITY LEVELS Sales to Total Assets (STA) 0.52 0.71 0.19 53.33% Partial (Revenue) Privatization 20 1.532 1.101 2.296* 2.276* 0.52 0.59 0.07 53.33% 19.52 23.48 0.027* 0.31 1.42 1.11 65.45% Total (Control ) Privatization 22 1.729* 1.564* 2.793* 2.521* 0.39 1.77 1.38 66.36% SHORT TERM (ST) EQUILIBRIUM Cash/Banks to ST Debt (CBTSTD) 0.23 0.16 -0.07 29.05% Partial (Revenue) Privatization 20 2.156* 3.027* 1.826* 1.342 0.21 0.13 -0.08 19.52% 20.95 22.05 0.772 0.18 0.16 -0.02 22.73% Total (Control ) Privatization 22 1.042 1.265 2.023* 2.201* 0.15 0.12 -0.03 27.27% LONG TERM (LT) EQUILIBRIUM Net cash flow to LT Debt (NCFTLTD) 25.23 18.66 -6.57 32.38% Partial (Revenue) Privatization 20 1.429 1.441 2.934* 3.180* 7.34 2.64 -4.70 31.90% 20.76 22.24 0.697 27.87 25.22 -2.65 33.64% Total (Control ) Privatization 22 1.225 1.431 3.296* 3.180* 8.87 6.45 -2.42 39.09% CAPITAL STRUCTURE Total Debt to Total Assets (TDTA) 0.79 0.62 -0.17 47.62% Partial (Revenue) Privatization 20 0.112 0.174 2.803* 3.180* 0.88 0.61 -0.27 61.90% 18..24 23.76 0.059* 0.71 0.67 -0.04 45.45% Total (Control ) Privatization 22 1.743* 1.770* 2.803* 3.180* 0.67 0.65 -0.02 59.09%

* rejection of H0 at five percent level of significance

124 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

Table 15. Comparisons of performance changes following privatization for firms that were restructured before privatization versus firms that were not restructured before privatization

This table presents comparisons of performance changes for firms that were restructured before privatization versus firms that were not restructured before privatization. The table presents the results of the Wilcoxon rank sum test (with is z-statistic) - that is employed as a test for significance for change in mean and median values between before and after privatization - and of the Kruskal-Wallis test firms that were restructured before privatization versus firms that weren’t - in mean terms and in median terms respectively (statistic mentions the 'p' value using the chi-squared approximation) - for each empirical proxy and each subsample of the pair. The table presents the number of useable observations, the mean and the median values of the proxy before and after privatization, their change in the proxy’s value after versus before privatization, the respective test of significance for the mean and median change, the mean rank of the KW test between Partial Restructured and Not Restructured subsample and the respectively statistic 'p' value for mean and median comparison.

Z- Z-statistics Z-statistics Z-statistics KW Results for differences between subsamples for Percentage of firms Mean Mean statistics for for for mean VARIABLE Mean After with improved N Before Change for difference significanc significanc Mean Rank KW test S (Median) performance (Median) (Median) difference in Medians e e in Means (After- Mean (Median) performan performan R NoR 'p' value (After- Before) ce ce PROFITABILITY Return on Sales (ROS) 0.07 0.31 0.24 63.75% Restructured 16 1.663* 1.569* 2.336* 2.491* 0.14 0.33 0.19 60.00% 25.55 17.33 0.039* Not 0.11 0.14 0.03 44.00% 25 0.503 0.329 3.496* 3.316* Restructured 0.23 0.25 0.02 40.00% OPERATING EFFICENCY Sales Efficiency (SALEFF) 0.83 1.89 1.06 78.75% Restructured 16 1.752* 1.698* 2.884* 2.894* 0.79 1.99 1.20 78.75% 21.13 21.73 0.877 Not 0.94 1.57 0.63 56.00% 25 2.333* 2.694* 3.783* 3.774* Restructured 0.95 1.59 0.64 53.08% CAPITAL INVESTMENT Real Capital Exp. to Sales (RCESA) 0.49 1.02 0.53 60.00% Restructured 16 1.651* 1.694* 2.521* 2.376* 0.29 0.53 0.24 53.75% 25.13 19.27 0.013* Not 0.63 0.59 -0.04 46.00% 25 0.913 1.095 2.656* 2.795* Restructured 0.37 0.38 0.01 38.46% REAL OUTPUT Real Sales (SAL) 0.81 2.54 1.73 75.00% Restructured 16 2.553* 2.212* 3.169* 3.049* 0.79 2.77 1.98 75.00% 24.45 18.96 0.056* Not 0.99 1.72 0.73 48.00% 25 0.989 1.441 3.591* 3.694* Restructured 0.99 1.56 0.57 49.23% EMPLOYMENT Total Employment (EMPL) 2101.00 1922.00 -179.00 37.50% Restructured 16 0.694 1.219 2.191* 2.914* 2002.00 1827.00 -175.00 38.75% 23.69 18.45 0.053* Not 2179.00 1838.00 -341.00 52.00% 25 2.384* 2.379* 2.491* 2.896* Restructured 2191.00 1848.00 -343.00 69.23% DIVIDEND POLICY Dividend to Sales (DIVSAL) 0.01 0.13 0.12 56.25% Restructured 16 1.230 1.458 2.666* 2.521* 0.01 0.11 0.10 50.00% 21.19 21.92 0.336 Not 0.02 0.05 0.03 52.00% 25 1.242 1.441 3.724* 3.823* Restructured 0.01 0.03 0.02 57.69% TREASURY Treasury Aplications (TA) 103220.00 119454.00 16234.00 37.50% Restructured 16 1.397 1.390 2.201* 2.023* 101221.00 119803.00 18582.00 31.25% 21.34 21.37 0.445 Not 99671.00 107345.00 7674.00 40.00% 25 0.957 0.892 3.408* 3.059* Restructured 110010.00 119888.00 9878.00 42.31% ACTIVITY LEVELS Sales to Total Assets (STA) 0.43 1.19 0.76 61.25% Restructured 16 2.329* 2.121* 2.023* 2.023* 0.39 0.66 0.27 61.25% 25.06 18.15 0.02* Not 0.58 0.82 0.24 54.00% 25 0.500 0.013 2.934* 2.666* Restructured 0.59 0.49 -0.10 48.46% SHORT TERM (ST) EQUILIBRIUM Cash/Banks to ST Debt (CBTSTD) 0.23 0.18 -0.05 31.25% Restructured 16 0.255 1.108 2.023* 1.604* 0.19 0.16 -0.03 18.75% 17.31 25.23 0.042* Not 0.19 0.15 -0.04 16.00% 25 2.716* 2.912* 1.826* 2.023* Restructured 0.14 0.11 -0.03 19.23% LONG TERM (LT) EQUILIBRIUM Net cash flow to LT Debt (NCFTLTD) 23.12 19.39 -3.73 42.50% Restructured 16 0.384 0.792 2.803* 2.803* 6.20 4.87 -1.33 42.50% 22.03 22.94 0.129* Not 28.45 24.01 -4.44 44.00% 25 1.766* 1.730* 3.516* 3.408* Restructured 9.70 3.65 -6.05 41.54% CAPITAL STRUCTURE Total Debt to Total Assets (TDTA) 0.70 0.69 -0.01 31.25% Restructured 16 0.776 0.647 2.023* 2.521* 0.73 0.69 -0.04 40.00% 17.19 26.31 0.0507* Not 0.81 0.59 -0.22 32.00% 25 1.654* 1.799* 2.521* 2.936* Restructured 0.67 0.50 -0.17 49.23%

* rejection of H0 at five percent level of significance

125 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

LOWER TAXES OR HIGHER EXECUTIVE BONUSES: HOW INVENTORY VALUATION CHOICES BEST EXHIBIT US CORPORATE GOVERNANCE FAILINGS

Kevin A. Diehl*

Abstract

This research seeks to update and finally determine for the Fortune 500 whether the market values the inventory valuation choice of last-in, first-out (LIFO) over first-in, first-out (FIFO) as some signal of reporting and management quality. The market can adjust LIFO earnings to FIFO earnings. Thus, the only issue then is that companies choosing FIFO pay higher taxes, which shareowners should disfavor. Indeed, only 20 percent of the Fortune 500 utilize LIFO to value any inventory. However, after Spearman correlations and logistic regression, the research statistically significantly shows that investors are willing to give premiums on the price of stock for the choice of LIFO. Thus, companies should choose LIFO to reduce taxes and increase their stock prices.

Keywords: Taxes, Executive Bonuses, Corporate Governance

* Western Illinois University (US), 3561 60th Street, Moline, IL 61265 Tel.: (309) 762-3999, Extension 62297 E-mail: [email protected]

Introduction and Literature Review therein executives’ bonuses and directors’ equity- based compensation (Morse and Richardson, 1983). This research for the first time seeks to update and From the reported earnings perspective, the finally determine for the Fortune 500 whether the market should not care as required disclosures permit market values the inventory valuation choice of last- investors to adjust LIFO earnings to FIFO earnings in, first-out (LIFO) over first-in, first-out (FIFO) as (Biddle and Lindahl, 1982). However, from the some signal of reporting and management quality. perspective of evaluating the overall management The choice between FIFO and LIFO inventory teams, investors should favor companies that select valuation has long been troublesome under US LIFO as it does decrease taxes payable to the GAAP. For financial accounting purposes, the theory government and can be adjusted to FIFO for is that the choice enables companies properly to comparing reported earnings (Biddle and Lindahl, match their inventory accounting to the inventory’s 1982). actual physical movement (Cook et al., 2007). The problem is best illustrated through the low However, in application, the choice of FIFO has percentage of Fortune 500 companies that select become grounds for earnings management (Cook et LIFO. As this current research pursuit shows, only 20 al., 2007). percent of the Fortune 500 report utilizing the LIFO Whatever choice is made for financial method for any part of their inventory. accounting purposes must also be utilized for tax There are some limitations on the choice of purposes. In periods of increasing prices, LIFO LIFO that lead to this low percentage. Because reports higher cost of goods sold and therein lower International Financial Reporting Standards (IFRS) net income. However, it also results in lower taxes do not permit the utilization of LIFO, there is some (Sunder, 1975). In periods of increasing prices, FIFO efficiency benefit from utilizing FIFO for valuing all reports lower cost of goods sold and therein higher the inventory of international companies (Fosbre et net income. Executives and the directors who review al., 2010). However, the higher tax cost from not their actions should favor LIFO as it generally utilizing LIFO for domestic reporting is greater than decreases taxes payable to the government, therein the cost saved for standardizing global inventory benefiting the companies (Hughes and Schwartz, reporting from choosing FIFO. 1988). In fact, as Hughes and Schwartz (1988) Companies with higher leverage ratios tend to remark, the companies utilizing FIFO are costing disfavor LIFO as this method does tend to report themselves billions of dollars annually in total. lower earnings (Gul, 2002). Creditors establish ratio However, these executives and directors tend to favor numbers necessary to stay solvent under the debt FIFO as it generally increases reported earnings and contracts (Gul, 2002). Thus, to increase their odds of

126 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

satisfying these ratios, companies choose the higher Hypothesis earnings reporting methodology, FIFO, (Gul, 2002). Indeed, staying solvent to remain in business should Investors are willing to give premiums on the price of be an important objective for any company (Gul, stock for companies that choose LIFO to report at 2002). In fact, Hughes and Trezevant (1998) find that least some of their inventory. companies that move to LIFO begin lowering their The measure for this test is whether the choice debt levels because of the implications of higher cost of LIFO is directly statistically significantly related to of goods sold reported under LIFO and its influence the percentage change in price from 2004 through on the ease with which debt contracts are satisfied. 2009 and is inversely statistically significantly related Whether investors value the choice of LIFO to the percentage change in earnings per share from over FIFO is tested through Spearman correlations 2004 through 2009 decreases. Separating the two and logistic regression with the choice of LIFO variables from the normal price to earnings or designated as 1 and the choice of FIFO designated as earnings to price ratios permits increased evidence of 0. The other variables are the number of financial the presence of this investor willingness to pay higher experts on the board of directors, the number of price to earnings per share ratios or lower earnings public accounting experience directors, the per share to price ratios, which would signal proof of percentage change in price to earnings per share this hypothesis. during the 2004 to 2009 time period, the ln of total The reason that investors would pay more is assets during those years, and the average debt to because of the higher quality of earnings. LIFO is equity ratio during the 2004 to 2009 time period. The more conservative than FIFO with regard to earnings. sample is the Fortune 500, which at least one other The other reason is that choosing LIFO signals prominent researcher has consistently utilized quality executive management and director review of (Yermack, 1997; Yermack, 2004). As various company activities. LIFO does result in lower taxes supporting data has to be hand collected, this sample payable to the government. size is efficient but still robust (Yermack, 1997; The control variables involve the number of Yermack, 2004). financial experts on the board, the number of The choice of LIFO is shown to be directly directors with public accounting experience, the ln of statistically significantly correlated to the percentage average total assets during 2004 through 2009, and change in price to earnings per share between 2004 the average debt to equity ratios during 2004 through and 2009. This finding supports the hypothesis that 2009. As the number of financial experts qualified investors are willing to give premiums on earnings of under Sarbanes-Oxley on the board increases, the companies utilizing LIFO because of the quality of expectation is that the choice of LIFO would reporting (more conservative earnings reporting decrease. The reason is that, better knowing the choice and therein showing the lack of earnings accounting rules under US GAAP, financial experts management in at least this regard) and the quality of would likely emphasize higher reported earnings the executives and directors in favoring the interests from FIFO rather than LIFO. Despite investors of the company over their own. favoring LIFO, directors still tend to believe the However, the choice of LIFO is inversely quality of their review of executives’ actions is based statistically significantly related to the presence of on whether the respective company is showing public accounting experience on the board of growing earnings, which is easier to do under FIFO. directors. This fact is qualified as it is discovered that As the number of directors on the board with the public accounting experience on boards tends to public accounting experience increases, the same be from auditing, not tax, backgrounds. Thus, the expectation exists that the utilization of LIFO would public accountants expectedly favor the method that decrease. The same reasoning is applicable as well. reports higher earnings as it is in their self-interest to There would be some overlap between financial the extent of their director equity compensation to experts and public accounting experience. However, report the higher earnings. It is also in keeping with financial experts would include more individuals. The their general experience with the executives of their public accounting experience in particular is included audit clients who have tended to favor the choice of as well to indicate whether the precise knowledge of FIFO for their bonus reasons as well. the auditing process and the accounting rules would The choice of LIFO is also inversely statistically lead to greater reluctance to utilize LIFO. significantly related to average debt to equity during The ln of average total assets acts as the control the 2004 to 2009 time period. This fact supports the on whether larger companies tend to select any importance to executives and directors who review method in particular compared to smaller companies. their actions of choosing the higher earnings The average debt to equity ratios are included as the reporting method, FIFO, to make meeting debt control on higher leveraged companies tending to contracts easier. disfavor LIFO based on the increased difficulty of satisfying debt contracts.

127 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

Methodology and Data Set Logistic regression has the same results with 3.120 statistically significant at the .05 level for %DPr and Whether investors value the choice of LIFO over -1.109 statistically significant at the .01 level for FIFO is tested through Spearman correlations and %DEPS. logistic regression with the choice of LIFO These findings support the hypothesis that designated as 1 and the choice of FIFO designated as investors are willing to give premiums on earnings of 0. The other variables are the number of financial companies utilizing LIFO because of the quality of experts on the board of directors, the number of reporting (more conservative earnings reporting public accounting experience directors, the choice and therein showing the lack of earnings percentage change in price to earnings per share management in at least this regard) and the quality of during the 2004 to 2009 time period, the ln of total the executives and directors in favoring the interests assets during those years, and the average debt to of the company over their own. equity ratio during the 2004 to 2009 time period. However, the choice of LIFO is inversely statistically significantly related to the presence of LIFO = + FinExp + PubAcct + %DPr public accounting experience on the board of + %DEPS + lnAssets + AveD/E + (1) directors. Under Spearman, it is -.263 at the .05 level of statistical significance, and, under logistic The sample is the Fortune 500, which at least regression, it is -2.793 at the .05 level of statistical one other prominent researcher has consistently significance. utilized (Yermack, 1997; Yermack, 2004). As various These results are qualified as the public supporting data has to be hand collected, this sample accounting experience on boards tends to be from size is efficient but still robust (Yermack, 1997; auditing, not tax, backgrounds. Thus, the public Yermack, 2004). For the logistic regression, Fortune accountants expectedly favor the method that reports 500 companies that utilize LIFO are paired with higher earnings as it is in their self-interest to the Fortune 500 companies that do not utilize LIFO. extent of their director equity compensation to report Companies are paired based on standard industrial the higher earnings. It is also in keeping with their code (SIC) and then based on ln of total assets within general experience with the executives of their audit the same SIC. clients who have tended to favor the choice of FIFO for their bonus reasons as well. Results The choice of LIFO is also inversely statistically significantly related to average debt to equity during Tables 1 and 2 show the following information. The the 2004 to 2009 time period. This result is shown choice of LIFO is shown to be directly statistically only through the logistic regression with -2.603 at the significantly correlated to the percentage change in .01 level. This fact supports the importance to price to earnings per share between 2004 and 2009. executives and directors who review their actions of Spearman correlations have it at .176 statistically choosing the higher earnings reporting method, FIFO, significant at the .05 level for %DPr and -.192 to make meeting debt contracts easier. statistically significant at the .01 level for %DEPS.

Table 1. Spearman Correlations

LIFO FinExp PubAcct %DPr %DEPS lnAssets AveD/E LIFO 1.00

FinExp -.105 1.00 (.152) PubAcct -.263** .129 1.00 (.030) (.294) %DPr .176** -.056 .004 1.00 (.013) (.443) (.973) %DEPS -.192*** .013 .347*** .170** 1.00 (.007) (.856) (.004) (.016) lnAssets -.013 .067 .068 .121* .059 1.00 (.860) (.364) (.583) (.089) (.405) AveD/E -.015 .003 -.180 -.077 -.119* -.060 1.00 (.839) (.970) (.149) (.283) (.096) (.401)

*** Statistically significant at the .01 level; ** statistically significant at the .05 level; * statistically significant at the .10 level. LIFO stands for choice of LIFO to value any inventory. FinExp represents the number of Sarbanes-Oxley qualified financial experts on the board of directors. PubAcct stands for the number of directors on the board with public accounting experience. %DPr represents the percentage change in price from 2004 through 2009. %DEPS stands for the percentage change in earnings per share from 2004 through 2009. lnAssets represents the ln of total assets on average over those years; AveD/E stands for the average debt to equity ratio during those years.

128 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

Table 2. Logistic Regression

Predicted Estimated Coefficients Variables Signs (x2) Intercept ? 10.428 (5.386) FinExp - -.637 (2.501) PubAcct - -2.793** (3.966) %DPr = 3.120** (5.753) %DEPS - -1.109*** (8.747) lnAssets ? -.208 (.322) AveD/E - -2.603*** (11.066) Observations 200 Likelihood ratio 47.612 Pseudo .632

Implications savings,” Journal of Accounting Research, 20, pp. 551-588. To get to 100 percent compliance with the LIFO 2. Cook, K., Huston, G. R., and Kinney. M. (2007), choice domestically for valuing inventory, “Managing earnings by manipulating production: shareowners should propose for vote at each the effects of cost structure and inventory company’s annual meeting that the board be required valuation method,” Working Paper, Texas. to enforce executives’ choice of LIFO over FIFO as it 3. Fosbre, An., Fosbre, Pa., and Kraft, E. (2010), enhances the value of the company. The choice of “Roadblock to US adoption of IFRS is LIFO LIFO does so through reduction of taxes payable to inventory valuation,” Global Journal of Business the government. Research, 4, pp. 41-49. Boards on their own should implement the 4. Gul, F. (2002), “Free cash, debt-monitoring, and requirement that executives’ compensation be based managers’ LIFO/FIFO policy choice,” Journal of on numbers adjusted to LIFO, therein encouraging Corporate Finance, 7, pp. 475-492. executives to take on LIFO and simultaneously 5. Hughes, P. J., and Schwartz, E. S. (1988), “The denying any benefit to bonuses from the choice of LIFO/FIFO choice: an asymmetric information FIFO to show higher earnings. approach,” Journal of Accounting Research, 26, Likewise, boards and executives should lobby supplement, pp. 41-58. government or the financial industry itself to have 6. Hughes, P. J., and Trezevant, R. (1998), “The financial ratios in debt contracts be based on LIFO capital structure and LIFO/FIFO choices: tests of numbers. As such, the financial industry should be cash signaling and tax-shield substitution,” willing to give greater leeway in the ratios as the Working Paper, California. LIFO method is more conservative in reporting. 7. Sunder, S. (1975), “Stock price and risk related In the end then, the choice of LIFO is more to accounting changes in [the] inventory efficient for companies. Thus, compensation and debt valuation,” The Accounting Review, 50, pp. 305- contracts should be based on LIFO numbers. 315. 8. Yermack, D. (1997), “Good timing: CEO stock References option awards and company news announcements,” Journal of Finance, 52, pp. 1. Biddle, G. C., and Lindahl, F. W. (1982), “Stock 449-476. price reactions to LIFO adoptions: the 9. Yermack, D. (2004) “Remuneration, Retention, association between excess returns and LIFO tax and Reputation Incentives for Outside Directors,” Journal of Finance, 59, pp. 2281-308.

129 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

SUBSCRIPTION FORM TO "CORPORATE OWNERSHIP & CONTROL", "CORPORATE BOARD: ROLE, DUTIES & COMPOSITION", “RISK GOVERNANCE & CONTROL: FINANCIAL MARKETS & INSTITUTIONS”, “JOURNAL OF GOVERNANCE AND REGULATION”

Copy this form and follow guidelines to fill it up. I would like to buy (underline what you choose):

For individual subscribers: 1. Journal of Corporate Ownership & Control: 1.1. Printed version of the journal - US$260 / €220. 1.2. Electronic version - US$185 / €150. 2. Corporate Board: role, duties & composition: 2.1. Printed version of the journal - US$240 / €190. 2.2. Electronic version - US$180 / €145. 3. Risk Governance & Control journal: financial markets & institutions: 3.1. Printed version of the journal - US$240 / €190. 3.2. Electronic version - US$180 / €145. 4. Journal of Governance and Regulation: 3.1. Printed version of the journal - US$240 / €190. 3.2. Electronic version - US$180 / €145.

For institutional subscribers: 1. Journal of Corporate Ownership & Control: 1.1. Printed version of the journal and electronic subscription - US$1240 / €890. 1.2. Printed version of the journal - US$1080 / €790. 1.3. Electronic version - US$730 / €650. 2. Corporate Board: role, duties & composition: 2.1. Printed version of the journal and electronic subscription - US$820 / €760. 2.2. Printed version of the journal - US$960 / €720. 2.3. Electronic version - US$650 / €590. 3. Risk Governance & Control journal: financial markets & institutions: 3.1. Printed version of the journal and electronic subscription - US$920 / €790. 3.2. Printed version of the journal - US$860 / €620. 3.3. Electronic version - US$550 / €490. 3. Journal of Governance and Regulation: 3.1. Printed version of the journal and electronic subscription - US$920 / €790. 3.2. Printed version of the journal - US$860 / €620. 3.3. Electronic version - US$550 / €490.

Underline one of the payment methods you prefer, and write amount to pay (if you prefer, you can pay by one cheque/bank transfer to subscribe to both journals): 1. I enclose a cheque for US$ / €______; 2. Send me an invoice for US$ / €______(indicate currency correctly). Write your contact details here:

Name______Position______Institution______Address______E-mail______Tel______Fax______

Please, send this form (with a cheque if you prefer to pay by cheque) at:

Dr. Alexander Kostyuk Publishing house "Virtus Interpress" Postal Box 36 Sumy 40014 Ukraine Ask for invoice by fax at +380-542-698125 (if you want to pay by bank transfer). Enter our web-site to download the subscription form at www.virtusinterpress.org. If you prefer, you can subscribe through our global subscription agents – SWETS INFORMATION SERVICES of EBSCO.

130 Journal of Governance and Regulation / Volume 1, Issue 2, 2012

LIBRARY RECOMMENDATION FORM PLEASE FORWARD THIS FORM TO YOUR INSTITUTION FOR ORDER CONSIDERATION

To: Subject Librarian ______From: ______Title: ______Department: ______Email: ______

I would like the library to order the titles selected below:

- Corporate Governance: An International Outlook - ISBN 978-966-96872-2-7 (390 pages, year of publishing - 2011, authors: Dr. Alexander Kostyuk, Professor at Ukrainian Academy of Banking (Ukraine), Dr. Carsten Gerner-Beuerle, Professor at London School of Economics (The UK), Dr. Rodolfo Apreda, Professor at Universidad del CEMA (Argentina) - price - EURO 56 Number of copies -______- Anti-Crisis Paradigms of Corporate Governance in Banks: a new institutional outlook - ISBN978-966-96872-0-9 (482 pages, year of publishing - 2010, authors - Alexander N. Kostyuk, Fumiko Takeda, Kaoru Hosono, introduced by Oliver Williamson, a Nobel Prize Winner 2009 - price - EURO 68 Number of copies -______- Corporate Governance: a textbook - ISBN978-966-96872-0-3 (400 pages, year of publishing 2007, authors - Alexander N. Kostyuk, Udo C. Braendle, Rodolfo Apreda) - price - EURO 58 Number of copies -______- Corporate Governance in a Transition Economy: a monograph - ISBN966-96566-0-5 (155 pages, year of publishing - 2005, author - Alexander N. Kostyuk) - price - EURO 38 Number of copies -______- Corporate Board Practices: a monograph - ISBN966-96566-1-3 (156 pages, year of publishing - 2006, author - Alexander N. Kostyuk) - price - EURO 38 Number of copies -______- Euroasian Perspectives Of The Banking Systems Development - ISBN 978-966-96872-2-7 (170 pages, year of publishing - 2011, authors: Dr. Oleksandr Kostyuk, Dr. Michelle Lin, Dr. Ghassan Omet - price - EURO 24 Number of copies -______

I want to order all books in the list. Total price - EURO 255.

I recommend this book(s) for the following reasons: - I need to refer to this information frequently in the course of my work. - This book(s) is directly related to my field and is an important source of information for my research. - It is critical that students have access to this information to best facilitate their course work and academic pursuits. - This is an obligatory course book - Additional reasons: Signature: ______Date : ______

Information on these books may be obtained from the Publisher's Website: http://www.virtusinterpress.org Please send all orders and inquiries to: Publishing company "Virtus Interpress" Postal Box 36 40014 Sumy, Ukraine E-mail: [email protected], [email protected]

131