Florida State University Libraries

Electronic Theses, Treatises and Dissertations The Graduate School

2007 Reverse Stock Splits: Motivations, Effectiveness and Stock Price Reactions Barry Marchman

Follow this and additional works at the FSU Digital Library. For more information, please contact [email protected]

THE FLORIDA STATE UNIVERSITY

COLLEGE OF BUSINESS

REVERSE STOCK SPLITS: MOTIVATIONS, EFFECTIVENESS AND STOCK PRICE REACTIONS

By

BARRY MARCHMAN

A Dissertation submitted to the Department of Finance in partial fulfillment of the requirements for the degree of Doctor of Philosophy

Degree Awarded: Summer Semester 2007

The members of the Committee approve the dissertation of Barry Marchman defended June 5, 2005

Gary Benesh Professor Directing Dissertation

Rick Morton Outside Committee Member

Stephen Celec Committee Member

Yingmei Cheng Committee Member

Approved:

Caryn L Beck-Dudley, Dean, College of Business

The Office of Graduate Studies has verified and approved the above named committee members.

ii

For RC JAM and Evelyn

iii ACKNOWLEDGEMENTS

A work of this nature is never done in a vacuum. I am grateful for all of the help and encouragement from family, friends, and colleagues. My deepest debt of gratitude and most sincere appreciation goes to Gary Benesh. Thank you for seeing me through. I am grateful to my committee members, Rick Morton, Yingmei Cheng, and Steve Celec for helping me finish. I would also like to thank Don Nast and Bill Anthony for their constant encouragement and their belief in me. I thank Bill Christiansen, James Ang, and David Peterson for all that they taught me. Pam Peterson got me started on this study and I acknowledge her contribution. I thank Scheri Martin for keeping me on track. I am grateful to Caroline Jernigan and Daiho Uhm for SAS help and I am thankful to all my FSU officemates, classmates, and other colleagues for their support and friendship. I am grateful to my friends and colleagues at SBI, especially Amos Bradford and Charles Evans, for giving me the means to provide for my family during this process. I thank Kenneth Gray and Colin Benjamin for their constant encouragement. I am grateful to my family. I thank my parents, Mac and Tillie Marchman, for all of their love, support, and encouragement through this process. I owe a special debt of gratitude to Jack and Dianne Steen for making this all possible. I am especially indebted to my biggest cheerleader and best friend without whom I could not have finished, Evelyn. I am ultimately thankful to God for all of these relationships and for His blessings.

iv TABLE OF CONTENTS

LIST OF TABLES...... viii LIST OF FIGURES...... x ABSTRACT ...... xi

CHAPTER 1: INTRODUCTION...... 1 1.1 Introduction and Motivation ...... 1 1.2 What Is a Reverse Stock Split?...... 3 1.2.2 A Brief Overview of the Evidence on Forward and Reverse Stock Splits ...... 4 1.2.2.1 Anecdotal ...... 4 1.2.2.2 Forward Split Empirical Findings ...... 6 1.2.2.3 Reverse Split Empirical Findings...... 7 1.3 Why Reverse Split?...... 7 1.3.1 Listing Requirements...... 8 1.3.2 Transaction Costs ...... 10 1.3.3 Marginability ...... 10 1.3.4 Signaling...... 11 1.4 Hypotheses ...... 12 1.4.1 Stock Price Reactions ...... 12 1.4.2 Financial Health...... 14 1.4.3 Listing Requirements...... 15 1.5 Outline of This Dissertation ...... 16

CHAPTER 2: LITERATURE REVIEW...... 17 2.1 Why Does a Firm Split Its Stock?...... 17 2.1.1 Asymmetric Information (Signaling)...... 19 2.1.2 Signaling Combined with Transaction Costs ...... 21 2.1.3 Trading Range...... 21 2.1.4 Institutional Reasons to Reverse Split...... 24 2.1.5 Margin Regulations ...... 25 2.1.6 Reorganization ...... 26 2.1.7 Liquidity ...... 27 2.1.8 Risk ...... 27 2.1.9 Investor Behavior in Relation to Stock Splits...... 28 2.1.9.1 Framing ...... 29 2.1.9.2 Price Preference...... 30 2.1.9.3 Herding...... 31 2.1.10 Optimal Tick Size (Market Microstructure)...... 31 2.2 Conclusion ...... 32 2.2.1 What We Know...... 32 2.2.2 What We Do Not Know ...... 33 2.2.2.1 Managers’ Self-interest...... 34 2.2.2.2 Shareholders’ Best Interest ...... 35

v CHAPTER 3: METHODOLOGY AND RESULTS...... 36 3.1 Stock Price Reactions ...... 36 3.1.2 Raw Returns Around the Announcement Day...... 38 3.1.3 Market Adjusted Returns Around the Announcement Day ...... 40 3.1.4 Market Model Adjusted Returns Around the Announcement Day ...... 41 3.1.5 Announcement Day Conclusions ...... 42 3.1.6 Ex-date Raw Returns ...... 42 3.1.7 Ex-date Market Adjusted Returns...... 44 3.1.8 Ex-date Market Model Adjusted Returns...... 44 3.1.9 The relation between ex-date returns and announcement date returns ...... 45 3.1.10 Analysis of Subsamples Based on Firm Categories...... 46 3.1.10.1 Industry...... 46 3.1.10.2 Delisting threat and Motivations for Reverse Splits ...... 49 3.1.10.3 Price ...... 52 3.1.10.4 Size of the Firm ...... 53 3.1.10.5 Consolidation...... 54 3.1.11 Long-run Returns Following Reverse Stock Split ...... 55 3.1.12 Long Run Buy and Hold Returns...... 55 3.1.13 Long-run Abnormal Returns ...... 59 3.1.14 Matched Firm Analysis ...... 62 3.1.15 Analysis of Subsamples Based on Firm Categories...... 64 3.1.15.1 Industry...... 64 3.1.15.2 Delisting Threat and Motivations for Reverse Stock Splits ...... 65 3.1.15.3 Price Considerations ...... 66 3.1.15.4 Size of the Firm ...... 66 3.1.15.5 Consolidation...... 67 3.1.15.6 Long-run Analysis Concluding Remarks...... 68 3.2 Market Movements and Reverse Stock Splits...... 68 3.2.1 Hypothesis...... 68 3.2.2 Testing ...... 69 3.2.3 Results ...... 70 3.3 Financial Distress and Future Performance ...... 70 3.3.1 Hypothesis...... 70 3.3.2 Testing ...... 71 3.3.3 Results ...... 72 3.3.4 Concluding remarks ...... 74 3.4: NASDAQ Regulatory Changes...... 74 3.4.1 NASDAQ Listing Rules – Before The Pilot Program ...... 75 3.4.2 The Pilot Program ...... 76 3.4.3 Hypotheses ...... 77 3.4.4 Reverse splits and the Moratorium...... 78 3.4.5 Delistings and the Moratorium...... 79

vi CHAPTER 4: CONCLUSIONS...... 81 4.1 Overview ...... 81 4.2 Results and Suggestions for Further Research...... 81 4.2.1 Stock Price Reactions – Announcement Date...... 81 4.2.2 Stock Price Reactions – Ex-date...... 82 4.2.3 Stock Price Reactions – Firm Characteristics...... 82 4.2.4 Stock Price Reactions – Long Run...... 83 4.2.5 Market Movements...... 83 4.2.6 Projecting Post Split Performance...... 84 4.2.7 NASDAQ Rule Change ...... 84 4.3 Contribution...... 85

APPENDICES ...... 87

BIBLIOGRAPHY ...... 169

BIOGRAPHICAL SKETCH...... 176

vii LIST OF TABLES

Table 1 Forward and Reverse Stock Splits and Split Ratios 1962-2002 ...... 87 Table 2 Average Stock Price per Year and Yearly Stock Splits 1962-2002 ...... 88 Table 3 Changes in Margin Requirements ...... 90 Table 4 Winners Versus Losers After Reverse Stock Split...... 90 Table 5 Summary of Stock Split Literature and Key Findings ...... 91 Table 6 The Daily Raw Returns of Reverse Stock Splits Around the Announcement Date ...... 96 Table 7 The Daily Market Adjusted Returns of Reverse Stock Splits Around the Announcement Date ...... 96 Table 8 The Daily Market Model Adjusted Returns of Reverse Stock Splits Around the Announcement Date...... 98 Table 9 The Daily Raw Returns of Reverse Stock Splits Around the Ex-date...... 100 Table 10 The Daily Abnormal Returns of Reverse Stock Splits Around the Ex- date - Market Adjusted...... 102 Table 11 The Daily Market Model Adjusted Returns of Reverse Stock Splits Around the Ex-date...... 106 Table 12 Ex-date Returns vs. Announcement Returns ...... 110 Table 13 Number of Reverse Stock Splits by Industry 1962-2003 ...... 111 Table 14 Ex-date Returns by Industry...... 112 Table 15 Returns Based on Reason Given ...... 114 Table 16 The Returns of Reverse Stock Splits by Price...... 115 Table 17 The Returns of Reverse stock splits and the Size of the Firm...... 116 Table 18 The Returns of Reverse stock splits and Consolidation ...... 117 Table 19 Reverse Stock Split Delistings Within 250 Days of Ex-date (1962- 2003)...... 118 Table 20 Average Buy and Hold Returns of Reverse stock splits from 1962 to 2003...... 119 Table 21 Average Market Adjusted Buy and Hold Abnormal Returns of Reverse stock splits from 1962 to 2003...... 121 Table 22 Matched Firm Statistics ...... 123 Table 23 Average Abnormal Returns of Reverse stock splits from 1962 to 2003 Based on Matched Firm...... 124 Table 24 Long Term Returns By Industry...... 125 Table 25 Long Term Returns by Reason Given and Low-Priced Versus High Priced...... 130 Table 26 Long Term Returns By Size, Consolidation and Post Split Price...... 131 Table 27 Monthly Stock Splits Related to Market Movements ...... 132 Table 28 Long Run Returns Regressed on Lagged Market Returns...... 133 Table 29 Regression Analysis of Financial Distress Variables – Definitions ...... 134 Table 30 Correlation Matrix: The Returns of Reverse Stock Splits and Financial Distress Variables...... 135 Table 31 Regression Analysis of Financial Distress Variables – Single Independent Variable...... 137

viii Table 32 Multivariate Regression Reduced Model...... 138 Table 33 Nasdaq Requirements September 2001 ...... 139 Table 34 Low-priced Nasdaq Stocks and Reverse Stock Splits 1998 - 2005...... 140 Table 35 Delistings Around the NASDAQ 2001 Moratorium ...... 142

ix LIST OF FIGURES

Figure 1 Reverse Stock Splits 1962-2002 ...... 144 Figure 2 Split Ratios of Stock Splits ...... 145 Figure 3 Distribution of 250-Day Buy and Hold Returns...... 146 Figure 4 Buy and Hold Returns of Reverse Stock Splits ...... 147 Figure 5 Market Adjusted Returns of Reverse Stock Splits...... 148 Figure 6 Matched Firm Abnormal Returns ...... 149

x ABSTRACT

Stock price reactions to reverse stock splits are examined in relation to various firm characteristics and market movements. It is found that the market reacts stronger on the ex-date than on the announcement date. The market reaction varies by industry and by reason given for the reverse stock split. Firms that split for regulatory reasons perform worse than firms that split for other reasons. In the long run, there is little evidence of a negative abnormal buy and hold abnormal return (BHAR) after a reverse stock split, but there is some evidence BHARs varying by industry. An examination of financial ratios and variables reveals that firms with better sales performance and higher operating-income-to-assets have better ex-date returns. In the long run, firms with lower debt relative to their assets do better after the reverse stock split. Operating income expressed as a percent of assets is also positively related to the 250-day BHARs. The NASDAQ minimum bid price rule change of 2001 is explored to determine if it had an impact on reverse stock splits or delistings. The evidence is mixed. The number of firms delisted, expressed as a percentage of firms that would have been delisted under the old rules, decreased. However, the percentage of reverse stock splits, expressed as a percentage of low priced firms, did not change.

xi CHAPTER 1: INTRODUCTION

1.1 Introduction and Motivation

The wealth effects of reverse splits have been studied previously with little disagreement in the literature. The consensus of Vafeas (2001), Desai and Jain (1997), Han (1995), Peterson and Peterson (1992), etc. is that a reverse stock split is a negative informational event. Given the harmony of the literature, what is the purpose of another study on reverse stock splits? There are three reasons that further study is warranted. First, more data are available. As documented in Table 1 and Figure 1, reverse stock splits have been more frequent in the late 1990s and 2000s than in previous periods. Since 1992 the number of reverse stock splits has almost doubled. This is analogous to the number of smokers doubling after the Surgeon General has proven conclusively that smoking is bad for your health. Research has shown that the market reacts negatively to reverse stock splits, yet the number of firms using this strategy has dramatically increased. The research needs to be updated to determine if the reverse stock split is still followed by a negative market reaction. Perhaps something has changed in the way the market perceives the reverse stock split or perhaps there are valid reasons for a firm to consolidate its shares. Second, the type of firm that is reverse splitting its stock has changed over the years. In previous reverse split studies, most of the sample consisted of established firms that fell into financial difficulty. In recent years, perhaps a consequence of the Internet bubble burst, a new type of reverse splitting firm has emerged.1 In the months preceding the bubble burst, the initial public offering (IPO) market was at its peak. New technology-based and Internet-based firms were in high demand. After the bubble burst, many of these technology and

1This refers to the beginning of the bear market of 2000 and 2001, when the NASDAQ composite fell dramatically. The beginning of this period of reversal is often referred to as the Internet Bubble burst.

1 Internet start-up firms lost a large percentage of their market value and their stock price fell below the minimum bid price required for continued listing. Some of these start-up firms chose to reverse split their stock in order to maintain compliance. Previous studies have examined reverse splits using primarily well established companies. Recent IPOs make up a large segment of this study.2 Third, a number of reverse splits with extraordinary post-reverse split performance have caught the attention of journalists who write for financial publications such as Forbes magazine. Journalists report that some reverse split stocks have experienced price increases of more than 100% in the months following the reverse split.3 They have speculated that a reverse split may not be the negative signal, or negative news event, that it was once thought to be. This study applies the rigors of academic research to the data to determine if these speculations are justified. Given the market rule changes, the changing composition of the sample, and perhaps a different dynamic in the market, results provided in this study may be beneficial to both regulators and corporate managers. The sheer number of reverse stock splits that have been announced during the past few years suggest that reverse stock splits are a strategic corporate decision that many managers now consider. Additionally, the 2001 change in the NASDAQ continued listing requirements that allowed low-priced firms more time to bring their bid price back into compliance (above $1.00) provides a unique opportunity to study the impact of such an event.4 In this dissertation, reverse stock splits are investigated on three fronts. First, the stock price reaction to the reverse stock split is examined. The reaction is examined in relation to certain firm characteristics and the reasons that a firm gives for reverse splitting. Daily abnormal returns around the announcement of

2Based on a regression analysis that revealed that that the number of days since market listing is inversely related to the date of the reverse stock spilt. Coefficient –0.42 and p-value<0.0001. 3David Simmons, Forbes, January 25, 2002, for example. 4This opportunity is not available with NYSE and AMSE listing standard changes because so few stocks are affected.

2 the reverse stock split are explored. The ex-date returns and long-run abnormal returns are also investigated. The methodology is similar to Vafeas (2001), Desai and Jain (1997), Han (1995), and Peterson and Peterson (1992). Second, accounting variables taken from the financial distress literature are tested for their usefulness in forecasting post reverse split performance. There is wide variation in firm performance after a reverse stock split and this analysis seeks to provide insight as to what firm specific variables are most related to post split performance. Third, reverse stock splits are examined in light of the changes in NASDAQ’s continued listing requirements that were introduced in September 2001. These changes granted additional time to low-priced firms to bring their bid price back into compliance (above $1.00). Even more time was granted if the company could maintain certain listing requirements besides minimum bid price. As a result this change, low-priced firms had less incentive to reverse split their stock. This study seeks to discover whether this rule change was effective.

1.2 What Is a Reverse Stock Split?

A forward stock split is the exchange of one share of stock for multiple shares. For example, in a 2-for-1 split, each stockholder receives two shares for each share held, and in a 3-for-1 split, each stockholder receives three shares for each share. When a firm splits its stock, the immediate value of the firm should not be affected. Theoretically, a holder of a $50 share that splits 2-for-1 is left with two shares worth $25 each. Companies may also reverse split stocks. Unlike a forward split, which leaves the shareholder with more shares, the reverse split leaves the shareholder with fewer shares. For example, a 1-for-2 reverse split requires the shareholder to trade two shares for one share of the new stock, and a 10-for-1 reverse split requires the shareholder to relinquish ten shares and receive one share in return. A reduction in the number of shares outstanding increases earnings per share and

3 stock price, but, like the forward stock split, it is a cosmetic change in ownership. Whether a firm forward splits or reverse splits its shares, the market value of the total number of shares, which is the market capitalization, should remain the same. Firms split their stocks at varying ratios. The most common forward split ratio is 2-for-1, and the most common reverse split ratio is 1-for-10, followed closely by 1-for-5. Table 1 presents the split ratios of both forward and reverse splits from 1962 to 2002. It is not shown in Table 1, but when the sample is divided into pre-2001 splits and 2001-2002 splits, the 1-for-10 split increased from approximately 19% to 25% of the total.5 This increase in large reverse splits (consolidation) may be due to the fact that the market as a whole, especially NASDAQ, was in sharp decline and companies used the reverse split to stay in compliance with market listing requirements. The forward split distribution has also changed in recent years. In the pre-2001 sample, the 2-for-1 split is most frequent (45%), but in 2001 and 2002, the smaller 3-for-2 split has about the same frequency as the 2-for-1 split (40%). The average share price per year for the entire market is given in Table 2. It is interesting to note that the average remains in the same general range throughout the years. The range does not seem to rise with inflation. Table 2 also shows the percentage of firms that split and reverse split their stock each year. The number shown is a percentage of firms listed in the CRSP database for that year.

1.2.2 A Brief Overview of the Evidence on Forward and Reverse Stock Splits6

1.2.2.1 Anecdotal

In the aftermath of the Internet Bubble, over 200 firms reverse split their stock. Regardless of the reason given by the company, the popular press usually

5The frequency of 1-for-10 splits was 20% for the entire sample. 6An extensive overview is presented in Chapter 2.

4 portrays the reverse split as a negative news event for the particular firm. This is consistent with the findings of academic researchers who, in general, observe negative abnormal returns following the reverse split. The reverse split is usually understood in the press and in academic research as a signal that bad news is on the horizon. There are some reports, however, that claim a reverse split may not necessarily be a negative event for a company. For example, Forbes magazine reports on “reverse split losers” and “reverse split winners,” with double- and triple- digit returns documented for the winners.7 A winner is defined as a firm that has positive returns after the stock split (the ex-date) and a loser is defined as a firm that has negative returns after the split. There are no market or risk adjustments in these returns, and therefore they can be taken only as anecdotal. This anecdotal evidence provided the catalyst for this study. In 2002, Forbes Magazine reported that there were 113 reverse stock splits in 2001 and 84 of these stocks were still trading in January 2002.8 Six companies were acquired, four went out of business, and six split to facilitate transactions such as spin-offs or special dividends. Contrary to the popular belief that reverse split stocks wind up trading on the pink sheets, the over the counter (OTC) bulletin board, or worse, about three quarters of 2001 reverse splits maintained a post-split price above $1 and remained listed.9 The fact that the reverse split stocks performed so well in the middle of a bear market is even more surprising. According to the conclusions of previous researchers, most reverse splits should have negative abnormal returns. This, coupled with the systematic decline in market prices, makes it interesting that such a large percentage of these firms had positive price increases. Of the 84 still trading on January 2002, the winners (39) were up an average of 66% and the losers (45) were down an average of 84%.

7Forbes, January 25, 2002 David Simmons 8 CRSP data revealed that there were actually 127 reverse stock splits in 2001 and 95 in 2000 with 102 still trading on January 31, 2001 9Popular belief means the consensus of previous researchers who show that reverse split stocks in general are followed by negative abnormal returns and then suffer delisting or bankruptcy. This is expanded and fully documented in the literature review in Chapter 2. The popular press also takes a dim view of the future of firms that reverse split. This, too, is expanded in Chapter 2.

5 Even among the losers, 22 (about half) maintained the $1 listing requirement. Internet-based firms, or “dot.coms,” made up about one fourth of the 2001 reverse splits. Of the 23 dot.com companies, the winners (10) were up an average of 79% and the losers (13) were down an average of 63%.10 A more comprehensive record of winners and losers is given in Table 4.11 Industry seems to have played a role in how the media covered some of these firms. Does industry play a role in post split firm performance? Is there any pre-split information like industry or financial ratios that is predictive of post-split performance? These questions are explored in this dissertation.

1.2.2.2 Forward Split Empirical Findings

On average, firms that split their stock have been characterized by positive abnormal returns following the stock split. This is curious because the stock split event adds no value to the firm; it is merely a cosmetic adjustment. Even if the stock split is viewed as a positive signaling event, the market should respond to the good news on the stock split announcement day rather than on the ex-date. If there were information leakage from the company about the upcoming split, then there should be some upward price drift leading up to the announcement day. Researchers find that forward stock splits are associated with a small pre- announcement price drift, a positive reaction on announcement day, and short- term and long-term positive abnormal returns. Researchers attribute the positive abnormal returns to either a sluggish response to the positive information (under- reaction) or asymmetric information. It may be that managers signal their belief in strong performance in the future by splitting the stock, but the market is skeptical and reacts slowly, waiting for confirmation. Other researchers contend that the forward split is the result of a price preference by managers, investors, and institutions.

10 Forbes, January 25, 2002, referencing Thomson Financial/First. 11 This table is referenced again in chapter 3.

6 1.2.2.3 Reverse Split Empirical Findings

Researchers have generally found that when a firm reverse splits its stock, the stock will suffer negative abnormal returns in both the short and long terms. A reverse split is often viewed as a last-ditch effort to artificially maintain the firm’s market listing or to delay inevitable bankruptcy. The price of the stock is not the cause of bankruptcy, but bankruptcy follows many reverse stock splits. This may be due to the firm’s losing market value because of increased idiosyncratic risk. The riskier the firm becomes, the lower its price becomes. Researchers provided empirical evidence indicating that, in general, negative abnormal returns follow reverse splits. As with forward stock splits, the market reaction to the reverse split is only partial at the time of the announcement. Perhaps investors are hoping for a turnaround, or perhaps the coverage by the press and stock analysts is so light on these thinly traded companies that it takes a while for the bad news to be disseminated into the market.12 Researchers have observed negative abnormal returns following reverse stock splits, but the anecdotal evidence suggests that this paradigm may need updating. Perhaps market reaction to the reverse split is conditioned on other information that is available to investors.13 It is possible that the market response has become more sophisticated in that corporate information is more readily available and more widely disbursed that in the past. Readily available information may lead to more firm specific market responses.

1.3 Why Reverse Split?

In theory, the total value of equity of the firm should be the same pre-split and post-split.14 If there are no restrictions on companies with regard to stock price and if the stock price does not affect investors’ transaction costs, the market

12Daniel, Hirshleifer, and Subrahmanyam (1988) model how analysts might overweight their own priors when valuing firms and thus underweight new information such as split announcements. 13Stocks that reverse split are sometimes delisted shortly thereafter. For example, PlanetRx.com did an 8-for-1 reverse split on December 14, 2000, and was delisted two weeks later. Eglobe did a 47-for-10 split on November 13, 2000, and was delisted nine days later. Why did these companies incur the expense of the reverse split? 14The market value of equity is shares outstanding times share price.

7 should be indifferent when a firm splits its stock, either forward or backward. The market, however, does not seem to view the split as a neutral-wealth event. In fact, researchers show that the market treats the reverse stock split as a negative event with both short-term and long-term negative abnormal returns. If reverse splits result in negative abnormal returns, why would a company choose to reverse split its stock? There are numerous reasons cited by companies to reverse split their stock. These reasons include (1) complying with listing requirements, (2) reducing the transaction costs of investors, (3) making the stock marginable, and (4) making the stock appear more reputable to investors by raising the share price of the stock.

1.3.1 Listing Requirements

The primary listing requirement of interest in this study is the minimum bid price for continued listing. The continued listing requirements are less stringent than the initial listing requirement. Each market has slightly different requirements for initial and continued listing, yet some requirements are the same. For example, the NYSE, the AMEX, and NASDAQ all have a $1 minimum bid price for continued listing.15 If a firm’s bid price drops below $1 for an extended period, the firm will be expelled from the market and no longer eligible for trade in that particular market. For example, both NASDAQ and the NYSE begin the delisting process when a firm trades for less than $1 for 30 consecutive days. However, the markets differ on initial bid price, market capitalization, and other qualitative public interest criteria. Firms not meeting the minimum bid price requirement may choose to

15All current requirements for initial listing and continued listing are tabulated in Table 33.

8 reverse split their stock.16 An interesting event to study is the changing of the rules. How do regulatory decisions affect the firms that they govern? Prior to 1997, NASDAQ had no minimum bid price, but in 1997 it saw the need to improve its image by divesting itself of low-priced stocks that were, as a group, highly speculative and prone to fraud. In 1997, NASDAQ first instituted its minimum bid price of $1 per share for continued listing. If a firm’s stock price fell below the minimum, the company then had 90 days to improve the price of the stock, either by reverse splitting or by taking actions to improve its market value. NASDAQ had a recent rule change involving grace periods granted to minimum bid price violators, but the NYSE and the AMEX had no such changes related to bid price or to any other criteria of interest to low-priced stocks. In addition, the vast majority of the stocks in the later sample are NASDAQ stocks. For example, in 2001 and 2002, there were no AMEX reverse splits and only five NYSE reverse splits per year. In contrast, there were more than 100 reverse splits on NASDAQ in each of these two years. Given the limited sample size and the stability of bid price requirements in the NYSE and the AMEX, the focus of this dissertation regarding rule changes will be on NASDAQ stocks and the effects of these rule changes on reverse splits and low-priced stocks in general. On September 27, 2001, NASDAQ issued a moratorium on the enforcement of the $1 minimum bid price rule in response to the depressed stock prices in the overall market. NASDAQ offered longer grace periods—180 rather than 90 days—for firms to bring their company into compliance. If criteria other than bid price were in order but the bid price was too low, even longer grace periods were granted. In January 2002, NASDAQ instituted a pilot program in which the moratorium was extended for two years. Companies in both the National and Small Cap Markets now had extra time to address bid price deficiencies. The program was slated to expire in December 2003 but was extended to December 2004. The pilot program was slightly different for the two

16The AMSE section 970 says, “The Exchange may recommend to the management of a company, whose common stock sells at a low price per share for a substantial period of time, that it submit to its shareholders a proposal providing for a combination (’reverse split’) of such shares.”

9 NASDAQ markets. The National market issuers had 180 days (formerly 90 days) to bring their stock’s bid price back into compliance. At the end of this time the firm could move to the Small Caps market if it wanted to stay listed. Firms in the Small Caps market were given 180 days to comply (formerly 90 days) and were given an additional 180 days (formerly 0 days) if certain other market capitalization criteria are met. The Small Caps were then given an additional 90 days to comply if they met certain non-price criteria at the end of the second 180-day period. These criteria (continued listing requirements) are shown in Table 33 and are discussed in Chapter 3.

1.3.2 Transaction Costs

Researchers have shown that the higher price range after a reverse stock split results in lower percentage bid-ask spreads, and thus the transaction costs are lowered. The lower bid-ask spread is probably due to the increased liquidity of the stock at its post reverse split price. However, a possible lost tax savings after a reverse stock split can also be considered a “transaction cost.” Volatility decreases after a reverse stock split, and firms (and investors) are less able to take advantage of tax laws that allow them to expense a paper loss.

1.3.3 Marginability

A stock is purchased on margin when the investor uses someone else’s money to fund a portion of his investment. An investor may borrow up to 50% of the funds for an initial investment. This requirement has fluctuated over the years from 10% to 50%, as shown in Table 3, but has remained fixed since 1974. Brokerage firms will monitor portfolios and “call” a margin if the equity that is less than fully funded drops in value. The investor must either divest or “cover” the

10 margin by sending additional funds to the broker. A stock must have a bid price of $5 or more to be considered marginable.17 Increasing the price of a stock through a reverse split allows more investors to use margin accounts to purchase the security. The stock is more marketable because there are more potential buyers. Institutional investors are also wary of low-priced stocks and generally avoid those that are unmarginable. When firms reverse split their shares, the resulting price is often greater than $5. Why would a firm reverse split to a price that is not marginable (i.e., less than $5)? Because, a low bid price is not the only criterion that a firm is struggling to meet in order to remain listed on an exchange. There are also minimum market valuations and minimum shares outstanding requirements. If a firm does not have enough shares outstanding to reverse split to over $5, then marginability is not a likely motive for the reverse split. Closely associated with marginability is institutional interest. For example, Judy Bruner, CFO of Palm, Inc., cites the loss of institutional investors as one reason for its 1-for-20 reverse split in October 2002.18

1.3.4 Signaling

Empirical studies on reverse stock splits are in agreement that, on average, reverse splits reduce the value of the splitting firm; abnormal negative returns tend to follow the reverse split. This observation suggests that reverse splitting conveys negative information about the firm. However, as mentioned earlier, some companies have recently experienced positive share price increases in the months after a reverse split. This observation suggests that the reverse split event is not always negative and that the market reaction to a reverse split may instead be driven by other factors.19 Perhaps the returns associated with a reverse split are

17Margin requirement rules, Securities Exchange Act of 1934. 18Quoted to Tim Reason, Reverse Psychology Today, CFO, Magazine for Senior Financial Executives, December 2002. 19Price increases are not necessarily indicative of positive abnormal returns, but this is interesting because, historically, decreasing prices follow reverse splits. This issue is addressed formally and

11 due to the information conveyed by a reverse split, conditional upon information about the motivation for the split or the financial health of the splitting firm. This study adds to the reverse split literature by showing how accounting variables or firm characteristics affect the signal of a reverse stock spilt.

1.4 Hypotheses

Hypotheses are developed and tested in this dissertation to explore the stock price reactions to reverse stock splits. The stock price reactions are examined in the context of the market environment, regulatory influences, the Internet Bubble burst, and the health of the stock as related to reverse stock splits.

1.4.1 Stock Price Reactions

The first hypothesis tested relates to the stock price reactions associated with reverse stock splits. Researchers have documented abnormal negative returns around the announcement date and negative abnormal returns around the ex-date. The increase in reverse splits during the late 1990s and early 2000s is shown in Figure 1. This wealth of new data provides an enticing reason to revisit this topic. Using the expanded data set (compared to previous studies), the abnormal returns surrounding the ex-date and the announcement date are examined. The primary hypothesis of this study, stated in the null, is:

There is no stock price reaction associated with the event of a reverse stock split. The typical company performing a reverse stock split is generally smaller and less well known and trading in its stock may be quite thin. It may be the case, then, that the effect on shareholders’ wealth is not confined to the event day, but

documented thoroughly in Chapter 3, in which abnormal returns following the reverse split are examined.

12 rather is spread out over a longer period as more information about the company is revealed. Announcement day effects are examined in an 11-day window around the announcement. Ex-date effects and long run effects are also examined. The first wealth effect is examined specifically with the following two hypotheses:

There is no short-term stock price reaction associated with the event of a reverse stock split.

There is no long-term stock price reaction associated with the event of a reverse stock split. One facet of this study deals with the regulatory environment and one of the question of interest is whether the reverse splits were motivated by the threat of delisting. If there is a difference in the returns for firms that have been forced to reverse split their stock versus firms that have voluntarily reverse split, then it is possible that the exchange regulations may be having effects in the market. The hypothesis, stated in the null, is then:

There is no difference in the stock price reaction on the stock of firms which reverse split their stock for voluntary versus regulatory reasons. Closely related to the wealth effects is the enormous loss in the value of NASDAQ composite stock index during 2000 and 2001. This loss of equity value in technology stocks has been referred to as the Internet Bubble burst. Laying aside the arguments concerning whether or not there was an Internet Bubble, the systematic loss of equity value during this period is given as the motivation for exchange rule changes and other regulatory changes. Did the systematic loss of equity value contribute to a firm’s decision to reverse split its stock? Firms that may never have considered reverse splitting their stocks may have been caught in the systematic downdraft of equity prices and therefore felt obligated to prop up their otherwise healthy company’s stock price. They may have felt that the price was too low due to exchange requirements, trading range goals, liquidity concerns, or other market considerations.

13 The next hypothesis, stated in the null, is:

The frequency of reverse stock splits is not related to prior market performance. A casual observation of the yearly reverse stock splits (Table 1) reveals no obvious pattern. When prices drop in times of market decline, it seems that more firms would reverse split their stock. This idea is tested. Perhaps a relationship exists that is not apparent from casual observation.

1.4.2 Financial Health

One of the motivations for the third hypothesis, is to determine if a distinction can be made between good firms that may have been caught in the downdraft of the 2000-2001 bull market versus firms that were destined for failure regardless. Some firms had little chance of survival, while other firms were simply caught in the wake of the NASDAQ plummet. The irrational exuberance (Alan Greenspan’s assessment) of the time preceding the plummet may have fueled the public’s hunger for IPOs of dot-com companies that were not quite strong enough to go public.20 The public’s demand for more and more tech companies may have led to an environment in which sound business valuation and common sense were replaced by a gambler’s mentality. IPOs in the high-tech industries, especially Internet-related firms, seemed extremely popular to investors. This section of the dissertation borrows variables from the financial distress literature and seeks to determine if the same variables that predict financial distress can be used to forecast a firm’s performance after a reverse stock split. If the distress variables for a firm do not indicate distress, then the bid price deficiency may be the result of getting caught up in a public panic to sell. For example, if sales were still strong

20Federal Reserve Board Chairman Alan Greenspan made these remarks in his speech at the Annual Dinner and Francis Boyer Lecture of The American Enterprise Institute for Public Policy Research in Washington, D.C., on December 5, 1996. The catchphrase later became the title of a 2000 book about the overpriced market. Irrational Exuberance, by Robert J. Shiller, won the 2000 Commonfund Prize for Best Contribution to Endowment Management Research. Shiller is the Stanley B. Resor Professor of Economics at Yale University.

14 and a business had a full pipeline of orders for the next several quarters, then perhaps the market’s valuation of the equity was unjust. If a firm in this situation reverse splits its stock, perhaps it sends a false signal to the market. Did the market see through the false signal? The next hypothesis, stated in the null, is:

The financial health of a stock does not help to predict firm’s returns after a reverse stock split.

1.4.3 Listing Requirements

The stock markets are interested in preserving the integrity and reputation of their exchanges. The markets have rules that specify the minimum bid price that a firm must maintain to remain listed on the exchange. The NYSE, the AMSE, and NASDAQ created these rules to enhance the reputation of the markets because low-priced stocks have been associated with price manipulation and stock scams. In 2000 and 2001, NASDAQ relaxed its minimum bid requirements and granted automatic extensions for firms that perhaps were caught in the downdraft of the bear market. The stated objective was to give the good firms adequate time to recover. Did the rule relaxation increase the probability that a low-priced stock would recover, or did the rule changes simply delay the inevitable bankruptcy or delisting? Were firms less likely to reverse split their shares in the aftermath of the rule changes? The next hypotheses, stated in the null, are:

The moratorium on the minimum bid price rules had no effect on whether or not a NASDAQ firm reverse split its stock.

The moratorium on minimum bid price rules did not result in fewer delistings for NASDAQ firms whose stock had a bid price of less than $1.00

15 1.5 Outline of This Dissertation

The remainder of this dissertation consists of a literature review followed by development and testing of the hypotheses. Chapter 2, the reverse split literature review, is followed by Chapter 3, in which the hypotheses are motivated and tested. Chapter 4 summarizes this research and its constraints while suggesting extensions for further research.

16 CHAPTER 2: LITERATURE REVIEW

2.1 Why Does a Firm Split Its Stock?

Stock splits present an enigma to market observers. A stock split produces no material change in a firm, yet through the years the stock split has remained a popular fixture in the U.S. financial markets. For example, in 1930 almost 20% of the firms listed on the NYSE had split their shares in the prior decade (Conroy & Harris, 1999). In modern times, stock splits still endure, with 5-10% of NYSE firms announcing a split each year. One of the results of the incessant stock split is that average stock prices in U.S. financial markets have remained remarkably constant over time. Over the last half-century, the average NYSE share price has remained in the $30-$40 range regardless of rising inflation, consumer prices, or firm equity value (Conroy & Harris, 1999). A stock split is, at one level, only a cosmetic change; the same-size pie is merely sliced into smaller pieces.21 The investor retains the same partial ownership of the company’s equity and the same voting power as before the split. Yet, in spite of this seemingly cosmetic change, empirical researchers continue to demonstrate that stock splits have real effects in the financial markets. In some ways, the effects of forward splits seem beneficial: equity value increases, trading volume increases, and the number of shareholders increases. In other ways, the effects of forward splits seem less desirable: shareholder risks increase, and some transaction costs, such as bid-ask spreads, are higher after forward splits.22 The reverse splits mirror the forward splits. Value decreases, volume decreases, the number of shareholders decreases, shareholder risks decrease, and transaction costs decrease.

21"You'd better make it four; I don't think I can eat six pieces."—Yogi Berra, when asked if he wanted his pizza cut into four or six pieces (http://www.brainyquote.com/quotes/quotes/y/yogiberra141844.html). 22Even if the bid-ask spread is the same pre- and post-split, the lower price of the stock makes the bid-ask spread a higher percentage of the stock price.

17 In an attempt to explain these varied benefits and consequences of stock splitting, the literature presents several reasons that a firm may reverse split its stock. These reasons include asymmetric information, moving the stock’s price to a more desirable trading range, exchange regulations, margin regulations, institutional limitations, reorganization, privatization, and liquidation. Key findings of major studies are presented in Table 5. Even though the literature presents several hypothetical reasons that a firm might split its stock, the truth is that no one knows the true motivations except the managers who actually make the decision to split. With this in mind, researchers conduct surveys. Unfortunately, the managers who are surveyed are not always forthcoming in their explanations. In a 1977 study of reverse splits, nearly half of the firms that reverse split their stock offered no public explanation at all for the split.23 Of the firms that made public announcements, the most common reason given was image improvement. This was followed by exchange requirements, then appealing to a different type of investor, and finally, reduction in shareholder service costs (Gillespie & Seitz, 1977). In a similar survey on forward splits, 70% of managers who responded stated that moving the stock to a preferred price range or improving liquidity was the reason for the stock split. Signaling was the primary motivation of only 14% of the responding managers (Baker & Powell, 1993). Studies of reverse splits in the late 1970s and early 1980s find that there is a small increase in volume and a decrease in the number of shareholders after the split, but the institutional interest remains unchanged. The studies conclude that reverse splits are not justified because they are, in general, detrimental to shareholder wealth. Moreover, reverse stock splits are not justified by the reasons cited by managerial surveys because there are no tangible benefits related to those reasons. In addition, Radcliffe and Gillespie (1979) studied only firms that survived their test period, so it is likely that they understated the negative effects of the reverse splits on shareholder wealth (Woolridge & Chambers, 1983).

23Gillespie and Seitz (1977) found that only 13 of 24 firms gave a public reason for the reverse split.

18 Another negative effect of the reverse split is the decrease of the tax option value of stocks. The investor has more possible tax alternatives when stock is more volatile. High volatility allows investors to take advantage of U.S. tax laws by offsetting capital gains with capital losses. Reverse splits decrease volatility and therefore decrease the tax option value. Forward splits, on the other hand, increase volatility (Lemoureux & Poon, 1987). Although researchers have shown that reverse splits are detrimental to shareholder wealth, firms continue to engage in the practice of reverse splitting their stocks. The stock split literature gives a variety of explanations for forward and reverse stock splits, and these are discussed in the following sections. The first explanation considered is asymmetric information, or signaling.

2.1.1 Asymmetric Information (Signaling)

The asymmetric information hypothesis is that managers will split the firm’s stock to send a signal to investors that conveys private or inside information about the firm. For some reason, the firm is unable to disclose information such as trade secrets, a revolutionary breakthrough, or some other secret competitive advantage. The managers cannot reveal their private information, but they feel that if their secrets were widely known, the company’s market value would improve. A stock split is a way to signal investors that the managers believe that the company is undervalued and that there is a positive expectation about the future. According to this theory, the main purpose of stock splits is to provide private information to the market. The positive abnormal returns that follow stock splits are often attributed to this signaling effect. A forward split is a sign of strength and confidence that the share price will continue to increase. A reverse split, on the other hand, is a sign of weakness and could be viewed as the only remaining option to stay listed on an exchange or as a means of complying with creditor demands. It is a negative signal to the market. Succinctly, the signaling view is that management can influence market perceptions of their firm by changing its share price.

19 Various researchers argue that forward stock splits are a favorable signal from management about future firm performance. They argue that managers of undervalued firms will split the firm’s shares to signal positive private information about the company. Even if the managers have no information to convey, the researchers suggest that managers are seeking to attract attention to the company in hopes that investors and analysts will notice the company is undervalued. The split is a costly signal to the market that a weaker competitor would not likely imitate. It is a show of strength by the corporation and a disclosure by management indicating they think the firm is more valuable than the current market price dictates. Reverse splits, however, have been shown to result in negative abnormal returns for the firm and are seen as a negative signal to the market (e.g., Brennan & Copeland, 1988; Brennan and Hughes, 1991; Conroy & Harris, 1999; Grinblatt, Masulis, & Titman, 1984; Ikenberry, Rankin, & Stice, 1996; Leland & Pyle, 1977; Ross, 1977). Some studies have concluded that even if managers are trying to signal the market, there is no evidence (or mixed evidence) supporting a change in information asymmetry after the split announcement. However, even if the split itself is not a signal, the evidence does seem to support the notion that the size of the split factor signals information (Brennan & Copeland, 1988; Brennan & Hughes, 1991). The studies discussed above conclude that forward stock splits convey positive information about a firm to the market, but more recently, researchers have extended this line of reasoning by examining whether the positive signal conveyed by the stock split announcement has an effect on companies in the same industry that did not have a stock split. They find that, on average, the firms that did not split had significant positive abnormal returns following the split announcement of their intra-industry counterpart (Tawatnuntachai & D’Mello, 2002). This finding seems to strengthen the signaling theory.

20 2.1.2 Signaling Combined with Transaction Costs

Theories that combine signaling with transaction costs offer further insight into stock splits. The view is that for a signal to be credible, it must be costly. One source of increased cost is the higher transaction costs of lower-priced shares after a forward split. (Brennan & Copeland, 1988). Transaction costs are reduced for reverse split stocks (Han, 1995). The primary transaction cost is the bid-ask spread, which is proportionally higher per share after the forward split and proportionally lower per share after the reverse split. The empirical findings in the 1990s support the view that investors respond to costly signals. Investors respond to signaling even more favorably when the firm is small. Specifically, returns to forward split announcements are negatively correlated with firm size, and the post-split price is positively correlated with the size of the split factor. The costly signaling explanation is that managers will not forward split unless they have exceptionally good information about the future of the firm (Ikenberry et al., 1996; McNichols & Dravid, 1990; Pilotte & Manuel, 1996). When surveyed, most managers deny the signaling aspect of the split and will respond that their primary goal is to move the stock price to a more favorable trading range, where the liquidity is improved and the price is more attractive to investors (Baker & Gallagher, 1980; Baker & Powell, 1993). Trading range manipulation and liquidity are discussed next.

2.1.3 Trading Range

What do executives say when asked why they split their stock? The answer seems to be consistent over time. A survey reveals that from 1900 to 1930, 90% of managers of stock-splitting firms cited a wider distribution in shares as their primary motive (Dolley, 1933). Presumably, the lower price per share not only increases the number of shares outstanding, it also facilitates trading. During the 1940s and 1950s, managers responding to a survey indicated that they wanted to

21 bring their share price down to the $15-$40 range (Dewing, 1953). Of the managers responding to a survey in the late 1980s (1987-1990), more than 70% cited a preferred price range as their primary motivation for splitting, and 14% cited signaling (Baker & Powell, 1993). Managers responding to a similar survey in the early 1990s reported that their main motivation for splitting is their belief that investors prefer stock prices in the $20-$35 range (Baker, Phillips, & Powell, 1995). Managers want to increase individual investor ownership because they feel that a broad base of ownership increases the value of the firm (Lakonishok & Lev, 1987). Trading range motivation exists when the firm’s management believes that there is a desirable price range in which their stock is more liquid. Firms will split the stocks when the price gets too high and out of range for investors who must purchase shares in round lots.24 Firms will reverse split stocks when the price gets too low. Margin requirements and institutional perception make moving up to a higher trading range look like an attractive option for low-priced stocks. In addition to perceptions, it has also been shown that traders actually have lower transaction costs after a reverse split (Peterson & Peterson, 1992; West & Brouilette, 1970). Uninformed trades, proxied by non-institutional trades, increase after a forward split, and these trades are more expensive to the trader (Easley, O’Hara, & Saar, 2001). Easley et al. interpreted their findings to be consistent with the trading range hypothesis since the number of small holdings increases—a stated goal of managers citing trading range as their motivation for splitting stock. An alternate explanation of their findings is that forward splits move tick sizes relative to the stock price to a desired level (Angel, 1997; Anshuman & Kalay, 2002). The idea is that the proportionately larger tick size may provide market makers (brokers) with an additional incentive to promote the stock to small investors.25 Brokerage firms

24A round lot is 100 shares. 25 This implies that stocks can be marketed or promoted at higher price levels and that irrational and naive investors can be taken advantage of by salespeople. This goes against economic theory that says in aggregate investors are rational and all-knowing. If promoters of consumer goods such as automobiles can benefit from promotional campaigns, why can’t promoters of equity offerings? This is beyond the scope of this study but an interesting study for later.

22 may encourage the split to preserve commission income (Brennan & Hughes, 1991). Supporting research demonstrates that there are indeed more small orders than large orders after a forward split, and the bulk of the orders are buys (Schultz, 2000). This is consistent with the trading range hypothesis and the view that forward splits act to promote stock trading among small individual investors. It is also consistent with studies that show increasing numbers of shareholders after forward stock splits (Lamoureux & Poon, 1987; Maloney & Mulherin, 1992). In addition, it has been shown that forward stock splits are followed by increased trading activity by small investors (Angel, Brooks, & Mathew, 2004). This is a desirable outcome because when there is incomplete information in a market, a company may have lower capital costs as the number of unique shareholders increases (Merton, 1987). The demand for the stock is higher; thus, the price is higher and returns are lower. The trading range motivation for forward stock splits is well established in the literature, but trading range is a motive for reverse stock splits as well (Peterson & Peterson, 1992). Forward and reverse splits are both used to place a stock in a more attractive trading range. For reverse splits, this generally results in lower transaction costs for the traders (Peterson & Peterson, 1992). Marketability is another motive for reverse splits. Low-priced securities can be seen as speculative or disreputable, but if a firm reverse splits its stock, the price is higher and the view is that the stock is more appealing or more available to the mass market. These marketing issues are important because companies desire to have their stock purchased by institutional investors. Institutional investors may have trouble justifying a portfolio with low-priced stocks. A stock is viewed as more marketable if it is listed on a major market, and NASDAQ and the AMEX do not allow low-priced stocks to persevere. In fact, a company is encouraged to reverse split if the stock trades at a low price for an extended period. The NYSE does not have a minimum share price for continued listing, but it can delist companies in the case of unusually low share prices, typically less than $1 for more than 30 days.

23 Low-priced securities are often seen as speculative or disreputable and cannot be bought with a margin account. Thus, companies have incentive to raise the price via a reverse stock split so that the stock is available to a broader base of individual investors at a more favorable trading price (Spudeck & Moyer, 1985). Research on why companies choose certain stock split ratios shows that the ratio is chosen to bring the equity price in line with the average market share price and, to a lesser extent, the industry-wide average share price (Lakonoshok & Lev, 1997). This supports the trading range theory. Lakonoshok and Lev have also presented evidence that forward splits are confirmations of past activity rather than signals of future prospects. Companies that forward split their stock are often companies that have recent abnormal growth in earnings and dividends. The high growth does not persist after the forward split. This is a strike against the signaling theory. Some researchers have suggested that the signaling theory is valid for stock splits because the empirical properties of stock dividends seem to support it. If stock splits and stock dividends are the same corporate event, except on different scales, this is a valid argument. Lakonoshok and Lev argued that empirical properties of stock splits and the empirical properties of stock dividends are fundamentally different, and therefore it is inappropriate to draw conclusions about splits based on the properties of dividends.

2.1.4 Institutional Reasons to Reverse Split

Some firms desire to appease institutional investors because they want to attract institutional capital. If a company has a low share price, institutional investors may be more prone to purchase a stock after a reverse stock split. The fact that institutional investors shy away from low-priced stocks gives firms a motive to reverse split their stock. The institutions are bound by covenants with their investors and often these covenants specify a minimum bid price.26 In

26Vincent Sbarra, a senior partner with HBC Capital, says, “A lot of investment funds have covenants that don't let them buy stocks under certain prices—usually $3 or $5.” Ulrico Font, senior analyst for Ned Davis Research, adds, “Everybody feels most comfortable with stocks priced

24 addition, institutions have to justify their investments and investors are more likely to second-guess the fund managers when a fund includes low-priced stocks that underperform (Lakonishok, Shleifer, & Vishny, 1992).

2.1.5 Margin Regulations

“Buying on margin” occurs when an investor borrows money from the broker to buy a security. The Securities Exchange Act of 1934 regulates margin accounts, but each brokerage firm has the power to place additional constraints on their customers. Margin requirements dictate the minimum share price and percentage of investment that can be borrowed from the brokerage firm. Currently, a customer can borrow up to 50% of the price of a new investment for securities that trade for $5 or more. Some brokerage firms keep a list of “unmarginable” securities. The brokerage firm has decided, for whatever reason, that the firms on this list are too risky to be bought on margin. Margin regulations can make a stock difficult to obtain for individual investors; thus, a company with low-priced shares may choose to increase the price through a reverse split. This allows more access to the security through margin accounts, resulting in a wider base of potential investors. However, bid price is not the only consideration that the brokers or the Securities and Exchange Commission (SEC) use to disqualify a stock from being “marginable”; therefore, a reverse split may not always be productive in this regard. Factors such as trading volume, public interest (percentage of shares outstanding), volatility, and perceived risk (e.g., IPOs) may weigh into a broker’s decision to designate a stock unmarginable. Margin regulations have changed over the years. In 1958, margin requirements were changed three times. They were reduced from 70% to 50% in January, raised back to 70% in August, and raised again to 90% in October. From 1960 to 1974, the margin requirements changed eight more times, fluctuating

between $5 and $50.” AT&T and Palm give similar reasons for reverse splitting (Tim Reason, CFO, Magazine for Senior Financial Executives, December 2002).

25 between 50% and 80%. The last change, in 1974, set the margin requirement at 50%, and this rule has not been changed since then.27 Generally, to be marginable, a stock cannot be a penny stock. In 1990, the SEC attempted to curb penny stock fraud and abuse by issuing the 1990 Securities Enforcement Remedies and Penny Stock Reform Act (PSRA). The centerpiece of this legislation is the severe restrictions placed on IPOs that are priced below $5 and not traded on major exchanges or a national automated quotation system. The reasoning is that low-priced IPOs are low-quality IPOs. The regulation succeeded in reducing the number of IPOs priced below $5, but it did not succeed in improving the overall quality of IPOs. Studies find that delisting risk, a proxy for IPO quality, did not decline significantly after the passage of the PSRA. As evidenced by the decline in abnormal returns earned by a portfolio of post- PSRA IPOs, the regulation had the unintended consequence of drawing speculative (i.e., more risky) IPOs in the higher price range (Beatty & Kadiyala, 2004).

2.1.6 Reorganization

In addition to the “marketing” considerations discussed previously, there could be a few other reasons that a firm chooses to reverse split its stock. When a company declares bankruptcy, creditors could dictate the split as a part of the reorganization. Reverse stock splits could also be used to reduce the number of shareholders so that the corporation can be privatized and avoid full disclosure requirements. Reverse splits can squeeze out minority shareholders if the managers desire to take a public company private. Some states will not allow investors to own fractional shares. Therefore, the reverse split forces minority shareholders to sell. In addition, reducing the number of shareholders reduces the cost of servicing those shareholders.

27This is from “Chronology of Significant Events,” California Department of Finance.

26 2.1.7 Liquidity

Han (1995) showed that the liquidity of equity is improved after a reverse stock split. He proxied liquidity with the bid-ask spread, trading volume, and the number of non-trading days. He compared split firms with control firms and concluded that for the reverse split stocks, the spread is reduced, the volume is increased, and there are fewer days during which the stock is not traded. In addition, he showed that bid-ask spread decreases and trading volume increases after the reverse split. Han (1995) says that reverse split improves the liquidity of a stock because the increased share price allows investors to buy the stock on margin. Brokers do not allow their customers to buy low-priced shares on margin. Therefore, a reverse split improves liquidity by allowing margin investors access to the stock. Another reason Han (1995) gave is that the increase in share price improves the market’s perception of the stock. As discussed above, institutional investors are more likely to purchase shares that they can justify to their clients. Low-priced stocks are not easily justified. The evidence for liquidity in forward splits is mixed, but one interesting study involving American Depository Rights (ADRs) found that when there is an ADR split announcement, both the underlying stock and the ADR will experience positive price gains (Muscarella & Vetsuypens, 1996). The stock price increases even when there is no stock split announcement in the firm’s home market. The ADR split is accompanied by increased trading activity in the underlying stock. This increased volume is cited as evidence of increased liquidity. Muscarella and Vetsuypens did not test reverse split ADRs.

2.1.8 Risk

Peterson and Peterson (1992) found that the total risk of returns of reverse split stocks declined after the reverse split. Brennan and Copeland (1998a) documented decreasing risk shifts for forward splits. However, Peterson and Peterson showed that systematic risk does not change after reverse splits. In

27 contrast, Brennan and Copeland found that systematic risk decreased around forward splits. Peterson and Peterson (1992) found positive wealth effects for companies that were forced to reverse split. Companies that were not forced to reverse split did not exhibit the same wealth effect. Their research showed that firms were less risky after a reverse split because of decreased nonsystematic risk.

2.1.9 Investor Behavior in Relation to Stock Splits

The idea of behavioral effects has been around since at least 1953, when A. S. Dewing noted that even though arguments for forward and reverse splits are equally sound, managers are reluctant to reverse split. He believed that a reverse split is an admission that the stock has been previously overpriced, and such an admission goes against human inclination since it may imply that the managers are either unethical or incompetent (Dewing, 1953). Thirty-five years later, behaviorists offer a model that shows how analysts, or the market as a whole, may underreact to new information such as a stock split announcement. People tend to overweight their own beliefs and perceptions. While new information may update or change these beliefs and perceptions, people tend to ascribe more weight to their prior beliefs than they do to the new information. For example, people tend to be slow to change their minds about religious and political views that they “know” to be true—even in the face of overwhelming evidence to the contrary (Barberis, Schleifer, & Vishny, 1988; Daniel, Hirshleifer, & Subrahmanyam, 1988). Are managers sensitive to underpricing or overpricing of their equity? Ikenberry, Lakonishok, and Vermaelen (1995, 2000) reported long-horizon return evidence in the United States and, more recently, in Canada. These authors found that, at least for repurchases, managers seem just as sensitive to underpricing, as their counterparts seem sensitive to overpricing. Ritter (1991) reported that managers appeared to be timing the market at a relative peak when initially issuing equity, since subsequent long-horizon abnormal returns were negative.

28 Seasoned equity offerings follow the same pattern as the IPOs (Loughran & Ritter, 1995; Speiss & Affleck-Garaves, 1995). Adding to the evidence of managers being able to determine whether their stock is over- or underpriced are Baker and Wurgler (2000), who reported that the aggregate flows of equity offerings appear to have predictive power for overall market returns. Lakonishok and Vermaelen (1990) examine equity repurchases and find that managers are often sensitive to underpricing. If managers know when their stock is underpriced, this supports the signaling hypothesis of stock splits. These managers split to signal their private information. The self-selection hypothesis is a synthesis of two explanations of the stock split phenomenon. This hypothesis combines the trading range hypothesis and the signaling hypothesis. Prices after a split are at or below the stock price of similar- sized firms. The five-day announcement return confirms prior research that the splits convey favorable information. The market reaction is greater for small firms, low book-to-market firms, and firms splitting to low-share prices. Investors underreact to a split announcement. The inverse relationship between pre-split run-up and post-split returns suggests that momentum is not a factor. The evidence is consistent with the self-selection hypothesis. Market underreacting to corporate events is not unique to stock splits. The market underreacts to many corporate events.

2.1.9.1 Framing

Framing occurs when an investor’s perception of gains and losses are linked to a reference point. For example, an investor may consider $10 to be a fair price for a security. If the investor buys at $8 (a 20% discount) and sells at $9.75, the investor may still feel a sense of loss because the security was “framed” in the mind of the investor to be worth $10. The investor feels that he lost $0.50 even though he in fact gained $1.75. Even if an investor frames the price correctly, losses and gains are not viewed in the same way. The utility (positive feelings) of

29 gains is not symmetric to the negative utility of a loss. Bad feelings about a loss must be offset by a greater gain. For example, when the investor loses $5, it may take a gain of $10-$15 to compensate for the emotional trauma of the loss. Other factors influencing the framing include a person’s habits, what he or she considers normal, and his or her view of the decision maker’s expectations (Kahneman & Tversky, 1979; Shefrin & Statman, 1984). Lewin (1996) discussed, in a broader sense, the links between economics and psychology and argued that investors may use a company’s split history to decipher the information content of the current stock split, even if investors have no particular price preference.

2.1.9.2 Price Preference

Do investors have a price preference? Theoretically, they should not, but public statements by fund managers underline the fund managers’ belief that investors prefer lower-priced stocks (Conroy & Harris, 1999). Designers of financial products recognize the behavioral aspects of investors as they market their products. For example, new securities are often designed to be priced within a certain trading range (Shefrin & Statman, 1993). Perhaps investors do not have a specific price level preference, but evidence suggests that regardless, the firm’s previous post-split performance is useful for interpreting the meaning of the current split. If managers typically split the stock to a certain price level, then post-split prices below that level are seen as an especially positive signal, and the market reacts accordingly (Pilotte & Manuel, 1996). The idea that managers have a preferred price level is supported without dispute in many studies (e.g., Baker & Powell, 1993). Firm-specific price preferences may be the result of a combination of behavioral and microstructure considerations. Market microstructure is discussed in the next section. Evidence further supports the price level preference in that analysts’ earnings forecasts tend to change consistent with the signal sent by a forward or reverse stock split (Conroy & Harris, 1999).

30 2.1.9.3 Herding

Does herding exist in financial markets? One model suggests that investors with little information will abandon their own information and follow other investors who are perceived to have better information, even if the “leader” has, in fact, very little information. The modelers call this conforming behavior. This conforming behavior leads individuals to join the herd or, as the researchers phrase it, “to converge on one course of action” (Bikhchandani, Hirshleifer, & Welch, 1992). Herding also occurs in mutual fund clients. Mutual fund clients often depend on the advice of a single broker for their information. Reliance on this “undiversified source” makes investors especially vulnerable to misinformation (Brennan, 1995). Further herding behavior may be caused by misleading performance information on commodity funds, as reported by the relatively few news outlets that cover them. Researchers have found that investors who are both misinformed and “unsophisticated” exhibit “puzzling behavior,” such as herding, in the marketplace (Elton, Gruber, & Rentzler, 1989).

2.1.10 Optimal Tick Size (Market Microstructure)

As noted in the previous section, evidence suggests that share price changes caused by stock splits are followed by market reactions. One explanation for this market reaction is that the historical prices may capture firm-specific market microstructure factors that have stabilized over time. One of these microstructure factors is tick size. A forward split results in a larger tick size relative to the share price. The bid-ask spread is a larger percentage of the share price. The regrettable higher transaction cost is offset by desirable results such as greater protection for limit orders, lower-cost trader negotiation, and fewer trading errors. The optimal relative tick size depends on various firm characteristics, but if there is an optimal tick size, then a stock split that results in a price below an optimal (liquidity-based) price can be viewed as a signal that conveys positive information to the market. The managers expect the price to rise after the split,

31 back to the optimally liquid price (Angel, 1997). In essence, Angel’s reasoning is an extension of the signaling hypothesis discussed previously. Angel also argued that the idea of an optimal tick size is a contributing factor to the consistency of the average NYSE stock price over the years. Further research based on the examination of intraday trades and quotes provides little support for the tick size hypothesis (Schultz, 1997).

2.2 Conclusion

2.2.1 What We Know

We know that firms continue to split their stock. In general, there are positive abnormal returns after a forward split and negative abnormal returns after a reverse split. Ikenberry et al. (1996) have shown that the market sometimes reacts to corporate events and that firms with low share price or negative pre-split returns have positive announcement returns, but negative returns in the year following the split. This means good news, followed by disappointment. The trading range hypothesis, blended with signaling, seems to be the prevailing theory as to why managers would take the costly action of splitting the firm’s stock. Companies that execute forward splits do so to target a specific price range. When a company splits, the split factor adds strength to the signal. Bigger factors amplify the signal for a split. A large factor in a forward split brings the price down below historic averages and the market expects the stock to increase further. Bigger factors in a reverse split mean that the reverse split is less severe and less negative of a signal. For example a split factor of 0.9 (10 shares to 9) is less severe than a split factor of 0.1 (10 shares to 1). We know that firms may reverse split their stock because of continued listing requirements, and there is evidence that the firms that split due to listing requirements outperform firms that reverse split for other reasons.

32 We also know that institutional preferences factor into the firm’s stock- splitting decision. Institutions shy away from low-priced firms and they seem to have a preferred trading range; thus, firms will forward or reverse split their stock to cater to these preferences.

2.2.2 What We Do Not Know

The most recent literature on stock splits concedes that stock splits are still one of the least understood phenomena in equity markets (Easley et al., 2001). Over the years, many attempts have been made to explain why a rational firm would choose to split its stock. The drawbacks are numerous. Stock splits are not costless to implement. For what seems to be a non-economic event, the company will pay underwriting costs, accounting costs, printing costs, postage, and other expenses associated with performing what appears to be a purely cosmetic operation. In addition, since exchange fees are based on the number of shares outstanding, a company must pay higher exchange fees when it issues a stock split (lower fees for a reverse split). There seems to be no financial reason for a firm to bear these costs. After all, there is no upward bound for stock prices. Berkshire Hathaway, for instance, is a firm that has chosen to forego stock splits. The “A” shares cost well over $10,000 per share and the CEO explains that the company is looking for a “certain type of investor,” implying that investors have stock price preferences and high per-share prices will drive away unsophisticated investors. Besides costs to the firm, empirical research has documented other negative effects of forward splits, such as greater volatility, larger proportional bid- ask spreads, and higher transaction costs. Reverse splits result in decreased volatility, increased liquidity, smaller proportional spreads, lower exchange fees, and lower transaction costs, so it may make more economic sense to reverse split than to forward split. However, reverse splits, even with all of the positive residuals, are historically followed by negative abnormal returns, or worse— delisting or bankruptcy.

33 Although there are many drawbacks, and although many studies have shown that forward splits are inefficient and reverse splits are detrimental to corporate wealth, firms continue to announce stock splits. One of the common explanations for a stock split is that it sends positive information to the market. The short-interest studies seem to confirm this for forward splits, but what happens to short interest following reverse splits? Is the reverse split a negative signal? Most firms that reverse split have prices too low to be shorted and low-priced securities in general have lower-option interest, so it is difficult to understand the impact of a secondary signal such as short interest or put interest. We still do not have a unifying theory as to why corporations split their stock. When the management of a firm chooses to split its stock, there can be only two logical reasons for the split: either the split is in the manager’s best interest or it is in the shareholder’s best interest.

2.2.2.1 Managers’ Self-interest

Perhaps the managers are under pressure from irrational shareholders to split the stock. Something about stock splits is exciting to shareholders. They have a sense of gain when they have more shares. Perhaps the managers think that an increased ownership base after a forward split will confound a hostile takeover and they can keep their job. Perhaps managers believe that the company and their jobs are more secure if the company is merged or privatized after the reverse split, even though there is no evidence to support this belief. Maybe managers think that a reverse split will improve the company’s image. Institutions will buy their stock if it is priced higher and maybe managers’ gain some satisfaction or prestige as they are courted by Wall Street.

34 2.2.2.2 Shareholders’ Best Interest

Perhaps, with the forward split, the managers really are signaling to the shareholders that the company is strong and its prospects for the future are good and perhaps managers really feel that a reverse split will buck the odds and result in positive recognition and increased value for the shareholder. Maybe the managers are reverse splitting in an effort to improve liquidity or reduce transaction costs. The empirical evidence seems to indicate that such attempts are futile, so why do managers reverse split?

35 CHAPTER 3: METHODOLOGY AND RESULTS

This chapter is divided into three sections. In section one, the stock price reaction to a reverse stock split is examined using event study methodology and long-run buy-and-hold return analysis. Section one also contains an analysis of post split firm returns based on characteristics such as: industry, firm size, consolidation, motivation for the reverse split, and the stock price at the time of the reverse split. In the second section, the information content of a reverse stock split is investigated further by regressing returns on financial distress variables. The final section contains an analysis of NASDAQ’s change in continued listing requirements in 2001 and its effects on reverse stock splits and delistings.

3.1 Stock Price Reactions

This section deals with the stock price reaction to a reverse stock split. Event study methodology is employed around the announcement date and the ex- date of the reverse stock split and long-run returns are examined using buy-and- hold methodology. This section begins with the analysis around the event itself and concludes with an analysis of the long-run returns. As discussed in Chapter 2, researchers have documented negative abnormal returns around both the announcement date and the ex-date of reverse stock splits. The reverse split seems to have gained in popularity in recent years and thus should be re-explored in light of all of the new data that is available. Previous studies were based on much smaller data sets than used in this study. For example, Woolridge and Chambers (1983) examined 57 reverse splits, Peterson and Peterson (1992) examined 196 reverse splits, and Vafeas (2001) examine 229 reverse splits. There are 2,075 reverse splits in my raw sample of which 1,956 are for ordinary common shares. Of these 1,956, announcement dates are hand collected for 1,494 firms. However, some of these companies made other announcements concurrent with the reverse stock split announcement. These firms are deleted with the result being a “pure” sample of

36 1,398 firms. Some companies in the sample announced the re-election of board members concurrent with the reverse stock split announcement, but this is considered to be a benign event and thus those companies are included in the “pure” sample.28 With 1,956 ex-dates and 1,398 pure announcements, it is clear that the reverse split has gained popularity in the time since the previous studies. Previous research indicates that the reverse stock split is a negative event for a company. If that is true, then why are so many firms utilizing the reverse stock split in recent years? Is the reverse stock split still a negative event? The primary hypothesis of this section, stated in the null, is as follows:

There is no stock price reaction associated with the event of a reverse stock split. Measuring the stock price reaction around an event date has some complications - especially for smaller more obscure companies. Under ideal circumstances and with perfect and instantaneous information, the market should respond the same day, if not instantaneously, to company news. Unfortunately, this study is not rampant with ideal circumstances. For example, the typical company performing a reverse stock split is a smaller and less well-known firm characterized by thinly traded stock. When trading is thin, there may be an information shortage because each trade, theoretically, conveys information to the market. In addition to light trading, there may be few, if any, analysts covering these firms. In these less than ideal circumstances, it may take longer for information to be disseminated. It may be the case, then, that the effect on shareholders’ wealth is not confined to the event day, but rather spread out over a longer period. Event studies typically use a multi-day announcement window to allow for the fact that press releases may be published after the close of trading. If an announcement was made before or during trading hours, then the market should react that same day. However, if the press release is published after hours

28 An announcement of the status quo remaining unchanged is more of a formality than an announcement. The re-election of board members announcement is deemed to have had no information content. It is equivalent to an announcement such as “corporate headquarters is at the same address.”

37 then the market reaction should occur the next day. Using a two-day announcement period (days t=0 and t+1) captures both possibilities and is the method employed in this study. The days t-5 to t-1 are also examined to see if any information leakage or rumor trading is present. Following the announcement period, the days t+2 to t+5 are examined to see if any patterns of information dissemination lag might be present. This results in an 11-day window over which the returns around the announcement of a reverse stock split are examined. The hypothesis of interest here is:

There is no short-term stock price reaction associated with the event of a reverse stock split. Results are presented for daily raw returns, daily market adjusted returns, and daily market model adjusted returns. The CRSP equally weighted index is used as a proxy for market because most of the sample firms are relatively small. Since small firm returns have a greater impact on the equally weighted index than they do on the value weighted index, the equally weighted index is more characteristic of the sample. However, for robustness, calculations involving a market index are repeated using the CRSP value weighted index as an alternate proxy for the market. If using the value-weighted index impacts the test results, it is noted in the discussion of that test.

3.1.2 Raw Returns Around the Announcement Day

Of the 1,398 firms in the pure sample, 44 are dropped because there are no usable estimation period returns in the CRSP database.29 Of the remaining 1,354 firms, an additional 91 are dropped because of missing return data for at least one of the days in the 11-day test period. An analysis using the remaining 1,263 firms is presented below. The larger sample of 1,354 firms was also examined, missing daily returns were ignored, and results are virtually the same as those presented here.

29 The estimation period of t-160 to t-40 is used to estimate market model coefficients.

38 The mean raw return for each day (-5 to +5) surrounding the announcement day is presented in Table 6. The mean cumulative return for three different periods is also presented. Of these three periods, the “announcement date” (day 0 to +1) is of primary interest. The cumulative periods referred to throughout this section are:

PRE-AD Days -5 to -1

AD Days 0 to +1

POST-AD Days +2 to +5

Using all 1,263 firms, the mean AD cumulative return (CR) is -5.01% (0.001 significance) as shown in the bottom section of Panel A (Days 0,+1).30 The average PRE-AD CR (Days -5,-1) is 1.93% (0.01 significance) and the average POST-AD CR (Days +2 +5) is 0.27% but not significantly different than zero. Some firms have the same announcement date and ex-date.31 To detach the announcement day effects from the ex-date effects, the firms with a unique announcement date (Panel B) are tested separately. In this sample of 721 firms, there is no other corporate event on the announcement day and the announcement day effect is not confounded by the ex-date. As such, this sample should provide the truest test of the stock price reaction to the announcement of a reverse stock split. The sample of firms with identical announcement and ex-dates (Panel C) is also tested to determine if there is a different stock price reaction than for the pure AD sample. The two groups of firms are hereinafter referred to as:

30 The Patell Z (Patell 1976) is used to determine whether a return is significantly different than zero. This test is published in many event studies (for example: Linn and McConnel, 1983; Schipper and Smith, 1986). The actual Patell’s Z is given in the tables, but for ease of reading and interpretation, the significance levels are reported in the text as 0.001, 0.01, 0.05, 0.1 or not significant. The phrase “5.01% (0.001 significance)” means that the value of 5.01% is significantly different than zero at the 0.001 level according to the Patell’s Z score. 31 All firms in the sample have an ex-date in the CRSP database. This is the day of the actual reverse-split. The announcement date is found by searching for the first available news article in Lexis Nexis. A Wall Street Journal article or a company press release is the preferred source.

39 UAD firms with Unique Announcement Day

IAE firms with Identical Announcement day and Ex day

Comparing the average AD returns of the two samples, it is found that the stock price reaction is strongest for the IAEs. The average AD cumulative return is -7.31% (0.001 significance) for the IAEs (Panel C) and -3.29% (0.001 significance) for the UADs (Panel B). Given the IAEs’ stronger stock price reaction and higher Z score, it appears that the market reacts stronger on the ex-date than it does on the announcement date. This could also be evidence that the market reacts stronger to an announcement when the event is immediately forthcoming.

3.1.3 Market Adjusted Returns Around the Announcement Day

Daily market adjusted returns are calculated as:

ARit = Rit - Rmt where ARit is the abnormal return of stock i at time t, Rit is the return of stock i at time t, and Rmt is the market return at time t that is proxied by the CRSP equally weighted or value weighted index. The average daily market adjusted returns are shown in Table 7. The average AD cumulative abnormal return (CAR) is -5.19% (0.001 significance) for the full sample as shown in Panel A. The average AD CAR is -3.43% (0.001 significance) for the UADs (Panel B) and -7.54% (0.001 significance) for the IAEs (Panel C). The comparison between Panels B and C reveals that the AD stock price reaction is strongest when the announcement date and the ex-date are the same. This is further evidence that the market reaction to reverse stock splits is stronger on the ex-date than on the announcement date. These results incorporate the CRSP equally weighted index as the proxy for the market. For robustness, the CRSP value weighted index is used as an alternate market proxy and the results are shown on the second page of Table 7. The results are nearly identical. The only difference is that the average PRE-AD

40 CAR and the average POST-AD CAR are slightly more significant for the IAEs (Panel C).

3.1.4 Market Model Adjusted Returns Around the Announcement Day

The analysis of raw returns around the announcement date reveals a significant negative stock price reaction to the reverse stock split (-5.01% CR from Table 6). This negative reaction remains about the same for the market-adjusted returns (-5.19% CAR from Table 7). The investigation of abnormal returns (CARs) around the announcement date proceeds by applying the market model. The daily abnormal returns are calculated using the market model estimates as:

∧ ∧ = − − ARit Rit α i β i (Rmt ) where ARit is the abnormal return of stock i at time t, Rit is the return of stock i at

∧ ∧ time t, α i is the intercept estimate from the market model, β i is the slope estimate from the market model and Rmt is the market return at time t that is proxied by the CRSP equally weighted or value weighted index. The coefficients are estimated using OLS regression for t equals -160 to -40 relative to the announcement date. This represents about six months of trading days in a period prior to the event, but it excludes a possible volatile trading period in the 40 trading days (about two months) prior to the announcement date. The estimation model is:

Rit = αi + βi (Rmt) + εit where Rit, αi, βi, and Rmt are defined previously and εit is the error term in the regression. The average market model adjusted returns around the announcement date are presented in Table 8. As with the raw returns and the market-adjusted returns, the strongest stock price reaction is seen for the IAEs (Panel C). The full sample of announcement dates, shown in Panel A, has an average AD CAR of -5.30% (0.001 significance). The PRE-AD CAR of 1.29% is also significant at the 0.01 level.

41 The UAEs (Panel B) have an average AD CAR of -3.51% (0.001 significance) and the IAEs (Panel C) have an average AD CAR of -7.68% (0.001 significance). Again, the stronger reaction and larger Z-score for the IAEs provides evidence that there is a stronger stock price reaction to a reverse stock split on the ex-date than on the announcement day. As with the market-adjusted returns, the CRSP value weighted index is used an alternate proxy for the market. The results are provided on the second page of Table 8. The average CARs and the significance levels are about the same regardless of the market proxy used.

3.1.5 Announcement Day Conclusions

The analysis of raw and abnormal returns around the announcement date is in harmony with the existing literature in that the mean abnormal returns are negative on the announcement date. However, the IAE analysis reveals an unexpected result. There are positive CARS in both the PRE-AD and POST-AD periods. This is not true for the full sample or for the UAEs. The PRE-AD cumulative returns are influenced most heavily by the positive and significant returns on day -1. Since only the IAE firms exhibit this phenomenon, it is most likely an ex-date effect rather than an announcement day effect. While explanations for the positive and significant PRE-AD CARs are rather hard to come by, the positive and significant POST-AD CRs and CARs for the IAEs may suggest an initial AD over-reaction to bad news (the surprise reverse stock split) followed by a correction.

3.1.6 Ex-date Raw Returns

Since the analysis of the average AD return provides evidence of an ex- date effect, the ex-date returns are also analyzed. Unlike the announcement date, which may be a little vague in terms of when the market assimilates the information, the ex-date is a precise point in time. Therefore no multi-day returns

42 are used in the analysis of ex-date returns. However, for comparison purposes, cumulative returns are presented for days -5 to -1 and +1 to +5. As with the announcement day returns, the sample is culled because of missing return data in the CRSP database. Of the 1,956 firms with ex-dates, 51 firms are dropped because no useable returns are available in the estimation period and 94 firms are dropped because at least one daily return is missing during the 11-day test period. The remaining 1,811 firms have an average ex-date raw return of -6.08% (0.001 significance) as shown Panel D of Table 9.32 For some firms, there was no announcement date found in the Wall Street Journal or in the Lexis-Nexis database. The average ex-date raw return for this subsample is -5.86% (0.001 significance) as shown in Panel E of Table 9. There are two assumptions that can be made about the firms with missing announcement dates. Either an announcement exists that was not reported in the two sources mentioned above, or there was no announcement. If the market knew about the reverse stock split prior to the ex-date, then there was an “announcement” and the firm should be included in the sample of firms in which the ex-date is not the announcement date. The average ex-date raw return for firms in which the announcement date is not the ex-date (and firms with missing announcement dates are included) is -5.78% (0.001 significance). Average cumulative returns and the average daily returns for this subsample are shown in Table 9 Panel F. If the market did not know about a firm’s reverse stock split prior to the ex- date, then there was no announcement and the firm should not be included in the sample of firms in which the ex-date is not the announcement date. The average ex-date raw return for firms in which the announcement date is not the ex-date (and firms with missing announcement dates are not included) is -5.74% (0.001 significance). Average cumulative returns and the average daily returns for this subsample are shown in Table 9 Panel G. This subsample is the purest sample of

32 As with the announcement day results, the results presented are virtually the same as the results using the full sample of 1,956 firms. Even with the full sample, 51 firms were dropped due to missing estimation period data.

43 ex-date returns. The ex-date is different from the announcement date and firms with missing announcement dates are not included. Based on the average raw returns presented in Table 9 there seems to be a stronger stock price reaction to reverse stock splits on the ex-date than on the announcement date. Comparing Panel G (purest ex-date returns) with Table 6 Panel B (purest AD returns), the average ex-date return is -5.74% with a Z-score of -20.669 while the average AD cumulative return is -3.29% with a Z-score of -8.156. The average stock price reaction on the ex-date is stronger and the Z- score is greater in magnitude. This suggests an ex-date effect, but this effect may not persist if the abnormal returns are considered. The following analysis of the ex-date abnormal returns mirrors the procedure used to examine announcement day returns.

3.1.7 Ex-date Market Adjusted Returns

Market adjusted returns are calculated as before and are presented in Table 10. The significant average ex-date returns persist. The average ex-date return for all firms is -6.18% (0.001 significance). For firms with no announcement date, the average ex-date return is -5.96% (0.001 significance). The average ex- date return for firms in which the announcement date is not the ex-date is -5.85% (0.001 significance) when firms with missing announcement dates are included (Panel F) and -5.81% (0.001 significance) when firms with missing announcement dates are not included (Panel G). The CRSP equally weighted index is used as a proxy for the market in calculating these results. For robustness, the CRSP value weighted index is used as an alternate proxy, but the results (last two pages of Table 10) change by only a couple of hundredths of a percent and the significance remains the same.

3.1.8 Ex-date Market Model Adjusted Returns

Market model adjusted returns are calculated as before and are presented in Table 11. The significant average ex-date return continues to persist. The

44 average ex-date return for all firms is -6.21% (0.001 significance). For firms with no announcement date, the average ex-date return is -5.97% (0.001 significance). The average ex-date return for firms in which the announcement date is not the ex-date is -5.86% (0.001 significance) when firms with missing announcement dates are included (Panel F) and -5.81% (0.001 significance) when firms with missing announcement dates are not included (Panel G). The average market model adjusted returns are nearly identical to the average market adjusted returns. In addition, the results are robust to the market proxy used. Using the CRSP equally weighted index versus the CRSP value weighted index has a negligible effect on the results (as shown on the last two pages of Table 11). The analysis of announcement day and ex-date returns shows that there is an ex-date effect that is stronger and more significant than the announcement day effect. The effect is present in raw, market adjusted, and market model adjusted returns.

3.1.9 The relation between ex-date returns and announcement date returns

In an efficient market the market’s full response to the reverse stock split should occur on or around the announcement day. In general, there is no information content on the ex-date. This is just a confirmation of a previous announcement. However, a strong significant ex-date effect is observed for our samples. Even though there is no obvious reason for a correlation to exist, it may be possible that the ex-date effect is somehow related to the announcement day effect, and we assess whether this is the case. Ex-date returns are regressed on announcement date returns for firms where the announcement date is different than the ex-date. Using the AD cumulative return as the independent variable and the ex-date return as the dependent variable, the OLS regression reveals a significant, positive relationship. When market adjusted or market model adjusted returns are used in the regression, the relationships remain positive and significant. The regression

45 results are presented in Table 12. As shown in the table, the results are robust to the market index used in the calculations.

3.1.10 Analysis of Subsamples Based on Firm Categories

In this part of the analysis, subsamples of firms are formed based on five firm specific criteria: the industry of the firm, the firm’s motivation for the reverse stock split, the firm’s stock price after the reverse stock split, the size of the firm, and the degree of consolidation.33

3.1.10.1 Industry

The first subsample analyzed is based on industry classification. Some industries seem to get more press than others leaving a causal observer with the impression that market downturns and reverse stock splits are concentrated in certain industries. For example, the market down turn of 2000-2001 has been called the Internet bubble burst even though firms from many industries suffered stock price reductions during that period. It may be true that firms in certain industries have a stronger stock price reaction to reverse stock splits and in this section we assess whether this is the case. The hypothesis of interest here, stated in the null, is:

The stock price reaction to a reverse stock split is not related to the industry classification of a firm.

Table 13 shows the actual concentration of reverse stock splits by industry. The industry classification assigned to a firm is based on the Fama and French (1997) industry definitions listed in Appendix C. According to Fama and French the firms within these classifications should have similar risk characteristics. There are seven ways to partition the market into industry classifications. The first

33 Consolidation is the number of shares that were turned into one share. A 10 to 1 reverse stock split was a larger consolidation than a 2 to 1 reverse stock split. This is the same idea as a split factor.

46 way is to partition the market into five broad industry groups. The remaining ways divide the market into 10, 12, 17, 30, 38, or 48 industry classifications. This type of grouping results in more firms per group than the ordinary method of grouping industries by the first two digits of the SIC. Having more firms in an industry group is helpful in the statistical analysis because larger samples have higher degrees of freedom and, all things being equal, that results in a greater statistical power (higher t-values). In addition, using this industry classification makes sense because firms with dissimilar SIC codes may be part of similar industries. With firms classified as described above, the average ex-date return is calculated for each industry classification. These results are presented in Table 14 and it appears that there are some differences across classifications. The differences are determined as follows. For each set of industry classifications some of the individual classifications do not have an average ex-date return that is statistically different from zero at the 5% level (a t-value greater than 1.96).34 For instance, there are five classifications in the first set, but only four of the classifications have an average ex-date return that is significantly different than zero. The largest and smallest significant average returns are compared within the category grouping. For example, in the Five Industries category, the returns of “Utils” (-4.77%) exceed the returns of “Shops” (-8.06%) by almost two to one, but the returns of “Utils” is not significant at the 5% level (t-value = -1.38) so it is ignored.35 The classification with the best significant average ex-date return is “Money.” Even though “Money” has a 2.91% better mean return than “Shops,” these two means may not be statistically different. Therefore a two-sample t-test for equal means is used to determine if the two population-means, u1 and u2, are equal (Snedecor and Cochran, 1989).

H0: u1 = u2

34 If a number is not statistically different than zero, it does not necessarily mean that the number is zero; it means that there is not enough data to confidently say what the number is. This is an important distinction because, in some groups, an average return of zero would be the best return for the group. 35 For the reminder of this section the phrase “returns of “Shops” (-8.06%)” is shorthand for “the average ex-date return for firms in the industry classification named “Shops” is -8.06%.”

47 Ha: u1 ≠ u2 The test statistic is Y −Y T = 1 2 s2 N s2 N 1 / 1 + 2 / 2 where Y is the sample mean, s is the sample standard deviation and N is the sample size. When “Money” and “Shops” (best and worst) are compared, the test of equal means indicates that there is no difference in means at the 5% level (T=1.78). There is therefore no insight in the analysis of the five broad industry classifications. However, when the industry classifications are narrowed, a pattern emerges. For instance, in the Ten Industries category the returns of “Oil” (-4.46%) exceed the returns of “Manuf” (-8.23%) almost two to one. In this category T is 2.89 indicating that the means are statistically different at the 5% level. Both returns are negative and significant so higher returns do not necessarily mean that one industry does well. It simply means that one industry does better in a relative sense than the other. In the Twelve Industries category the returns of “Energy” (-4.46%) exceed the returns of “BusEq” (-8.29%) almost two to one. A T of 2.82 indicates that the means are statistically different at the 5% level. In the Seventeen Industries category the returns of “Oil” (-4.38%) exceed the returns of “Clths” (-12.8%) almost three to one but a T of 1.79 indicates that the means are statistically different at only the 10% level. The next worst classification is “Retail” and the difference in the returns of “Oil” and “Retail” is significant at the 5% level (T=2.22) The Thirty Industries category is even more dramatic. The returns of “Oil” (- 4.37%) exceed the returns of “Clths” (-17.40%) almost six to one with a T 2.43. This seems to be the same result as in the Seventeen Industry definitions, but the “Oil” and “Clths” classifications are slightly different in the two categories.

48 Appendix C identifies the SICs that are associated with each classification in each group. In the Thirty-eight Industries category, the returns of “Ptrlm” (+4.51%) exceed the returns of “Apprl” (-20.39%) with a T of 4.64. And finally in the Forty-eight Industries category, the returns of “BldMt” (- 4.16%) exceed the returns of “Clths” (-17.40%) more than four to one (T=2.37). The returns of “Oil” (-4.37%) are the second best in the group and the mean returns of “Oil” and “Clths” are significantly different at the 5% level (T=2.43) but this is not shown on the table.

Industry Conclusions It seems that the industries that are non-discretionary, like food, energy and oil, do better than the discretionary industries like retail, clothes, apparel, and business equipment. “Oil,” “Petroleum,” or “Energy” are the best performers in most categories. Interestingly, one of the classifications with the worst returns, the “business equipment” category from “Twelve Industries,” contains many of the famed “Internet companies.” When people talk about an Internet bubble burst, they are more often than not referring to firms in this industry category. These results may be interpreted to mean that companies that provide essential non-discretionary items or services may fare better than other companies in periods of financial distress. However, the top performer in “Forty-eight Industries” is “building materials,” so we will not be dogmatic in our interpretations other than to say that we have shown a difference in stock price reactions to reverse stock splits across industries.

3.1.10.2 Delisting threat and Motivations for Reverse Splits

One facet of this study deals with the regulatory environment. Specifically, firms that reverse split their stock due to the threat of delisting are analyzed separately from those that reverse split for other reasons. If there is a difference in the results for these different groups of firms, then it is possible that the exchange

49 regulations may be having an impact on the market’s reaction to reverse stock splits. Signaling theory, as documented in the literature, asserts that managers are sending information to the market when they split or reverse split their stock. The best signals are those that are easily recognizable by the market, relatively inexpensive to use, and hard to copy by under-performing competitors. Stock repurchases and dividend increases are positive signals to the market that under performing companies would find hard to duplicate. The literature reveals that forward splits are generally followed by positive abnormal returns and reverse splits are generally followed by negative abnormal returns. The conclusion is that this is evidence of signaling. Forward stock splits are a positive signal, and reverse stock splits are a negative signal. Though reverse stock splits are generally considered a negative signal, the statements by management at the time of the reverse stock are rarely negative in tone. Rather managers tend to put a positive face on the reverse stock split announcement. They are “improving liquidity,” “appealing to institutional investors,” “making the price more attractive,” or giving some other positive reason. If the reverse stock split is a negative signal, then managers seem to be trying to counter the negative signal with their press releases. Many companies are in poor financial condition when they resort to a reverse stock split to prop up their share price. They may be admitting that they have no other means of addressing the problems they currently face. However, it may also be true that some managers do not desire to reverse split their stock but are forced to do so for regulatory reasons. They may believe that the company’s long-run prospects are good and may be very reluctant to send a negative signal to the market, except that an even worse fate, market delisting, is in their short- term future if they do not act. In order to stay listed on a market exchange, a minimum price must be maintained in the short term. The current low price may be due to market conditions or industry conditions (systematic) rather than to specifics of the particular company (idiosyncratic). Reverse stock splits undertaken because of regulatory pressures rather than financial pressures might produce a

50 different wealth effect, on average, and this is the issue that is addressed below. The next wealth effect hypothesis, stated in the null, is then:

There is no difference in the wealth effects on the stock of firms which reverse split their stock for voluntary versus regulatory reasons. Peterson and Peterson (1992), in a very small sample, find positive wealth effects for companies that reverse split for regulatory reasons. Two-day announcement period abnormal returns were significantly positive for “pure” firms that reverse split for regulatory reasons. Pure firms are those firms that had no announcements other than the reverse stock split. Abnormal returns were defined as mean adjusted returns where the firm’s return was adjusted by its own average post announcement return.36 The above hypothesis is tested by grouping firms into subsamples based on public statements concurrent with the reverse split announcement and then comparing the average daily returns of companies that split because of minimum bid price requirements with those of companies that split for other reasons. The reasons identified were given in press releases found in the Wall Street Journal or in the Lexis-Nexis database. Many firms gave no reason at all and these firms are not included in any of the samples for these tests. In our sample the most common reasons include, threat of delisting (384), reorganization (159), attracting attention (99), merging with another firm (97), moving to a desired trading range (74), increasing liquidity (35), or maintaining NMS status (36). The returns are shown by reason in Table 15. The “N” shown is the number of firms that have valid returns for days -2 to +2. For example, there are 384 firms citing “delisting threat” as the reason for the reverse stock split, but only 373 of these firms have non-missing returns in the CRSP database for all five days.

36 In their study, the daily return was averaged from day t+2 to day t+101 with t=0 being the announcement day.

51 The average ex-date return for firms that cite “delisting threat” is -6.31% (0.001 significance). This is the worst average return for any motivation. Other significant average ex-date returns were found for firms that reverse split their stock due to: a reorganization (-5.74%, 0.001 significance), their desire to remain on the NMS (-5.51%, 0.001 significance), a merger or acquisition (-3.51%, 0.05 significance), a trading range adjustment (-2.46%, 0.01 significance), or to attract attention to the firm (-3.35%, 0.001 significance). Ex-date returns are used throughout this study, but the announcement date (AD) returns yielded similar results. Earlier studies (cited above) found that firms that reverse split for regulatory reasons outperformed firms that reverse split for other reasons. The results of this test, taken from a larger sample, do not support these earlier findings.

3.1.10.3 Price

After the individual reasons given are analyzed, a second test is conducted. Motivations are tested based on pre-split price rather than the reason given by management. The assumption is that any firm with a stock price of less than one dollar is reverse split for regulatory reasons regardless of the managers’ given reason. The rationale is that compliance is at least a partial motivation when the price is under one dollar. This type of motivation test is unique in the literature. Table 16 shows the results of this testing. Stocks with pre-split prices below $1.00 have an average ex-date return of -8.75% to -8.58% and stocks with pre- split prices above $1.00 have an average ex-date return of -2.65% to -2.49%. The means are significantly different as shown by the large t-values (two sample t-test of equal means). Stocks in danger of delisting due to low price have worse returns than stocks that reverse split for other reasons. Related to price is the post split price that the firm is trying to achieve. The rationale is that firms that aim for a higher post split price have the lowest expectations regarding future performance. Consolidation to the higher price will

52 buy the company more time to turn around before it is threatened by delisting again. On the other hand, a post split price of at least $5 may be desirable because many institutions avoid stocks that are priced below $5. In addition, a common margin requirement for traders is that the stock bought on margin be priced at least $5 per share. The average returns according to target post split price are also shown in Table 16. Target post split price is defined as the average price for the twenty days before the announcement date times the number of shares consolidated. The table shows the difference in returns for stocks with post splits prices over and under $5. Firms that attempt to reverse split to the higher price levels have, on average, better event day returns than those firms that attempt to split to a lower price level. Higher priced firms, $10 and above are also examined. The results indicate that stocks aiming for a higher trading range have better (or less bad) returns on the event day.

3.1.10.4 Size of the Firm

Large firms may have more resources at their disposal and thus may be more resilient than small firms. Ericsson and Priceline.Com are two large firms that received a lot of press coverage around the time of their reverse stock split. The fact that they have recovered well (they have positive post split abnormal returns) may leave the general public with the impression that large companies are less susceptible to the negative stock price reaction that normally follows a reverse stock split. Tests are conducted to assess whether this is the case. Size is calculated as the number of shares multiplied by the price per share and the test results are given in Table 17. The Ex-day returns are regressed on firm size. There is no relationship between size and ex-date returns. The coefficient of size is not significantly different than zero for any model. Results are similar for AD returns. The conclusion from this test is that the market reaction to reverse stock splits is not a function of firm size.

53 3.1.10.5 Consolidation

A reverse stock split consolidates the firm's shares into fewer shares. Companies with the worst prospects for the future generally need to consolidate the most. For example the company splitting 10 for 1 may have a dimmer outlook than the company splitting 2 for 1. The larger consolidation may also be a negative consideration if it results in the company being near continued listing limits on shares outstanding or number of shareholders. If the reverse stock split sends a signal, a larger share consolidation may send a stronger signal and the market may react less favorably to the larger consolidation. To test this idea, ex- date returns and AD cumulative returns are regressed on the number of shares consolidated such that:

R0 = α + β(number of shares consolidated). Both AD returns and ex-date returns are discussed here because the AD results differ from the ex-date results. The results are presented in Table 18. If the AD cumulative return is the dependent variable, the regression coefficient is -0.0007 with a t-value of -1.81. The coefficient is marginally significant so there is a slight relation between the AD CRs and the number of shares consolidated. Results are similar for the market adjusted AD CAR, but when the market model adjusted AD CAR is the dependent variable, the regression coefficient is -0.0008 and significant (t-value = -2.19). Even though there is a significant relation between market model adjusted AD CARs and consolidation, it may be argued that there is not a strong “economic” relation. A 10 to 1 consolidation is only 0.4% worse than a 5 to 1 consolidation (10X0.0008 - 5X0.0008). However, considering this is a daily return and 0.4% compounded daily for 250 trading days (about one year) is 171%, it is economically significant. A “normal” average daily return for the market would be around one tenth of this difference. For example, a 0.04% return compounded daily for a 250-days would yield 10.5% annually.

54 The AD CR is marginally related to consolidation, but if the ex-date return is the dependent variable, the regression coefficient is -0.0013 and the t-value is -6.09. This is evidence of a significant relation between the number of shares consolidated and the ex-date return. The negative coefficient means that larger consolidations will generally result in more pronounced negative reactions on the ex-date. As with the AD returns, the coefficient seems small, but as demonstrated above the relationship is economically and statistically significant. Results are similar when the market adjusted ex-date return or the market model adjusted ex- date return is regressed on consolidation. These results show that consolidation has a stronger effect on the ex-date returns than on the announcement date returns. This concludes the analysis of the market reaction to reverse stock splits on the announcement day and on the ex-day. The next section begins the analysis of long-run returns after the reverse stock split.

3.1.11 Long-run Returns Following Reverse Stock Split

The analysis of long-run returns after the reverse stock split follows a similar sequence as before. The raw return analysis is followed by an investigation of market adjusted abnormal returns and then by an investigation based on a matched firm method. Long-run returns are further examined by analyzing subsamples of firms based on five firm specific criteria: the industry of the firm, the firm’s motivation for the reverse stock split, the firm’s stock price after the reverse stock split, the size of the firm, and the degree of consolidation.

3.1.12 Long Run Buy and Hold Returns

Raw returns are examined first. In this section, each stock’s long-run “buy- and-hold” return (BHR) is calculated for holding periods of one trading day to 250 trading days after the ex-date. The 250-day holding period corresponds to approximately one calendar year. Over the course of the year, some firms have missing returns. Some firms are missing many returns in a row over several

55 different periods. The return data starts and stops and starts again. The firms with sporadic or missing returns are removed from the sample leaving a sample size of 1,245 firms. Of these firms, 27 are dropped because there is no return data after the reverse split.37 There are 1,218 remaining firms. Though firms with missing returns are removed from the sample, the firms delisted before the end of the 250-day period are not removed as long as returns are consecutive up to the delisting day. Of the 1,218 firms remaining in the sample, 239 are delisted by the end of the 250-day period. These firms are delisted at a rate of about 20 per month. Since these firms stay in the sample, and must be dealt with in the calculations, it is useful to examine the nature of the delistings. Table 19 provides a breakdown of the reasons for delisting as recorded in the CRSP database. Of the 239 firms that disappear from the sample, 24 firms (10%) merged with other firms. These 24 firms have an average daily return of positive 1.2% on the delisting day. Most of the firms, 214, were delisted for non- compliance with exchange rules.38 The average daily return on delisting day is -15.4% for these firms. The most common non-compliance issue is bid price (57), followed by insufficient float or assets (32), insufficient capital (29), non-payment of exchange fees (22), and insufficient equity (16). Interestingly, only 11 of the 239 firms gave bankruptcy as a reason for delisting. Delisting does not result in a total loss to the investor. In fact, only one company in the sample declared bankruptcy in the year following delisting. Furthermore, the average delisting return taken from the CRSP database is -13.5% on the day that a firm is delisted. If delisting is not a total loss, then what return should be used when a firm is delisted? Two methods are employed here. First, it is assumed that a delisted firm is sold and the portfolio rebalanced. The second method employed here substitutes the return of a broad-based market portfolio for the return of the delisted security for the remaining days in the test

37 Some firms were merged with other firms; others were taken private via the reverse split. 38 The remaining firm (of the 239) was liquidated. This firm is missing return data, but it has a positive liquidation price of $0.375 per share and is therefore not a total loss.

56 period. The latter assumption neutralizes the effect of the delistings when the market adjusted model or market model is used to analyze returns. In this section raw returns are examined first. The buy-and-hold return (BHR) for a firm is calculated as

T BHRiT = ∏ 1( + rit ) −1 t =1 where BHRiT is the raw return for stock i over the holding period T and rit is the return on the security i (including dividends) on day t after the ex-date. The mean return for each holding period is calculated as

N ∑BHRiT BHR = i =1 T N where N is the number of firms in the sample. As discussed above, missing returns are handled two different ways. One way to handle missing returns is to replace the missing return with the average portfolio return for that trading day. When a firm’s return is not reported by CRSP, the return for that day is replaced by the average return of all firms in the sample for that trading day such that:

NT ∑r jt j =1 rit = NT where NT is the number of firms in the sample on day T and t is the number of trading days after the ex-date. The mean calculated in this manner is the return an investor would realize by selling the delisted firms and using the proceeds to rebalance the equal weighted portfolio. The 250-day mean BHR calculated in this way is 8.80% (t-value=2.25) as shown on Table 20. The second method of dealing with missing returns is to replace them with zero. The assumption here is that once the firm is delisted, it neither adds to nor reduces portfolio value. The 250-day mean BHR calculated in this way is 5.60%

57 (t-value=1.44).39 Since the average daily return is higher than zero in the latter days, replacing returns with zero lowers the average BHR. These positive mean returns seem to be in conflict with the idea of a reverse stock split being a negative event. However, upon closer examination of the distribution of the BHRs (where missing returns are replaced with the average return for that trading day), the skewness is 6.0 which indicates a positive bias from normal and the kurtosis is 53.1 which indicates very thin tails compared to the normal distribution. These two indicators reveal that there are a few very large values that are driving the average up. In this case, the median may be a better indicator of the average than the mean. The 250-day median return (50% quartile) is -0.24%. Other quantiles of interest are the 75th (BHR = 24.6%) and the 25th (BHR = -58.3%). Even though the median may be a better indicator of the average, the question of whether an investor would be better off or worse off for buying and holding reverse stock splits is more closely related to the mean return. For the 250-day holding period, history shows that an investor would have lost on about 3 out of 4 firms, but that 1 out of 4 firms would have returned better than 24.6%. This high return on a few firms would have offset the many losses and resulted in an 8.80% raw gain. The distribution of these returns is shown in Figure 3. Table 20 shows the BHR for reverse split stocks for 10-day intervals through 250 days. The data confirm previous studies that show that returns to reverse stock splits are generally negative for a period following the ex-date. The average BHRs are negative for the first 180 days, which represents about 8.6 months of calendar time. However, after this period, the mean BHR turns positive and by day 250 the BHR is 8.80%. These BHRs are plotted in Figure 4. Alternatively, the missing returns for the delisted stocks can be replaced with the market return for that day. The assumption here is that for the remaining days in the test period, a delisted firm will, on average, earn the same return as

39 Replacing the returns with zero implies that the delisted firm is sold for cash and the cash is held until the end of the 250-day period. There are no further losses or gains on that firm. If it is cost prohibitive to rebalance the portfolio, this may mimic investor behavior.

58 the market. In this model, the CRSP equally weighted index return is used as a proxy for the market. The model then becomes: N ⎛ T ⎞ ∑⎜∏ 1( + rit ) −1⎟ BHR = i =1 ⎝ t =1 ⎠ T N where

T ∏ 1( + rit ) −1= BHRiT t =1 is the buy-and-hold return for firm i for T days as before except that rit is replaced with the market return when returns are missing. N is the number of firms in the sample and T is the holding period. When returns are calculated in this manner the mean BHR is 7.59% with a t-value of 1.95. Table 20 (second page) shows the BHR for reverse split stocks for 10-day intervals through 250 days. If the CRSP value weighted index is substituted for missing returns, then the 250-day BHR is 6.27% with a t-value of 1.61.

3.1.13 Long-run Abnormal Returns

The analysis now turns to abnormal returns. The BHAR provides insight into how an investor would have fared against a benchmark. The BHRs may be positive and significant over longer periods, but the same may not be true for the abnormal returns. The buy-and-hold abnormal return, or BHAR, for security i is calculated for each holding period T as

T T BHAR r E r iT = ∏ 1[ + it ] − ∏ 1[ + ( it )] t=1 t=1 where E(rit) is the return on a reference portfolio. The alternative is to use cumulative abnormal returns (CARs) calculated as

T CAR = [R − E(R )] iT ∑t=1 it it where E(Rit) is defined as the expected return of a reference portfolio such as a market index. However, Barber and Lyon (1997) argue that for long-run returns, BHARs are more meaningful than CARS. Barber and Lyon document that CARs

59 are a positively biased predictor of BHARs, which may lead to improper inferences; that is, there may be significant CARs when in fact, no significant BHARs are present. Furthermore, CARs assume a daily rebalancing of the portfolio and investors do not manage their portfolio that way. Barber and Lyon (1997) also document problems with the mean BHAR; it could be biased due to new listings, rebalancing the benchmark portfolio may cause bias, and the multi- year returns of individual firms are skewed positive. The skewness, when compounded, may cause bias. The bias can be controlled by having a large sample size or by careful construction of the benchmark portfolio. Fama (1998) argues that the BHAR methodology can yield significant abnormal returns when none are present due to the “bad model problem.” The benchmark is an imperfect projection of expected returns. Fama (1999) argues for the calendar-time portfolio method because: this approach accounts for cross correlation in firm returns, the distribution of the calendar time abnormal returns are approximately normal (ease of statistical interpretation), and the bad model problem is not as pronounced when monthly returns are used. Lyon, Barber, and Tsai (1999) and Loughran and Ritter (1999) counter Fama (1998) by arguing that the calendar time abnormal returns have low power to detect abnormal returns and they do not reflect the investor’s experience. Furthermore, they offer two bootstrap procedures that can be used for statistical interpretation of the mean BHAR when the sample is not sufficiently large. Mitchell and Stafford (2000) offer a rebuttal to Lyon et. al. arguing that the BHAR methodology is flawed because it assumes independence of multi-year abnormal returns and this assumption produces test statistics that are too large. In this study there is a large sample size, the benchmark portfolio is the market index (or a carefully chosen matched firm portfolio), and the long-run returns are only 250 days (one year of trading days versus a multi-year approach). For these reasons, BHARs are used in this analysis without reservation. The mean BHAR is calculated for each holding period T as

60 N ⎧ T T ⎫ ∑ ⎨∏ 1[ + rit ] − ∏ 1[ + E(rit )]⎬ BHAR = i =1 ⎩ t =1 t =1 ⎭ T N where E(rit) is the daily return of the CRSP equally weighted index. If rit is missing, it is replaced by the average return of the remaining firms in the sample (as before).

The mean 250-day BHAR is -23.19% with a t-value of –6.11. If E(rit) is the daily return of the CRSP value weighted index, then the 250-day BHAR is not significantly different than zero (t-value = -0.93). Values for the mean BHAR for holding periods from 10 days to 250 days, in 10-day increments, are provided in Table 21 and plotted in Figure 5. As with the raw returns, the analysis is also conducted by substituting the CRSP equally weighted index daily return for missing returns in the sample. In this case, the mean BHAR is calculated as

N ⎧ T T ⎫ ∑ ⎨∏ 1[ + rit ] − ∏ 1[ + E(rit )]⎬ BHAR = i =1 ⎩ t =1 t =1 ⎭ T N where the daily return, rit, is replaced with the daily CRSP equally weighted return when rit is missing. The 250-day BHAR calculated in this way is -24.4% with a t- value of -6.46. These numbers are shown on the second page of Table 21. If an alternative market proxy, the CRSP value weighted index, is used then the 250- day BHAR is -6.10% with a t-value of -1.59. This is not significantly different than zero at normal acceptance levels. To summarize, the average 250-day BHR following a reverse stock split is positive and significant, but the BHAR results are mixed. If the CRSP equally weighted index is used in the abnormal return calculations, then the average 250- day BHAR is negative and significant. However, if the CRSP value weighted index is used in the abnormal return calculations, then the 250-day BHAR is not significantly different than zero. To further investigate this issue, another method of calculating abnormal returns is employed.

61 3.1.14 Matched Firm Analysis

When calculating market adjusted or market-model BHARs, a proxy for the market is used in calculating an expected return for a firm. The matched firm method uses a companion firm, rather than the market, as a basis for determining expected returns. Each sample firm is matched to an out of sample firm based on several firm specific characteristics and all things being equal, the sample firm should have the same returns as the matched firm. If the firms are matched well, then the only distinguishing characteristic between the firms is a corporate event (in this case the reverse stock split). Any difference in the returns of the two firms is considered to be the “abnormal” return that is associated with the corporate event. The matched firm method of calculating abnormal returns may be preferable here due to the unique characteristics of firms that reverse split their stock. Typically, these firms are much smaller than the average size of firms in the CRSP indices, so adjustment based on market movements may lead to questionable conclusions. Pairing the sample firms with firms of similar characteristics and then testing for differences may then be the preferred alternative, and at the very least, it provides a robustness check. Firms in the sample are matched with out-of-sample firms based on size, industry, and price. Size is proxied by market value of equity and industries are matched according the 2-digit SIC codes. The matched firms are priced within a dollar of the original firm and they are the same size plus or minus 15 percent. In the case where several matches are found, the best price match is chosen. Sample firms are not allowed to match other sample firms, even in different time periods. However, a matched firm is allowed to match more than one firm from the sample as long as the matches occur in different months. The average sample firm price is $3.73 (standard deviation 4.957) and the average matched firm price is $3.71 (standard deviation 4.961). The average market value of equity of the sample firm is $47,168,470 (standard deviation $202,323,470) and the average market value of equity of the matched firm is $46,648,060 (standard deviation

62 $192,322,130). These statistics are tabulated in Table 22. These matching criteria are too constraining for some firms and no match is found. The resulting sample size is 691 firms. In the matched firm method, returns are “abnormal” if the average return of the sample of reverse stock splits is significantly different from the average return for the sample of matched firms. The buy-and-hold abnormal return (BHAR) is calculated as

T T BHAR r E r iT = ∏ 1[ + it ] − ∏ 1[ + ( it )] t=1 t=1 where BHARiT is the abnormal return for stock i for the holding period T, and E(rit) is the return of a firm matched to the sample firm based on size, price, and industry. Each firm’s long-run BHAR is calculated for holding periods of one trading day to 250 trading days and the average matched firm BHARs are calculated as

N ⎧ T T ⎫ ∑ ⎨∏ 1[ + rit ] − ∏ 1[ + E(rit )]⎬ BHAR = i =1 ⎩ t =1 t =1 ⎭ T N where N is the number of firm in the sample for holding period T. Table 23 presents the matched firm adjusted BHARs. The 250-day matched firm adjusted returns are -8.26% with an unpaired sample t-value of -1.01. The sample average raw return is 23.27% and the matched firm average raw return is 31.53%. The unpaired sample t-value is calculated as BHRS − BHRM t = T T var(BHRS ) var(BHRM ) T + T N N where BHRST is the buy-and-hold return of the sample for holding period T,

BHRMT is the buy-and-hold return of the matched firm for holding period T, and “var” is variance of the sample for holding period T. Figure 6 shows the BHARs plotted from day 0 to day 250 from the ex-date. Table 23 reveals that the BHARs are significantly negative in the short run, but after about 8 months, the matched

63 firm returns and the sample returns are not significantly different from each other. Similar results (long-run BHAR = 0) were found above when the BHARs were calculated using the CRSP value weighted index as a market proxy. To summarize, the average 250-day BHR after a reverse stock split is positive and significant, but the average 250-day BHAR is either not statistically different than zero or negative and significant. Two tests show no significant 250- day returns, but one test (the market adjusted BHAR using the CRSP equal weighted index as the market adjustment) shows negative and significant BHARs.

3.1.15 Analysis of Subsamples Based on Firm Categories

The analysis of long-run returns continues by examining subsamples of firms based on firm characteristics. As in the investigation of announcement day and ex-day returns, subsamples are created based on industry, motivation, price, size, and consolidation.

3.1.15.1 Industry

Firms are grouped into industries based on the Fama and French (1997) industry classifications and the average 250-day BHR by industry is presented in Table 24. As before, there are seven sets of classifications in which firms are grouped into 5, 10, 12, 17, 30, 38, or 48 industry classifications. When industries are divided into five classifications, there is a significant difference between the best and worst industry BHR. In determining the best and worst BHR, the BHRs are ignored that are not statistically different from zero at the 10% level.40 In the Five Industries category the returns of “Shops” (27.22%) exceed the returns of “Utils” (-28.22%) by 55.44%.41 A t-value of 19.78 indicates that the means are statistically different.42

40 With the ex-date returns, 5% significance was used. A significance level of 10% is used here because there are so few 250-day BHRs that are significant at the 5% level. 41 As before, the phrase “returns of “Shops” (27.22%)” is shorthand for “the average 250-day buy- and-hold return of firms in the industry classification "shops” is 27.22%.” 42 As with the ex-date industry returns in the earlier section, a two-sample t-test is conducted here.

64 In the Ten Industries category, “Shops” (27.64%) leads “Money” (21.98%) and these means are significantly different as evidenced by the two-sample t-test of 2.14. In the Twelve Industries category the returns of “Chems” (51.29%) exceed the returns of “Shops” (-25.91%) more than two to one. Again, the two-sample t- test of 3.07 indicates the means are statistically different at the 5% level. In the Seventeen Industries category the returns of “Other” (28.97%) exceed the returns of “Durable” (-21.65%) and the two-sample t-value is 32.28. The Thirty Industries have few significant returns, but “ElcEq” (60.49%) exceeds “Servs” (33.52%) by nearly two to one. The two-tailed t-test of 2.43 indicates that these are statistically different. In the Thirty-eight Industries category the returns of “Elctr” (41.51%) exceed the returns of “Garbg” (-38.00%) and “Food” (-20.04%) with two-sample t-statistics of 14.42 and 10.64 respectively. And finally in the Forty-eight Industries category the returns of “ElcEq” (60.49%) exceed the returns of “Books” (-15.25%) with a two-sample t-value of 7.35. The returns of “Food” were also on the low end (–17.03%), but not statistically different than zero at the 10% level (t-value = -1.23). In conclusion, there are significant differences in the 250-day BHRs across industry classifications, but there is no clear pattern as there was for the average ex-date returns. Most classifications do not have significant BHRs and most classifications that do have significant BHRs are only significant at the 10% level. Therefore any conclusions drawn concerning industry would have little evidentiary support.

3.1.15.2 Delisting Threat and Motivations for Reverse Stock Splits

The average 250-day BHRs for different motivations are provided in Table 25. The BHRs are highest for the subsample of firms threatened with delisting (51.92% with a t-value of 3.72). The only other significant mean 250-day BHR

65 was for the “No Reason Given” (16.96% with a t-value of 2.88). For market adjusted returns (BHARs), using the CRSP equally weighted index as the market, there are no motivational groups with a mean BHAR that is significantly different than zero.

3.1.15.3 Price Considerations

To continue the motivation analysis we assume that, regardless of their public statements, low priced firms reverse split their stock for regulatory reasons. We assess whether there is a long-run performance difference for firms that have compliance problems and those who do not. Table 25 shows the results of the testing. Stocks with pre-split prices below $1.00 have an average 250-day BHR of 27.07% (t-value=4.82) and an average BHAR of 0.13% (t-value=0.56). Stocks with pre-split prices above $1.00 have an average 250-day BHR of 2.74% (t-value=0.41) and average 250-day BHAR of -0.11% (t-value=-0.25). A two-sample t-test of equal means is calculated as Y −Y T = 1 2 s2 N s2 N . 1 / 1 + 2 / 2 T=32.11 for the BHRs and T=94.39 for the BHARs indicating that the means are not equal in either case. However, finding a significant difference in means for the two average BHARs that are indistinguishable from zero seems questionable. Since the average BHARs of firms with prices above $1.00 and below $1.00 are both insignificant, there is no evidence that firm that split for regulatory reasons perform any differently than firms that split for other reasons.

3.1.15.4 Size of the Firm

Larger firms have more resources and they may be able to recover from the reverse stock split over time. The 250-day buy-and-hold raw returns (BHRs) and the 250-day market adjusted returns (BHARs) are regressed on the size of the firm. The CRSP equally weighted index is used as a proxy for the market and the

66 number of shares multiplied by the price per share is used as a proxy for company size. Neither the BHRs, nor the BHARs, exhibit any significant relationship with size. The regression coefficient is zero to four decimal places for both regressions. The conclusion from these tests is that firm size has no effect on 250- day BHRs or BHARs. The regression results are presented in Table 26.

3.1.15.5 Consolidation

Finally, the long-run returns are regressed on consolidation. As stated before, a firm reverse splitting 10 for 1 has a dimmer outlook on the future than a company reverse splitting 2 for 1. In addition, if larger consolidation is indicative of a lower quality firm, then the long-run returns should be lower for firms with larger consolidation. This idea is tested by regressing 250-day returns on number of shares consolidated:

R0 = α + β(number of shares consolidated). Both BHRs and BHARs are tested. The results are presented in Table 26. When the 250-day BHR is the dependent variable, the regression coefficient is -0.0001 with a t-value of -1.21. When the 250-day market adjusted BHAR is the dependent variable, the regression coefficient is -0.0013 with a t-value of -1.34. Both regression coefficients are outside of normal acceptance levels but a weak argument can be made that the negative coefficient indicates that larger consolidations result in a lower 250-day return. Related to consolidation is the target price of the consolidation. A larger price target may indicate that the firm has a dim outlook on the future. Both BHRs and BHARs are regressed on the post split price and no significant relation is found. The long-run returns are not related to the post split price. The regression results are presented in Table 26.

67 3.1.15.6 Long-run Analysis Concluding Remarks

This concludes the analysis of the long-run returns after a reverse stock split. In summary, the average BHR and BHAR increase as the year progresses. However, the 8.8% 250-day average raw return is offset by the average CRSP equally weighted market index return of 32%. The average CRSP value weighted return of 12.3% is also higher than the average BHR, but when the BHAR is calculated using the value weighted index, the 250-day BHAR is not significantly different from zero. The 250-day average BHAR calculated using the matched firm method is also insignificant. Different industry groupings have significantly different 250-day returns, as do different motivational groupings. However, no relation to long-run returns was found for size, consolidation, or post split price.

3.2 Market Movements and Reverse Stock Splits

3.2.1 Hypothesis

Reverse stock splits, in general, are preceded by a period of negative returns for the firm. As the price of a firm declines, it sometimes becomes necessary to reverse split the stock in order to maintain compliance with market listing requirements. Since a reverse stock split is generally associated with declining prices, it seems likely that there would be more reverse stock splits in periods immediately following substantial market declines. A cursory review of Table 1, Reverse Stock Splits by Year, reveals no obvious relation between market performance and the frequency of reverse stock splits, but further analysis is warranted. And even if no relation is found between the frequency of reverse stock splits and prior market performance, it is possible that the post reverse stock split performance of firms may vary with the market environment that preceded the reverse stock split. This issue is examined below as well.

68 First, the relation between the frequency of reverse stock splits and lagged market returns is examined. The null hypothesis here is:

The frequency of reverse stock splits is not related to prior market performance.

3.2.2 Testing

To test this hypothesis, the frequency of reverse stock splits in a given month is regressed on a lagged market return. The frequency of reverse stock splits in a given month is calculated as the percentage of total firms in the market that reverse split in that month. The independent variable is either the 12 month or the 24 month lagged buy and hold return on the CRSP equally weighted index. The equally weighted index is used because it gives greater weight to small firms, and these are the firms that are most likely to be candidates for reverse stock splits. The regression equations is:

FRSt = α + βRm,lagged + εt where FRSt is the percent of total firms that reverse split in month t and Rm,lagged is either the 12 or 24 month lagged buy and hold return on the Equally Weighted CRSP Index. Results are presented in Table 27. The relationship is marginally significant (t=-1.87) when the market return is lagged 12 months, and more significant when the market return is lagged 24 months (t-value –3.67). This suggests that the frequency of reverse stock splits is more pronounced after market downturns. To assess whether post reverse split stock returns are related to market conditions prior to the reverse stock split, the 250-day BHRs are regressed on lagged market BHRs.

BHRi,250 = α + βRm,lagged + εt where BHRi,250 is the 250-day BHR of firm i after the reverse stock split, and

Rm,lagged is either the 12 or 24 month lagged buy and hold return on the CRSP equally weighted index. The conjecture is that firms that are forced to reverse stock split partly because of price declines attributable to general market declines

69 may perform better on average after the split than firms that reverse split due to idiosyncratic reasons only.

3.2.3 Results

The regression results are provided in Table 28. As conjectured, there is an inverse relationship between 250-day BHRs and lagged market returns. In both cases (12 month and 24 month lagged market returns), the relationship between the BHRs and the lagged market return is negative and significant, indicating an inverse relationship between long run BHRs on stocks that reverse split and lagged market returns. Continuing along this path, the 250-day market adjusted BHARs obtained using the equally weighted CRSP Index are regressed on lagged market returns. These results are also provided in Table 28. Here too, a significant inverse relationship is found between the 250-day BHARs and the lagged market returns, for both the 12-month and the 24-month lag. In summary, evidence is provided that the frequency of reverse stock splits is higher subsequent to market declines. Also, it appears that stocks that reverse split their stock after an extended market decline tend to do better than firms that reverse split after a period where the market has performed well.

3.3 Financial Distress and Future Performance

3.3.1 Hypothesis

The performance after a reverse stock split varies widely by firm and the purpose of this section is to investigate pre-split information that may be useful for forecasting post split performance. Reverse stock splits are thought to be negative signals to the market because, on average, reverse splitting firms have been shown to have abnormal negative returns around the announcement date and the ex-date. In addition, the average long run BHAR following a reverse stock split has been shown to be negative on average. However many firms in the

70 sample do very well after the reverse stock split. Depending on how “do very well” is defined, 1/4 to 1/3 of the sample firms do well after the reverse stock split. Table 4 shows “winners” versus “losers” with various holding periods and various return hurdles for being classified as a winner or loser. The lowest hurdle is 20% raw returns and the highest hurdle is 200%. The holding periods range from 30 days to 250 days (about one year of trading days). Since some firms in the sample do well after the reverse stock split, this section of the analysis explores the possibility that there may be some public information that is useful for forecasting which firms will do better (or worse) after the reverse stock split. For example, firms that are generally healthy might be expected to outperform firms that are generally unhealthy. The financial distress literature provides some financial ratios that can be used to infer the health of a particular stock. If these ratios are predictive of firm performance after a reverse stock split, then investors may be able to exploit this information. The hypothesis to be tested, stated in the null, is:

The financial health of a stock does not help to predict firm’s returns after a reverse stock split.

3.3.2 Testing

To test this hypothesis, the ex-date returns and the long run returns are regressed on the financial distress variables posited by Altman (1968), Theodossiou, Kahya, Saidi, and Philippatos (1996), and Kahya and Theodossiou (1999). In these articles, the authors modeled financial distress using debt, liquidity, and operational performance ratios in a discriminant function framework. The researcher collects many variables and seeks to determine which are useful in forecasting bankruptcy. Compustat variables include: debt to assets, operating income to assets, inventory to sales, sales, assets, net working capital to assets, current assets to current liabilities, quick assets to current liabilities, EBIT to assets, retained earnings to assets, long-term debt to total assets, market value of equity to long-term debt, and fixed assets to assets. Regression analysis is used

71 to determine which variables are most useful in identifying winners and losers. The coefficients (βs) are estimated using maximum likelihood and they show whether an independent variable is positively or negatively related to the dependent variable. The model is

Ri = β0 + β1dta + β2opincta + β3invtsls + β4lgsales + β5lgassets + β6nwcta + β7cacl

+ β8qacl + β9ebita + β10reta + β11ltdta + β12mveltd + β13fata.

The model is used in several regressions where Ri is either the ex-date raw return or the long run BHR as calculated before. The independent variables remain the same for each regression and dta is debt to assets, opincta is operating income to assets, invtsls is inventory to sales, lgsales is the log of sales, lgassets is the log of assets, nwcta is net working capital to assets, cacl is current assets to current liabilities, qacl is quick assets to current liabilities, ebita is EBIT to assets, reta is retained earnings to assets, ltdta is long-term debt to total assets, mveltd is market value of equity to long-term debt, and fata is fixed assets to assets.

3.3.3 Results

A correlation matrix of these variables is presented in Table 30 and the regression results are presented in Table 31. The correlation matrix shows that the ex-date returns are related to the following firm specific variables: assets, sales, the EBIT to assets ratio, and the operating income to assets ratio. The 250- day BHRs are positively related to the EBIT-to-assets ratio, the operating-income- to-assets ratio, and the net-working-capital-to-assets ratio. The 250-day BHRs are inversely related to the retained-earnings-to-assets ratio, the debt-to-asset ratio and the long-term-debt-to-assets ratio.43 Given the coefficients in the correlation matrix, it seems that firms with larger sales, more assets, larger operating-income-to-assets, and more earnings- to-assets outperform on the ex-date, and companies with higher income, higher earnings, lower net working capital and lower debt do better in the long run.

43 For the long run return results presented in the correlation matrix and the dependent variable in the regressions is the BHR. Using BHARs (market model adjusted or market-adjusted) yields the same results with only slight changes in the p-values.

72 The results of the full model with all independent variables included are presented in Table 31. The regression coefficients are shown for BHRs of 30, 60, 90, 125, and 250 days. Regressing the returns of firms that reverse stock split on financial distress variables is unique in the literature but a related study, Kiang, et al (2005), uses Altman’s Z and a type of neural network to forecast which firms will eventually go bankrupt. In this model, lgsales has the only significant coefficient but even then it is only significant for the ex-date returns. It is not significantly correlated to any of the long run returns. Since the correlation matrix reveals that there is multicollinearity in many of the variables, a reduced model is proposed based on the work of Theodossiou, Kahya, Saidi, and Philippatos (1996). These authors use discriminate analysis to find that the variables DA, OIA, ITS, and S are significant predictors of financial distress.44 Using only these variables the reduced model becomes:

Ri = β0 + β1dta + β2opincta + β3invtsls + β4lgsales. The results of this regression are given in Table 32. For the ex-date returns

(RI is the ex-date return) there is a positive relation to opincta (operating income to assets) at the 1% significance level (t-value = 2.54) and there is positive relation to lgsales (sales) at the 10% significance level (t-value = 1.65). There is no significant relation to dta or invtsls. The regression results presented use the ex-date return as the dependent variable, but the results are similar if the dependent variable is market adjusted ex- date return or market model adjusted ex-date return. The market adjusted ex-date returns are significantly related to opincta (t-value 2.54) and lgsales (t-value 1.59) and the market model adjusted ex-date returns are significantly related to opincta (t-value 2.40) and lgsales (t-value 1.83). When the dependent variable is the 250-day (12 month) BHR, then the only marginally significant relations are with opincta (t-value = 1.65), and dta (t-value =

44 Discriminate analysis is used to eliminate variables form a model. Theodossiou et al also find change in employment to be a significant predictor of financial distress.

73 -1.59). When the 250-day BHAR is used as the dependent variable, the only significant (inverse) relationship is with dta. Both the market adjusted BHARs and the market model adjusted BHARS are significantly related to dta with t-values of -2.14 and -2.11 respectively.

3.3.4 Concluding remarks

Even though the correlation matrix shows a significant relation between returns and some of the accounting variables the conjecture that reverse stock splits are a negative signal to the market is not strongly challenged by these results. Many of the accounting variables are significantly correlated with each other and isolating the variables that have the most effect on returns is difficult. In the full model, only one variable was significantly related to ex-date returns and there were no variables with a significant relationship to the long run returns. In a reduced model, variables were used that have been shown to be significantly related to financial distress. Significant relationships were found for both ex-date and long run returns. Ex-date BHRs and BHARs are found to be related to operating-income-to-assets and sales. Long run BHARs are related to debt-to-assets and long run BHRs are related to operating-income-to-assets. These results suggest that in the short-run, firms with better sales and better operating income (in relation to assets) will fare better on the days around the reverse stock split. In the long run, a company with less debt will probably not be hurt by the reverse stock split as much as companies with more debt.

3.4: NASDAQ Regulatory Changes

As discussed earlier, many of the firms in this sample reverse split their stock in order to comply with NASDAQ’s minimum bid price requirement. If the minimum bid price requirement is relaxed, it is likely that fewer firms would reverse

74 split their stock and fewer firms would be delisted. The NASDAQ rule change of 2001 provides an opportunity to test this conjecture. The rule change was instituted after a culmination of extraordinary market conditions. The NASDAQ, home of many Internet start-up firms, suffered a substantial loss in the value during 2000 and 2001. 45 Some viewed this severe market decline as a mandate for regulatory changes. As revisions in minimum bid price rules were being discussed, the impetus for change was accelerated by the September 11, 2001 attacks on the World Trade Center and the Pentagon. The market was in disarray and in an effort help firms make it through the difficult times, NASDAQ announced on September 27, 2001 a suspension of minimum bid price rules and public float requirements. A pilot program soon followed in which the minimum bid price and continued listing requirements were temporarily relaxed. The following sections describe the rules before the pilot program and the rules during the pilot program. Two tests are conducted to test the effectiveness of the rule change.

3.4.1 NASDAQ Listing Rules – Before The Pilot Program

Each market demands that its members meet certain minimum standards that its policy makers set forth. NASDAQ has two sets of standards: a National Market standard for larger companies and a Small Cap standard for smaller companies. A firm must meet the more stringent “initial listing requirements” to be accepted into the market and maintain the somewhat less stringent “continued listing requirements” to remain in the market. For example, to be listed on the NASDAQ Small Cap Market, a firm must have a minimum bid price of $4.00 and 1 million shares on the market. To stay listed the firm must maintain a minimum bid price of $1.00 and 500,000 shares on the market. There are additional listing

45NASDAQ, as proxied by the NASDAQ 100 Trust (QQQQ) lost 75.48% of its value from March 27, 2000 to September 21, 2001. This percentage was calculated as (ending price minus beginning price) divided by (beginning price) and the prices are adjusted for splits and dividends. Source: Yahoo.com.

75 requirements concerning assets, capitalization, net income, market value, number of market makers, number of shareholders, etc. These requirements are given in Table 33. Prior to the rule change, a firm that failed to meet the continued listing standards for 30 consecutive days would receive a citation from NASDAQ. The firm then had a 90-day “grace period” to bring their stock back into compliance. If a firm’s infraction was low price, then it could use this 90-day period to consolidate shares and thereby increase the share price of the stock. During the 90-day period a company that did not meet continued listing requirements for at least 10 consecutive days was subject to delisting. During the suspension period beginning September 27, 2001, companies were not cited for failing to meet minimum bid price or public float requirements, and therefore not subject to delisting. When the suspension period ended in January 2002, NASDAQ immediately implemented a pilot program that extended the grace period to 180 days so that companies had more time to bring their bid price and public float into compliance.

3.4.2 The Pilot Program

Under the pilot program, “Small Cap” firms that failed to meet minimum bid price requirements or public float requirements for 30 consecutive days were given 180 days to bring their stock back into compliance. If the firm’s minimum bid price were still out of compliance at the end of this 180-day grace period, the company was given another 180 days to comply if it complied with other listing criteria. It had to have either • $5 million in equity, • $50 million market value, or • $750,000 in net income from continued operations in the most recent fiscal year or two of the last three fiscal years. At the end of the second 180-day grace period, firms with non-compliant price or float were given an additional 90 days to comply if the equity, market

76 value, or net income criteria (above) were met. The firm was finally subject to delisting at the end of this 450-day period. The larger companies listed on NASDAQ’s “National Market System” (NMS) were also given consideration under the pilot program. Their grace period was increased from 90 days to 180 days. A NMS company still only had one grace period in which to meet continued listing requirements (for at least 10 consecutive days). But, if the NMS firm could not comply with the NMS requirements, it had the option of participating in the Pilot Program (as described above) if it voluntarily lowered its status to “Small Cap.” As shown in Table 33, the requirements for the NMS issuers are more complex than the requirements for the Small Cap market issuers. There are three ways for a NMS company to gain initial acceptance and there are two ways to qualify for continued listing.

3.4.3 Hypotheses

Since the majority of firms that execute reverse splits cite regulatory reasons as their motivation, the action is probably not voluntary. The consolidation is a costly and inconvenient endeavor – like adding pollution controls. It is not something the company chooses to do to enhance shareholder wealth. If the regulations were relaxed, the firm would probably not choose to bear the cost and inconvenience of the reverse stock split. Therefore if the rule change and Pilot Program were effective, the number of reverse stock splits should decline from what it would have been under the old rules. In addition, if the companies were given more time to bring the firm into compliance and if the overall market condition (rather than firm performance) were responsible for temporary compliance issues, then one result of the Pilot Program should have been fewer delistings. The hypotheses, stated in the null, are:

The moratorium on the minimum bid price rules had no effect on whether or not a NASDAQ firm reverse split its stock.

77 The moratorium on minimum bid price rules did not result in fewer delistings for NASDAQ firms whose stock had a bid price of less than $1.00

3.4.4 Reverse splits and the Moratorium

The number of reverse stock splits before and after the moratorium is calculated as a percentage of low-priced firms. The time “before” the moratorium is defined as month t-30 to month t-7 where month t=0 is September 2001. This filters out any noise (rumor activity) that may be present in the data in the 6 months prior to the rule change. The time “after” the moratorium is defined as t=1 to t=24. A “low-priced” firm is defined as one whose bid price has been below $1.00 for 30 consecutive trading days. This was the pre-moratorium delisting flag for NASDAQ and in the “after” period it represents the firms that would have been out of compliance before the rule change. The monthly number of low-priced firms, number of reverse stock splits, and ratios of reverse-splits-per-low-priced-firms are shown in Table 34. This table also shows rolling averages. The mean percentage of reverse stock splits per low-priced firm is calculated for months t-30 to t-6 and compared to the mean percentage for months 1 to 24. The Wilcoxon non-parametric test of equal means is calculated to be 1644. This corresponds to a two-tailed p-value of 0.30. This number means that when the median difference of reverse stock splits is zero, then there is a 30% chance that the difference found would be as much or greater than calculated. Lower p-values provide stronger evidence of unequal means. A p-value of 0.30 is not enough evidence to say that the percentages of reverse stock splits before and after the rule change are different. As measured by the number of reverse stock splits, the intent of the rule change, to help firms make it through a systemic decline was not realized. If the rule change were effective, then the post-moratorium percentage of reverse stock splits would be lower than the pre-moratorium percentage of revere stock splits.

78 However, after the rule change, firms were still engaging in this costly and often involuntary activity at about the same rate.

3.4.5 Delistings and the Moratorium

If the rule change was successful then there should be fewer companies delisted in the months following the change. To ascertain whether this is the case, a comparison is made between the number of firms that were delisted before the rule change and the number of firms that were delisted after the rule change. This number is calculated as a percentage of firms that would have been delisted under the old rule. The number of firms that would have been delisted in a period is the number of firms that have had a price of less than $1.00 for 120 consecutive days. A firm was out of compliance for 30 days, warned, and then (should have been) delisted 90 days later. If the price was greater than $1.00 for 10 consecutive days during this time, the firm was not counted because it was considered to be back in compliance. If the rule change were effective, then there should be fewer delistings after the rule change. In the 24 months prior to the rule change (t-30 to t-7) there were 187 NASDAQ firms that were delisted due to low share price (out of 240 firms eligible for delisting). In the 24 months after the rule change there were 133 NASDAQ firms that were delisted because of share price (out of 700 meeting old delisting criteria). Before the rule change 78% of the firms that met the low price delisting criteria were delisted. After the rule change only 19% of the firms that met the old low price delisting criteria were delisted. It seems that the rule change had an immediate impact on delistings. The delistings before and after the rule change are shown on Table 35. Also shown is the delistings in six-month increments. After the rule change the delisting ratio drops from an average of 32% (for the six-month increments) to single digits. The two-sample t-test of equal means yields a t-value of 967 indicating that the means are not equal at the 0.0001 level. The number of delistings shown for t-30 to t-7 is the sum of the delistings in the six-month periods.

79 The number of “firms that should have been delisted” for t-30 to t-7 is not the sum of the values in the six-month increments because firms could be eligible for delisting in more than one period. In conclusion, the 2001 NASDAQ moratorium and subsequent pilot program did not seem to have an impact on the percentage of firms that reverse split their stock. It did, however impact the number of delistings. The pilot program allowed firms up to 450 days to bring their price back into compliance. If the rule change were not effective, it would merely delay delistings and a wave of delistings would occur at the end of the first 450-day grace period after the change. This is not the case. In the 15+ months (450 days) after the rule change there was no new wave of delistings. If the goal of the regulatory change was to help companies weather the market downturn, then according to these delisting measurements, it was a success. On the other hand, there was no reduction in reverse stock splits after the rule change, and as such, the two tests conducted here do not provide collaborating evidence.

80 CHAPTER 4: CONCLUSIONS

4.1 Overview

Despite numerous studies that have shown negative market reactions to reverse stock splits, firms continue to consolidate their shares. In fact, the total number of reverse stock splits has nearly doubled since 1992. The increase in available data, some unusual market conditions in 2001, and a NASDAQ rule change provide an impetus to re-examine the reverse stock split. Stock price reactions to reverse stock splits are examined in relation to various firm characteristics and market movements. This analysis is followed by an examination of financial ratios and variables that may be predictive of post split performance. Finally, the NASDAQ minimum bid price rule change of 2001 is explored to determine if it had an impact on reverse stock splits or delistings.

4.2 Results and Suggestions for Further Research

4.2.1 Stock Price Reactions – Announcement Date

The announcement day results of this study confirm previous research. On average, there is a negative cumulative abnormal return for the announcement day periods examined. Announcement dates are classified two ways. Firms that issue a press release before the ex-date constitute one sample, and firms that issue no press release until the ex-date are placed in the second group. When the first news is on the ex-date the negative and significant reaction on that “announcement-day” is followed by positive and significant post announcement day returns. In the case where there is no previous announcement and the ex- date is the announcement day, this evidence may suggest that an initial AD overreaction to bad news is followed by a correction.

81 4.2.2 Stock Price Reactions – Ex-date

One unique contribution of this study is the finding that the market appears to react stronger to the reverse stock split on the ex-date than on the announcement day. The ex-date returns are greater in absolute magnitude than the return on the announcement day for firms that announce prior to the ex-date. This finding is somewhat peculiar given that the stock becomes more liquid and more tradable after the price is adjusted to levels above $1.00. Many institutions, as a matter of policy, shy away from low-priced stocks. In addition, many brokerage firms do not allow their customers to trade low-priced shares on margin. Why the wider availability of the stock would lead to the large negative ex-date reaction is puzzling. More research is required to determine the exact cause of the reaction. One possible explanation is increased short activity on the stock, as shorting would generally have been impossible at pre-consolidation price levels.

4.2.3 Stock Price Reactions – Firm Characteristics

Further insight into the stock price reaction is gained by considering firm specific characteristics of firms in the sample. For example, significant differences in stock price reactions across some industries are detected. In addition, the firm’s motivation for reverse splitting is a factor in how the firm performs on the ex-date. In an apparent conflict with the earlier literature, the results of this study show firms reverse splitting for regulatory reasons underperform (rather than outperform) firms that split for other reasons. Specifically, this study reveals that firms that reverse split for regulatory reasons perform worse than do firms that split for any other reason. This finding is confirmed by comparing firms with pre-split prices below $1.00, which are assumed to reverse split for regulatory reasons, to firms with pre-split prices above $1.00. The relation between a firm’s ex-date performance and its post-split target price is also examined. Findings indicate firms with higher post split prices (greater than $5.00) tend to perform better on the ex-date.

82 The ex-date return is not related to the size of the firm, but there is a mild relationship with consolidation. The more shares a firms consolidates, the worse the firm performs on the ex-date.

4.2.4 Stock Price Reactions – Long Run

The mean 250-day BHR (250 days from the ex-date) is positive, but the median 250-day BHR is negative. The mean is skewed due to a small number of reverse stock split firms having large positive returns for the 250-day period. Two tests of long-run 250-day BHARs reveal no significant results, but one test (the market adjusted BHAR using the CRSP equal weighted index as the market adjustment) shows negative and significant BHARs. The BHRs are examined as related to various firm characteristics and it is found that the 250-day BHRs vary some by industry. However, there is no significant size, price, motivation, or consolidation effect for the 250-day BHRs.

4.2.5 Market Movements

Evidence is presented that the frequency of reverse stock splits (as a percentage of total firms) increases after periods of market decline. In addition, it is found that firms that reverse split their stock after a period of market decline have higher 250-day BHRs than firms that reverse split their stock after a period where the market has performed well. In other words, a market effect of sorts is detected. Specifically, the 250-day BHARs are greater for firms that reverse split following periods of substantial market declines. This may suggest an overreaction followed by a correction, but further research is needed to understand why the BHARs vary as they do in different market environments. Further research could also be conducted as to a possible trading strategy to take advantage of this difference in these BHARs.

83 4.2.6 Projecting Post Split Performance

Variables taken from the financial distress literature are related to ex-date returns and BHARs. This is unique in the literature. Some of these variables are significantly related to post split returns. The results suggest that firms with better sales performance and higher operating-income-to-assets have better ex-date returns. In the long run, companies with lower debt-to-assets do better after the reverse stock split. Operating income expressed as a percent of assets is also positively related to the 250-day BHRs. These results provide evidence that the post reverse stock split performance is dependent on the health of the company at the time of the reverse stock split. A generally healthy company that reverse splits its stock should not be automatically considered to be a bad investment. There is a correlation between financial distress and reverse splits, but it is unclear if there is causation. More research is required to determine if there is a way for investors to exploit this relationship.

4.2.7 NASDAQ Rule Change

The NASDAQ moratorium on minimum bid price requirements and the extended compliance time did not impact the number of reverse stock splits. This finding suggests that the rule change may not have had the intended effect. Firms were still assuming the cost and inconvenience of the reverse stock split at a relatively high rate. On the other hand, the number of delistings (as a percentage of firms that would have been delisted under the old rules) dropped significantly after the rule change. Using this metric, the rule change was successful because more firms were able to remain listed on the market. Further research could be conducted into other rule changes in other markets at various times to see if NASDAQ or another market has more effective policies.

84 4.3 Contribution

This dissertation is useful to investors because the reverse stock split is shown to be a conditional event. It is true that on average, the reverse stock split is followed by negative BHARs, but if a firm is in a non-discretionary industry, has a high target price, little debt and a strong operating-income-to-assets ratio, it may do better in the long run. This dissertation is useful to regulators because it provides a performance measure to assess whether the rule change was effective. If the goal of the regulatory change was to retain firms in the market, then the rule change was successful.

85

APPENDIX A: TABLES

86 Table 1 Forward and Reverse Stock Splits and Split Ratios 1962-2002 This data, from the Center for Research on Security Prices is also represented in Appendix B Figure 1 and Figure 2.

Reverse Stock Splits by Year

1962 2 1976 11 1990 106 1963 5 1977 6 1991 99 1964 11 1978 9 1992 157 1965 5 1979 6 1993 114 1966 4 1980 5 1994 92 1967 1 1981 15 1995 101 1968 3 1982 25 1996 96 1969 3 1983 36 1997 106 1970 0 1984 36 1998 182 1971 5 1985 48 1999 114 1972 6 1986 30 2000 73 1973 12 1987 71 2001 124 1974 5 1988 56 2002 85 1975 7 1989 70

Split Ratios of Stock Splits 1962 to 2002

Reverse Stock Splits Forward Stock Splits

Split ratio Split ratio Number of splits Number of splits 1 for x x for 1

1 for 2 or less 199 <1.25 for 1 359 1 for 2.01-3.49 255 1.25 for 1 (5 for 4) 1,339 1 for 3.5-4.99 281 1.251-1.5 for 1 543 1 for 5 364 1.5 for 1 (3 for 2) 4,848 1 for 5.01-9.99 188 1.51-1.99 for 1 74 1 for 10 398 2 for 1 6,673 1 for 10.0-24.99 161 2.01-2.99 for 1 117 1 for 25-34.99 27 3 for 1 615 1 for 35-49.99 44 3.01-4.99 for 1 134 1 for 50 or more 65 5 or more for 1 126

87 Table 2 Average Stock Price per Year and Yearly Stock Splits 1962-2002 This table provides the yearly number of firms in the market, the number of forward stock splits and the number of reverse stock splits. Percentages are a percent of total.

Total Firms In Percent Forward Stock Reverse Stock Percent Reverse Average Price Year The Market Forward Stock Splits Splits Stock Splits of All Firms Splits 1962 1606 71 4.42 4 0.25 23.31 1963 2071 61 2.95 5 0.24 25.18 1964 2117 116 5.48 11 0.52 26.27 1965 2164 146 6.75 5 0.23 27.16 1966 2191 174 7.94 4 0.18 26.30 1967 2206 119 5.39 1 0.05 30.24 1968 2218 241 10.87 3 0.14 33.41 1969 2297 220 9.58 3 0.13 28.83 1970 2413 50 2.07 0 0.00 19.96 1971 2520 110 4.36 5 0.20 23.14 1972 2929 159 5.43 6 0.20 23.15 1973 5672 210 3.70 12 0.21 14.65 1974 5393 101 1.87 5 0.09 10.54 1975 5255 124 2.36 7 0.13 11.15 1976 5249 264 5.03 10 0.19 12.85 1977 5215 296 5.68 6 0.12 13.22 1978 5093 410 8.05 9 0.18 14.39 1979 5010 368 7.35 6 0.12 15.24 1980 5053 548 10.85 5 0.10 16.34 1981 5389 586 10.88 15 0.28 15.77 1982 5544 310 5.59 25 0.45 13.28 1983 5917 883 14.92 36 0.61 17.19 1984 6424 414 6.44 36 0.56 14.10 1985 6424 587 9.14 49 0.76 15.29 1986 6601 833 12.62 29 0.44 16.99 1987 7125 682 9.57 70 0.98 15.77 1988 7170 267 3.72 55 0.77 13.85 1989 6976 402 5.76 70 1.00 15.36 1990 6871 262 3.81 106 1.54 13.24 1991 6804 304 4.47 98 1.44 14.74 1992 6977 490 7.02 157 2.25 15.96 1993 7419 554 7.47 115 1.55 17.87 1994 8151 422 5.18 92 1.13 17.37 1995 8373 512 6.11 101 1.21 19.11 1996 8840 642 7.26 98 1.11 21.12 1997 9152 745 8.14 104 1.14 23.36 1998 9035 737 8.16 182 2.01 25.63 1999 8580 488 5.69 114 1.33 25.84 2000 8398 512 6.10 73 0.87 25.93 2001 7818 230 2.94 124 1.59 25.22 2002 7289 222 3.05 124 1.70 26.00 2003 6883 244 3.54 90 1.31 28.41 2004 6769 329 4.86 43 0.64 34.97 2005 6757 349 5.17 57 0.84 36.10

Source: Center for Research in Security Prices

88 Table 3 Margin Requirements Margin requirements refer to how much money an investor may borrow to purchase securities. A margin requirement of 80% means that the investor must use 80% of his own money when purchasing securities. Margin requirements may be different for different brokers. This table shows how SEC minimum margin requirements have changed over the years.

Date Margin Requirement Before August 5, 1958* 50% August 5, 1958 70% October 16, 1958** 90% July 28, 1960 70% July 10, 1962 50% November 5, 1963 70% June 8, 1968 80% May 6, 1970 65% December 6, 1971 55% January 3, 1974 50% Present 50% *Prior date unknown. **Highest level in 11 years. Source: California Department of Finance.

89 Table 4 Winners Versus Losers After Reverse Stock Split Various returns are shown below. A “winner” has a buy and hold raw return (BHR) greater than the return in the left column and a “loser” has a BHR less than that rate. The BHRs are computed for various trading days as shown. The numbers of winners and losers varies because firms with the exact return are not counted. As a reference point, the average return for the CRSP equal-weighted index in this study is about 32%.

Trading Days Return 30 60 90 125 250 20% winners 620 560 539 533 456 losers 1282 1342 1363 1369 1446 % winners 33% 29% 28% 28% 24%

30% winners 587 517 509 488 403 losers 1315 1385 1393 1414 1499 % winners 31% 27% 27% 26% 21%

50% winners 534 455 436 398 318 losers 1368 1447 1466 1504 1584 % winners 28% 24% 23% 21% 17%

100% winners 424 341 297 264 181 losers 1478 1561 1605 1638 1721 % winners 22% 18% 16% 14% 10%

150% winners 347 266 204 183 122 losers 1555 1636 1698 1719 1780 % winners 18% 14% 11% 10% 6%

200% winners 293 196 142 124 82 losers 1609 1706 1760 1778 1820 % winners 15% 10% 7% 7% 4%

90 Table 5 Summary of Stock Split Literature and Key Findings

Summary of studies of reverse stock splits Authors Year Type of Firms Period Key Findings Studied Gleason, Rosenthal, and Lee 2004 1,072 NYSE, AMEX, and 1992-2001 Negative signal to the market. Reverse splits are a confirmation of Nasdaq reverse splits poor performance, especially when done for the purpose of staying listed. Kim, Jain, Jiang, Mcinish, and 2003 1,363 NYSE, AMEX, and 1983-2002 Presplit prices of $5 have positive abnormal returns. Postsplit prices Wood Nasdaq reverse stock greater than $5 have increased institutional ownership splits

Jing 2002 116 reverse splits on the 1991-2001 Negative abnormal CARs around announcement date (-4%) and the Hong Kong exchange ex-date (-10%). 12 month performance of reverse splits is inferior to matched firms (-10.6% vs. 1.21%). Liquidity improvements evidenced by greater adjusted volume post reverse split. Transactions costs (bid ask spread) reduction after reverse split.

Vafeas 2001 229 reverse splits 1985-1993 Reverse split firms have inferior earnings to matched firms in pre split years. Investors expect inferior earnings to continue. Desai and Jain 1997 5596 forward stock splits 1976-1991 Forward splits, on average, have one-year BHAR of 7.05% and three- and 76 reverse stock split year BHAR of 11.87% after the announcement month. Reverse splits, announcements on average, have one-year BHAR of -10.76% and three-year BHAR of -33.90%. Underreaction to split event is cited.

Han 1995 136 NYSE, AMEX, 1963-1990 Liquidity is improved after reverse stock split Nasdaq reverse splits Maloney and Mulherin 1992 446 Nasdaq 1.25 for 1 or 1984-1996 Reverse splits are followed by negative abnormal returns. Forward greater splits splits do not increase volume. Forward splits result in enlarged ownership base, increased number of small trades, small buy orders by individuals after split.

Peterson and Peterson 1992 1057 NYSE, AMEX, 1973-1989 Risk of returns of reverse split stocks decline after reverse split. Nasdaq Lakonishok and Lev 1987 1015 stock splits and 1963-1982 Pre split price (and dividend) run up precedes forward splits. 1257 stock dividends Managers want to increase individual investor ownership. Managers have a target price - market wide average and industry averages. Percentage price appreciation during the year before split

Lemoureux and Poon 1987 213 forward splits and 49 1962-1985 Reverse stock splits decrease the tax option value of stocks by reverse splits. NYSE decreasing volatility. Forward splits result in enlarged ownership base, and AMEX. increased number of small trades, small buy orders by individuals after split. Forward stock split volatility increases.

Spudeck and Mayer (1985) 1985 21 NYSE and AMEX reverse splits Follow-up to Woolridge and Chambers indicates that there are no negative BHARs around announcement date or ex-date. Grinblatt, Masulis, and Titman 1984 1,762 splits and 1967-1976 Reverse splits are followed by negative abnormal returns and forward dividends splits are followed by positive abnormal returns. Stock dividends can also be considered stock splits.

Woolridge and Chambers 1983 57 NYSE and AMEX 1973-1983 Reverse splits are not justified because they are followed by negative reverse splits BHARs around announcement date and ex-date Radcliffe and Gillespie 1979 NYSE, AMEX 1960-1976 Reverse splits, in general, are detrimental to shareholder wealth Gillespie and Seitz 1977 Most common reasons for reverse split - image improvement, exchange requirements, appealing to different type of investor, reduction in shareholder service costs West and Brouileter 1970 Traders have lower transactions costs after reverse split

91 Table 5 (Continued)

Summary of studies of forward stock splits Authors Year Type of Firms Period Key Findings Studied Admati and Pfleiderer (1988) 1988 Splits not only attract informed traders, but also noise traders because of lower post-split share prices. Angel 1994 Firms desire to control the relative tick size at which their shares trade. Angel 1997 Firm specific market microstructure factors are stable over time and a firms optimal price can be understood in terms of maintaining an optimal relative tick size, which itself, depends on firm specific characteristics. A larger relative tick size "means greater protection for limit orders, fewer trading errors adn lower costs of negotiation between traders" (Schultz, 1997). What does tick size mean now that there is no tick (decimalization)?

Angel, Brooks, and Matthew 2004 Return Volatility increases after splits - does not support trading range hypothesis. Trading activity by small investors increases. Asquith, Healy, and Palepu 1989 Existence of positive excess returns consistent with signaling hypothesis. Firms that forward split have superior earnings to matched firms in pre split years. Baker and Gallager 1980 CFOs of 100 NYSE 1978 Managers use splits to increase ownership by individual investors firms (conflict with Szewczyk and Tsetsekos) Baker and Phillips 1994 Managers frequently justify splits on the basis that they improve liquidity and marketability Baker and Powell 1993 136 Managers 1987-1990 70% of managers surveyed cited a preferred price range or a stock's liquidity as the reason to forward split. Only 14% pointed to signaling Baker, Phillips, and Powell 1995 Survey Artilce on 1978-1993 Review article on splits. Managers want price in $20 to $35 range forward splits Boehme and Sorescu 2002 NYSE, AMEX, Nasdaq 1927-1998 Equally weighted calendar time portfolios yield positive abnormal with dividend returns but value weighting the same portfolios yields no resumptions significant abnormal return. Changes in risk loadings help explain long term returns Boehme, Danielsen, and 2003 5,550 stock splits and 1949-2000 No positive 12 month CTARS. Abnormal returns from Sorescu 556 large stock announcement to delivery date. Post announcement drift dividends NYSE, attributable to trading frictions rather than behavioral biases. AMEX and Nasdaq Seeks to resolve dispute between Byun and Rozeff versus Inkenberry and Ramnath.

Brennan and Copeland 1988 1,034 forward splits The evidence on the reduction and the extent of information asymmetry is mixed. Decreasing risk shifts for forward splits and systematic risk decreases after forward splits. Splits are costly since the fixed cost element of commissions increases the per share trading costs of low priced stocks. The size of the split factor provides signal. Signals must be costly to be credible (harder for other firms to copy). One cost may be the increased transaction costs of lower priced shares

Brennan and Copeland 1988 1,034 forward splits Beta increases on both announcement date and ex-date. Beta increase becomes permanent after ex-date. Firm specific event increases systematic risk. Brennan and Hughes 1991 Brokerage firms want to preserver commission income. The size of the split factor provides signal. Wider spreads give brokerage firms the incentive to cover (provide information about) firm and bring in new investors Byun and Rozeff 2002 12,747 forward stock 1927-1996 Market efficiency is not challenged by stock split anomaly. There splits are few if any 12-month CTARs or abnormal returns compared to match firms. Find abnormal returns in 2-1 splits. Abnormal returns calculated after ex-date.

92 Table 5 (Continued)

Chen, Roll and Ross 1995 Evidence of momentum in stock returns over a six-month to 1 year time period. Since stock splits are preceded by a run up, the post split abnormal returns may be attributed to momentum rather than the split. Conroy and Harris 1999 Stock splits and stock 1925-1996 A firm's history of stock splits plays a crucial role in both the distributions (>25%) design and effect of current splits. Post split prices below NYSE previous post split price is especially good information. Conroy, Harris, and Benet 1990 133 NYSE forward 1981-1983 No improved liquidity post split based on daily inside bid/ask splits of 1.2 for 1 or quotes from specialists. Liquidity is worse after split. greater Conroy, Harris, and Benet 1990 133 NYSE forward 1981-1983 Forward splits do not increase trading volume, but bid-ask spread splits of 1.2 for 1 or increases greater Corwin and Lipson 2000 469 intraday trading 1995-1996 Forward splits result in enlarged ownership base, increased halts in NYSE common number of small trades, small buy orders by individuals after split. stocks Volatility increases. Desai, Nimalendran, and 1998 Forward splits result in enlarged ownership base, increased Venkataraman number of small trades, small buys orders by individuals after split. Volatility increases. Dennis and Strickland 2002 Quarterly ownership 1990-1993 Abnormal stock split announcement returns are greatest for firms composition of 1,392 with low pre split institutional ownership. Liquidity changes most NYSE, AMEX, and for firms with low pre split institutional ownership. Largest post Nasdaq forward splits split institutional ownership increases for firms with low pre split institutional ownership.

Desai and Jain 1997 5596 forward stock 1976-1991 Forward splits, on average, have one-year BHAR of 7.05% and splits and 76 reverse three-year BHAR of 11.87% after the announcement month. stock split Reverse splits, on average, have one-year BHAR of -10.76% and announcements three-year BHAR of -33.90%. Underreaction to split event is cited. Dolley 1933 88 companies 90% of managers want wider distribution of shares and thus split their stock. Dravid 1987 Return Volatility increases after splits - does not support trading range hypothesis Dubofsky and French 1986 Return Volatility increases after splits - does not support trading range hypothesis Dubofsky 1991 Return Volatility increases after splits - does not support trading range hypothesis Easly, O'hara, and Sarr 2001 75 NYSE forward splits 1995 Microstructure study indicates that forward splits are positive signals. Larger ownership base protects management from takeovers. Trading range helps attract more investors Fernando, Krishnamurthy, 1999 194 Mutual fund splits 1978-1993 Splitting funds experience significant increases (relative to non- and Spindt splitting matched funds) in net assets and shareholders. Evidence of marketability increase. Survey of mutual fund mangers - lower net asset value per share attracts small investors Fernando, Krishnamurthy, 1999 194 Mutual fund splits 1978-1993 Firms going public appear to use offering price to influence and Spindt investor interest in the issue Grinblatt, Masulis, and 1984 Managers forward split stock when they are optimistic Titman Ikenberry and Ramnath 2002 3,028 NYSE, AMEX, 1988-1997 One year BHAR of 9% for forward stock splits attributed to market and Nasdaq forward underreaction. Financial analysts tend to underestimate firms stock splits of five-for- earnings (forecast bias) four or greater. Ikenberry, Rankine, and Stice 1996 1,275 NYSE & AMEX 1975-1990 Post forward split positive abnormal returns of 7.93% for 12 firms with two-for-one months, 12.15% for 36 months — market underreacts. stock splits Announcement excess return of 3.38% Market reacts to other corporate events. Managers split stock when they are optimistic (self select by conditioning on expected future performance).

93 Table 5 (Continued)

Kadiyala and Vetsuypens 2002 NYSE splits with short 1990-1994 Short interest should be used to measure signaling strength. interest data Short interest declines in light of positive information. Short interest does not decline around stock splits, but it does decline for the subsample characterized by favorable industry-adjusted pre-split performance. It increases significantly for firms that have post-split liquidity improvements.

Kryzanowski and Zhang 1996 Forward splits result in enlarged ownership base, increased number of small trades, small buy orders by individuals after split. Lakonishok and Lev 1987 1015 stock splits and 1963-1982 Pre split price (and dividend) run up proceeds forward splits. 1257 stock dividends Managers want to increase individual investor ownership. Managers have a target price - market wide average and industry averages. Percentage price appreciation during the year before split had no explanatory power for split ratios

Maloney and Mulherin 1992 446 Nasdaq 1.25 for 1 1984-1996 Forward splits do not increase volume, but they result in enlarged or greater splits ownership base, more small trades. McNichols and David 1990 all stock dividends and 1976-1983 Existence of positive excess returns consistent with signaling splits hypothesis. Managers split when they are optimistic Merton 1987 Theoretical Wider distribution of stock may lead to lower capital costs in a market with incomplete information Muscarella and Vetsuybens 1996 ADRs Price of ADR and underlying increase on the announcement of ADR split even when underlying does not split. Forward splits result in enlarged ownership base, increased number of small trades, small buy orders by individuals after split. Nayar and Rozeff 2001 forward splits Record date not really important Nayek and Prabhala 2001 forward splits Disentangling Dividend Information in Splits: A decomposition using conditional event study methods. 54% of split announcement effects attributed to dividend information. Splits and dividends are partial substitutes. Ohlson and Penman 1985 1,257 forward splits of 1962-1981 Forward splits are followed by increased return volatility. 2-1 or greater. NYSE Lamoureux and Poon (1987) extend with reverse splits. only. Pilotte and Manuel 1996 776 firms that forward 1970-1988 Investor learning. Investors use a firms previous post split split their stock at least earnings performance to interpret a newly announced split. twice Rozeff 1988 167 Mutual fund splits Behavioral concerns in the context of mutual fund splits. A particular price level may frame how investors classify losses or gains. No evidence o f increased inflows vs. non split. Schultz 2000 2 for 1 splits or greater 1993-1994 Forward splits result in enlarged ownership base, increased on CRSP, NASDAQ, number of small trades, small buy orders by individuals after split. NYSE, AMEX Questions tick size hypothesis based on his examination of intraday trades and quotes. Sheikh 1989 CBOE optionable 1976-1983 Return Volatility increases after splits - does not support trading stocks with split factors range hypothesis greater than 0.25 with no mergers Szewczyk and Tsetsekos 1993 175 NYSE and AMEX 1972-1986 Firms that split have lower managerial ownership than their stock splits with dist matches. Inverse relationship between managerial ownership ratio of 1.25 or greater and post split abnormal returns. Tawatnuntachai and D'Mello 2002 327 NYSE, AMEX, and 1986-1995 Positive signal of splitting firm is contagious to firms in the same Nasdaq split factor industry. Non-splitting industry firms have positive abnormal <=0.5 returns upon announcement of split. Abnormal 1 year CAR of 3.82%+E36

94 Table 6 The Daily Raw Returns of Reverse Stock Splits Around the Announcement Date These panels show the average raw returns around the announcement date of reverse stock splits from 1962 to 2003. The number of positive versus negative returns is given followed by the standardized abnormal return test, Patell's Z (Patell, 1976). This test is used to determine whether the mean is significantly different than zero. The symbols $, *, **, and *** denote statistical significance at the 0.10, 0.05, 0.01 and 0.001 levels, respectively, using a 1-tail test. The symbols (, <, <<, <<<, or ), >, >>, >> show the direction of the generalized sign test (not shown) and denote statistical significance at the 0.10, 0.05, 0.01 and 0.001 levels, respectively, using a 1-tail test. Cumulative returns are given in the bottom panel for the three periods shown.

Panel A Panel B Panel C

All Firms with Firms with Unique Annoncement Day (UAD) Firms with Identical Announcements of Reverse Stock Splits (Announcement Date is not the Ex-Date) Announcement Date and Ex Date (IAE)

Mean Mean Mean Abnormal Positive: Patell Abnormal Positive: Patell Abnormal Positive: Patell Day N Return Negative Z N Return Negative Z N Return Negative Z

-5 1263 0.17% 392:444 0.542 721 0.42% 228:237 0.981 542 -0.17% 164:207 -0.304 -4 1263 0.32% 383:434 1.692 $ 721 0.29% 213:237 0.676 542 0.36% 170:197 1.803 $ -3 1263 0.13% 385:438 0.485 721 0.18% 211:247 0.614 542 0.06% 174:191 0.031 -2 1263 0.60% 388:427 1.550 721 0.38% 209:243 0.618 542 0.88% 179:184 1.653 $ -1 1263 0.72% 412:415> 3.145 ** 721 0.24% 212:247 0.529 542 1.35% 200:168>>> 4.190 *** 0 1263 -3.38% 365:646 -12.572 *** 721 -0.72% 195:286 -2.293 * 542 -6.92% 170:360 -16.545 *** +1 1263 -1.63% 382:585 -7.307 *** 721 -2.57% 195:327 -9.241 *** 542 -0.39% 187:258> -0.497 +2 1263 0.68% 442:485>>> 1.665 $ 721 -0.18% 243:277>> -1.114 542 1.81% 199:208>> 3.826 *** +3 1263 -0.22% 409:515> -0.263 721 -0.17% 229:277 -0.092 542 -0.28% 180:238 -0.295 +4 1263 0.11% 412:476> 0.681 721 -0.47% 208:262 -0.967 542 0.89% 204:214>>> 2.154 * +5 1263 -0.30% 377:521 -2.411 * 721 -0.41% 208:289 -2.513 * 542 -0.15% 169:232 -0.783

Mean Mean Mean Cumulative Positive: Patell Cumulative Positive: Patell Cumulative Positive: Patell Days N Return Negative Z N Return Negative Z N Return Negative Z

(-5,-1) 1263 1.93% 660:527>>> 3.316 *** 721 1.52% 356:314>>> 1.529 542 2.49% 304:213>>> 3.298 *** (0,+1) 1263 -5.01% 425:738>> -14.057 *** 721 -3.29% 239:388> -8.156 *** 542 -7.31% 186:350> -12.051 *** (+2,+5) 1263 0.27% 601:606>>> -0.164 721 -1.22% 333:344>>> -2.343 * 542 2.27% 268:262>>> 2.451 *

95 Table 7 The Daily Market Adjusted Returns of Reverse Stock Splits Around the Announcement Date These panels show the average daily market model adjusted returns around the announcement date of reverse stock splits from 1962 to 2003. Abnormal returns are the daily return minus the daily market index. The market adjusted returns presented on this page were obtained using the CRSP equal-weighted index as a proxy for the market. The results on the following page were obtained using the CRSP value-weighted index as a proxy for the market. The number of positive versus negative returns is given followed by the standardized abnormal return test, Patell's Z (Patell, 1976). This test is used to determine whether the mean is significantly different than zero. The symbols $,*,**, and *** denote statistical significance at the 0.10, 0.05, 0.01 and 0.001 levels, respectively, using a 1-tail test. The symbols (, <, <<, <<<, or ), >, >>, >> show the direction of the generalized sign test (not shown) and denote statistical significance at the 0.10, 0.05, 0.01 and 0.001 levels, respectively, using a 1-tail test. Cumulative abnormal returns are given in the bottom panel for the three periods shown.

Panel A Panel B Panel C

All Firms with Firms with Unique Annoncement Day (UAD) Firms with Identical Announcements of Reverse Stock Splits (Announcement Date is not the Ex-Date) Announcement Date and Ex Date (IAE)

Mean Mean Mean Abnormal Positive: Patell Abnormal Positive: Patell Abnormal Positive: Patell Day N Return Negative Z N Return Negative Z N Return Negative Z

-5 1263 0.09% 558:705 0.193 721 0.38% 333:388) 0.966 542 -0.30% 225:317 -0.819 -4 1263 0.25% 570:693> 1.294 721 0.21% 319:402 0.356 542 0.31% 251:291> 1.565 -3 1263 0.06% 538:725 -0.044 721 0.11% 307:414 0.225 542 -0.01% 231:311 -0.326 -2 1263 0.49% 576:687> 0.933 721 0.31% 331:390) 0.222 542 0.73% 245:297 1.168 -1 1263 0.58% 568:695) 2.342 * 721 0.11% 312:409 -0.044 542 1.19% 256:286> 3.626 *** 0 1263 -3.48% 437:826<<< -13.277 *** 721 -0.79% 270:451<< -2.718 ** 542 -7.06% 167:375<<< -17.131 *** +1 1263 -1.71% 501:762( -7.783 *** 721 -2.64% 274:447< -9.647 *** 542 -0.48% 227:315 -0.755 +2 1263 0.57% 556:707 1.026 721 -0.26% 311:410 -1.652 $ 542 1.68% 245:297 3.471 *** +3 1263 -0.29% 524:739 -0.679 721 -0.24% 303:418 -0.416 542 -0.35% 221:321 -0.556 +4 1263 0.01% 551:712 0.080 721 -0.56% 304:417 -1.443 542 0.77% 247:295) 1.786 $ +5 1263 -0.36% 524:739 -2.777 ** 721 -0.48% 297:424 -2.920 ** 542 -0.20% 227:315 -0.871

Mean Mean Mean Cumulative Cumulative Cumulative Abnormal Positive: Patell Abnormal Positive: Patell Abnormal Positive: Patell Days N Return Negative Z N Return Negative Z N Return Negative Z

(-5,-1) 1263 1.47% 629:634>>> 2.110 * 721 1.13% 344:377>> 0.771 542 1.93% 285:257>>> 2.332 * (0,+1) 1263 -5.19% 439:824<<< -14.892 *** 721 -3.43% 255:466<<< -8.743 *** 542 -7.54% 184:358<<< -12.648 *** (+2,+5) 1263 -0.06% 563:700 -1.175 721 -1.54% 314:407 -3.216 ** 542 1.90% 249:293) 1.915 $

The results on this page are based on using the CRSP equally weighted index as a market proxy.

96 Table 7 (Continued)

Panel A Panel B Panel C

All Firms with Firms with Unique Annoncement Day (UAD) Firms with Identical Announcements of Reverse Stock Splits (Announcement Date is not the Ex-Date) Announcement Date and Ex Date (IAE)

Mean Mean Mean Abnormal Positive: Patell Abnormal Positive: Patell Abnormal Positive: Patell Day N Return Negative Z N Return Negative Z N Return Negative Z

-5 1263 0.11% 595:668 0.307 721 0.42% 353:368) 1.110 542 -0.30% 242:300 -0.811 -4 1263 0.34% 624:639> 1.727 $ 721 0.27% 351:370 0.594 542 0.43% 273:269> 1.951 $ -3 1263 0.12% 583:680 0.257 721 0.17% 330:391 0.446 542 0.07% 253:289 -0.122 -2 1263 0.58% 609:654) 1.397 721 0.40% 343:378 0.545 542 0.82% 266:276 1.504 -1 1263 0.62% 608:655 2.580 ** 721 0.14% 334:387 0.033 542 1.27% 274:268> 3.900 *** 0 1263 -3.44% 473:790<<< -12.925 *** 721 -0.73% 296:425<< -2.485 * 542 -7.04% 177:365<<< -16.862 *** +1 1263 -1.68% 518:745<<< -7.504 *** 721 -2.62% 283:438<<< -9.481 *** 542 -0.43% 235:307 -0.520 +2 1263 0.64% 585:678 1.386 721 -0.24% 325:396 -1.553 542 1.81% 260:282 3.906 *** +3 1263 -0.21% 554:709 -0.297 721 -0.20% 315:406 -0.248 542 -0.23% 239:303 -0.167 +4 1263 0.04% 564:699 0.343 721 -0.53% 312:409 -1.268 542 0.80% 252:290 1.985 * +5 1263 -0.29% 556:707 -2.377 * 721 -0.42% 311:410 -2.613 ** 542 -0.13% 245:297 -0.615

Mean Mean Mean Cumulative Cumulative Cumulative Abnormal Positive: Patell Abnormal Positive: Patell Abnormal Positive: Patell Days N Return Negative Z N Return Negative Z N Return Negative Z

(-5,-1) 1263 1.77% 641:622>>> 2.803 ** 721 1.39% 349:372 1.220 542 2.29% 292:250>>> 2.872 ** (0,+1) 1263 -5.11% 446:817<<< -14.445 *** 721 -3.35% 261:460<<< -8.461 *** 542 -7.47% 185:357<<< -12.291 *** (+2,+5) 1263 0.17% 579:684 -0.473 721 -1.39% 315:406 -2.841 ** 542 2.26% 264:278 2.555 *

The results on this page are based on using the CRSP value weighted index as a market proxy.

97 Table 8 The Daily Market Model Adjusted Returns of Reverse Stock Splits Around the Announcement Date

These panels show the average market model adjusted returns around the announcement date of reverse stock splits from 1962 to 2003. The market model adjusted returns presented on this page are obtained using the CRSP equal-weighted index as a proxy for the market. The results on the following page are obtained using the CRSP value weighted index as a proxy for the market. The estimation period for the market model parameters is t-160 to t-40 where t is the announcement date. The number of positive versus negative returns is given followed by the standardized abnormal return test, Patell's Z (Patell, 1976). This test is used to determine whether the mean is significantly different than zero. The symbols $,*,**, and *** denote statistical significance at the 0.10, 0.05, 0.01 and 0.001 levels, respectively, using a 1-tail test. The symbols (, <, <<, <<<, or ), >, >>, >> show the direction of the generalized sign test (not shown) and denote statistical significance at the 0.10, 0.05, 0.01 and 0.001 levels, respectively, using a 1-tail test. Cumulative abnormal returns are given in the bottom panel for the three periods shown.

Panel A Panel B Panel C

All Firms with Firms with Unique Annoncement Day (UAD) Firms with Identical Announcements of Reverse Stock Splits (Announcement Date is not the Ex-Date) Announcement Date and Ex Date (IAE)

Mean Mean Mean Abnormal Positive: Patell Abnormal Positive: Patell Abnormal Positive: Patell Day N Return Negative Z N Return Negative Z N Return Negative Z

-5 1263 0.06% 575:688 0.468 721 0.38% 342:379) 1.116 542 -0.36% 233:309 -0.573 -4 1263 0.24% 579:684 1.644 721 0.25% 331:390 0.719 542 0.23% 248:294 1.680 $ -3 1263 0.07% 574:689 0.597 721 0.18% 336:385 0.979 542 -0.07% 238:304 -0.218 -2 1263 0.42% 576:687 1.033 721 0.29% 330:391 0.430 542 0.58% 246:296 1.080 -1 1263 0.50% 586:677) 2.458 * 721 0.06% 313:408 0.055 542 1.09% 273:269>> 3.689 *** 0 1263 -3.53% 454:809<<< -12.915 *** 721 -0.82% 286:435< -2.396 * 542 -7.13% 168:374<<< -16.950 *** +1 1263 -1.77% 498:765<<< -7.748 *** 721 -2.69% 275:446<<< -9.612 *** 542 -0.55% 223:319 -0.742 +2 1263 0.46% 570:693 1.061 721 -0.33% 322:399 -1.538 542 1.51% 248:294 3.394 *** +3 1263 -0.32% 529:734 -0.251 721 -0.23% 306:415 -0.018 542 -0.43% 223:319 -0.363 +4 1263 -0.10% 546:717 0.083 721 -0.64% 301:420 -1.342 542 0.61% 245:297 1.674 $ +5 1263 -0.44% 555:708 -2.573 * 721 -0.52% 321:400 -2.575 * 542 -0.34% 234:308 -0.957

Mean Mean Mean Cumulative Cumulative Cumulative Abnormal Positive: Patell Abnormal Positive: Patell Abnormal Positive: Patell Days N Return Negative Z N Return Negative Z N Return Negative Z

(-5,-1) 1263 1.29% 630:633>>> 2.773 ** 721 1.17% 346:375> 1.475 542 1.46% 284:258>>> 2.531 * (0,+1) 1263 -5.30% 438:825<<< -14.611 *** 721 -3.51% 259:462<<< -8.491 *** 542 -7.68% 179:363<<< -12.510 *** (+2,+5) 1263 -0.40% 546:717 -0.840 721 -1.71% 310:411 -2.737 ** 542 1.35% 236:306 1.874 $

The results on this page are based on using the CRSP equally weighted index as a market proxy.

98 Table 8 (Continued)

Panel A Panel B Panel C

All Firms with Firms with Unique Annoncement Day (UAD) Firms with Identical Announcements of Reverse Stock Splits (Announcement Date is not the Ex-Date) Announcement Date and Ex Date (IAE)

Mean Mean Mean Abnormal Positive: Patell Abnormal Positive: Patell Abnormal Positive: Patell Day N Return Negative Z N Return Negative Z N Return Negative Z

-5 1263 -0.02% 583:680 0.070 721 0.29% 349:372> 0.765 542 -0.43% 234:308 -0.776 -4 1263 0.19% 569:694 1.502 721 0.19% 325:396 0.588 542 0.19% 244:298 1.613 -3 1263 0.05% 567:696 0.445 721 0.17% 329:392 0.831 542 -0.10% 238:304 -0.279 -2 1263 0.44% 572:691 1.141 721 0.26% 327:394 0.377 542 0.67% 245:297 1.307 -1 1263 0.52% 577:686 2.543 * 721 0.05% 308:413 -0.012 542 1.15% 269:273>> 3.895 *** 0 1263 -3.57% 458:805<<< -13.030 *** 721 -0.87% 288:433< -2.586 ** 542 -7.17% 170:372<<< -16.908 *** +1 1263 -1.81% 491:772<<< -7.838 *** 721 -2.70% 280:441<< -9.621 *** 542 -0.64% 211:331< -0.868 +2 1263 0.49% 567:696 1.120 721 -0.36% 322:399 -1.646 $ 542 1.61% 245:297 3.609 *** +3 1263 -0.34% 531:732 -0.431 721 -0.29% 303:418 -0.217 542 -0.39% 228:314 -0.407 +4 1263 -0.10% 555:708 0.046 721 -0.66% 302:419 -1.424 542 0.65% 253:289 1.712 $ +5 1263 -0.44% 551:712 -2.613 ** 721 -0.52% 314:407 -2.535 * 542 -0.32% 237:305 -1.066

Mean Mean Mean Cumulative Cumulative Cumulative Abnormal Positive: Patell Abnormal Positive: Patell Abnormal Positive: Patell Days N Return Negative Z N Return Negative Z N Return Negative Z

(-5,-1) 1263 1.18% 621:642>>> 2.549 * 721 0.95% 335:386 1.140 542 1.48% 286:256>>> 2.577 ** (0,+1) 1263 -5.39% 431:832<<< -14.756 *** 721 -3.57% 252:469<<< -8.632 *** 542 -7.80% 179:363<<< -12.569 *** (+2,+5) 1263 -0.38% 538:725 -0.939 721 -1.83% 296:425( -2.911 ** 542 1.55% 242:300 1.924 $

The results on this page are based on using the CRSP value weighted index as a market proxy.

99 Table 9 The Daily Raw Returns of Reverse Stock Splits Around the Ex-date These panels show the average raw returns around the ex-date of reverse stock splits from 1962 to 2003. The number of positive versus the number of negative returns is shown and the value of the standardized abnormal return test, Patell's Z (Patell, 1976) is given. These tests are used to determine whether the mean is significantly different than zero. The symbols $,*,**, and *** denote statistical significance at the 0.10, 0.05, 0.01 and 0.001 levels, respectively, using a 1-tail test. The symbols (,< or ),> etc. correspond to $,* and show the significance and direction of the generalized sign test (not shown).

Panel D Panel E

All Ex Date with Ex Dates No Announcement Date Found

Mean Positive: Patell Mean Positive: Patell Day N Return Negative Z N Return Negative Z

-5 1811 0.43% 499:617 1.535 403 1.35% 91:116 0.722 -4 1811 0.20% 500:593 0.443 403 0.11% 92:120 -0.503 -3 1811 -0.17% 500:602 -0.722 403 -0.15% 97:113 0.723 -2 1811 0.77% 510:572 4.737 *** 403 1.98% 115:100>> 6.100 *** -1 1811 -0.15% 510:623 -0.427 403 -0.72% 91:134 -0.731 0 1811 -6.08% 545:1209>> -32.171 *** 403 -5.86% 121:273>>> -19.791 *** +1 1811 0.10% 620:773>>> 0.497 403 0.23% 140:149>>> 0.6 +2 1811 0.24% 594:721>>> 0.566 403 -0.41% 118:152>> 0.376 +3 1811 -0.62% 532:773> -3.403 *** 403 -0.60% 101:149 -2.165 * +4 1811 0.23% 593:721>>> 0.078 403 -0.15% 129:149>>> -0.926 +5 1811 -0.33% 536:730> -2.647 ** 403 -0.68% 105:147 -2.008 *

Mean Mean Cumulative Positive: Patell Cumulative Positive: Patell Days N Return Negative Z N Return Negative Z

(-5,-1) 1811 1.09% 857:782>>> 2.489 * 403 2.58% 169:166>>> 2.823 ** (0,0) 1811 -6.08% 545:1209>> -32.171 *** 403 -5.86% 121:273>>> -19.791 *** (+1,+5) 1811 -0.38% 779:964>>> -2.196 * 403 -1.61% 161:214>>> -1.843 $

100

Table 9 (Continued)

Panel F Panel G

Ex Date is Not Announcement Date Ex Date is Not Announcement Date Missing Announcements Included Missing Announcements Not Included

Mean Positive: Patell Mean Positive: Patell Day N Return Negative Z N Return Negative Z

-5 1287 0.64% 338:418 1.861 $ 884 0.32% 247:302 1.758 $ -4 1287 0.11% 335:403 -0.639 884 0.11% 243:283 -0.432 -3 1287 -0.25% 334:417 -0.835 884 -0.30% 237:304 -1.498 -2 1287 0.75% 340:393 4.872 *** 884 0.19% 225:293 1.754 $ -1 1287 -0.75% 320:458 -2.942 ** 884 -0.76% 229:324 -3.058 ** 0 1287 -5.78% 375:866> -28.210 *** 884 -5.74% 254:593 -20.669 *** +1 1287 0.31% 436:524>>> 0.984 884 0.34% 296:375>>> 0.782 +2 1287 -0.45% 398:525>>> -2.093 * 884 -0.47% 280:373>> -2.782 ** +3 1287 -0.76% 358:543 -3.934 *** 884 -0.83% 257:394 -3.286 ** +4 1287 -0.02% 398:516>>> -1.154 884 0.03% 269:367) -0.767 +5 1287 -0.38% 374:505> -2.686 ** 884 -0.25% 269:358) -1.885 $

Mean Mean Cumulative Positive: Patell Cumulative Positive: Patell Days N Return Negative Z N Return Negative Z

(-5,-1) 1287 0.51% 567:573>>> 1.036 884 -0.44% 398:407>>> -0.66 (0,0) 1287 -5.78% 375:866> -28.210 *** 884 -5.74% 254:593 -20.669 *** (+1,+5) 1287 -1.31% 524:702>>> -3.973 *** 884 -1.17% 363:488>>> -3.550 ***

101 Table 10 The Daily Market Adjusted Abnormal Returns of Reverse Stock Splits Around the Ex-date These panels show the average daily market adjusted returns around the ex date of reverse stock splits from 1962 to 2003. Abnormal returns are the daily return minus the daily market index. The market adjusted returns presented on this page use the CRSP equal- weighted index as a proxy for the market. The results on the following page use the CRSP value-weighted index as a proxy for the market. The positive versus negative returns is given followed by the standardized abnormal return test, Patell's Z (Patell, 1976). These tests are used to determine whether the mean is significantly different than zero. The symbols $,*,**, and *** denote statistical significance at the 0.10, 0.05, 0.01 and 0.001 levels, respectively, using a 1-tail test. The symbols (,< or ),> etc. correspond to $,* and show the significance and direction of the generalized sign test (not shown).

Panel D Panel E

All Ex Date with Ex Dates No Announcement Date Found

Mean Positive: Patell Mean Positive: Patell Day N Return Negative Z N Return Negative Z

-5 1811 0.37% 794:1017 0.995 403 1.34% 181:222 0.756 -4 1811 0.15% 830:981>> 0.169 403 0.07% 177:226 -0.623 -3 1811 -0.26% 772:1039 -1.499 403 -0.25% 171:232 0.306 -2 1811 0.61% 773:1038 3.581 *** 403 1.82% 176:227 5.569 *** -1 1811 -0.27% 740:1071 -1.356 403 -0.81% 151:252( -1.067 0 1811 -6.18% 556:1255<<< -32.726 *** 403 -5.96% 124:279<<< -19.640 *** +1 1811 0.02% 766:1045 -0.029 403 0.14% 180:223 0.295 +2 1811 0.13% 769:1042 -0.286 403 -0.58% 165:238 -0.516 +3 1811 -0.69% 707:1104<< -4.010 *** 403 -0.66% 155:248 -2.358 * +4 1811 0.13% 780:1031 -0.551 403 -0.27% 170:233 -1.22 +5 1811 -0.39% 776:1035 -2.886 ** 403 -0.76% 173:230 -2.064 *

Mean Mean Cumulative Cumulative Abnormal Positive: Patell Abnormal Positive: Patell Days N Return Negative Z N Return Negative Z

(-5,-1) 1811 0.61% 849:962>>> 0.845 403 2.17% 179:224 2.210 * (0,0) 1811 -6.18% 556:1255<<< -32.726 *** 403 -5.96% 124:279<<< -19.640 *** (+1,+5) 1811 -0.80% 749:1062 -3.471 *** 403 -2.13% 158:245 -2.622 **

The results on this page are based on using the CRSP equally weighted index as a market proxy.

102 Table 10 (Continued)

Panel F Panel G

Ex Date is Not Announcement Date Ex Date is Not Announcement Date Missing Announcements Included Missing Announcements Not Included

Mean Positive: Patell Mean Positive: Patell Day N Return Negative Z N Return Negative Z

-5 1287 0.60% 574:713 1.526 884 0.26% 393:491 1.33 -4 1287 0.07% 587:700> -0.85 884 0.07% 410:474> -0.605 -3 1287 -0.34% 552:735 -1.513 884 -0.38% 381:503 -2.034 * -2 1287 0.59% 537:750 3.812 *** 884 0.02% 361:523 0.834 -1 1287 -0.86% 493:794<< -3.682 *** 884 -0.88% 342:542< -3.724 *** 0 1287 -5.85% 390:897<<< -28.482 *** 884 -5.81% 266:618<<< -21.100 *** +1 1287 0.23% 545:742 0.513 884 0.27% 365:519 0.42 +2 1287 -0.55% 526:761 -2.918 ** 884 -0.54% 361:523 -3.175 ** +3 1287 -0.82% 495:792<< -4.455 *** 884 -0.90% 340:544< -3.783 *** +4 1287 -0.11% 543:744 -1.665 $ 884 -0.04% 373:511 -1.185 +5 1287 -0.44% 557:730 -2.897 ** 884 -0.30% 384:500 -2.100 *

Mean Mean Cumulative Positive: Patell Cumulative Positive: Patell Days N Return Negative Z N Return Negative Z

(-5,-1) 1287 0.05% 575:712) -0.316 884 -0.91% 396:488 -1.877 $ (0,0) 1287 -5.85% 390:897<<< -28.482 *** 884 -5.81% 266:618<<< -21.100 *** (+1,+5) 1287 -1.70% 507:780< -5.108 *** 884 -1.50% 349:535( -4.393 ***

The results on this page are based on using the CRSP equally weighted index as a market proxy.

103 Table 10 (Continued)

Panel D Panel E

All Ex Date with Ex Dates No Announcement Date Found

Mean Positive: Patell Mean Positive: Patell Day N Return Negative Z N Return Negative Z

-5 1811 0.39% 830:981 1.182 403 1.38% 183:220 0.936 -4 1811 0.21% 856:955 0.515 403 0.07% 180:223 -0.675 -3 1811 -0.23% 819:992 -1.442 403 -0.25% 187:216 0.177 -2 1811 0.66% 832:979 3.900 *** 403 1.87% 195:208 5.781 *** -1 1811 -0.20% 807:1004 -0.826 403 -0.73% 171:232 -0.759 0 1811 -6.14% 577:1234<<< -31.885 *** 403 -5.94% 128:275<<< -18.821 *** +1 1811 0.06% 798:1013 0.187 403 0.17% 188:215 0.267 +2 1811 0.20% 810:1001 0.013 403 -0.52% 174:229 -0.524 +3 1811 -0.62% 766:1045<< -3.627 *** 403 -0.61% 169:234 -2.060 * +4 1811 0.18% 815:996 -0.145 403 -0.22% 188:215 -0.927 +5 1811 -0.34% 798:1013 -2.495 * 403 -0.68% 177:226 -1.692 $

Mean Mean Cumulative Cumulative Abnormal Positive: Patell Abnormal Positive: Patell Days N Return Negative Z N Return Negative Z

(-5,-1) 1811 0.82% 865:946) 1.488 403 2.33% 184:219 2.442 * (0,0) 1811 -6.14% 577:1234<<< -31.885 *** 403 -5.94% 128:275<<< -18.821 *** (+1,+5) 1811 -0.52% 776:1035< -2.713 ** 403 -1.86% 162:241< -2.207 *

The results on this page are based on using the CRSP value weighted index as a market proxy.

104 Table 10 (Continued)

Panel F Panel G

Ex Date is Not Announcement Date Ex Date is Not Announcement Date Missing Announcements Included Missing Announcements Not Included

Mean Positive: Patell Mean Positive: Patell Day N Return Negative Z N Return Negative Z

-5 1287 0.63% 593:694 1.736 $ 884 0.28% 410:474 1.462 -4 1287 0.10% 594:693 -0.668 884 0.11% 414:470 -0.351 -3 1287 -0.33% 578:709 -1.566 884 -0.37% 391:493 -2.011 * -2 1287 0.61% 575:712 3.993 *** 884 0.03% 380:504 0.909 -1 1287 -0.79% 545:742< -3.192 ** 884 -0.82% 374:510< -3.341 *** 0 1287 -5.81% 401:886<<< -27.650 *** 884 -5.76% 273:611<<< -20.650 *** +1 1287 0.27% 568:719 0.648 884 0.32% 380:504 0.601 +2 1287 -0.50% 553:734( -2.785 ** 884 -0.49% 379:505 -3.008 ** +3 1287 -0.78% 537:750<< -4.212 *** 884 -0.85% 368:516< -3.692 *** +4 1287 -0.06% 573:714 -1.299 884 0.01% 385:499 -0.941 +5 1287 -0.41% 562:725 -2.611 ** 884 -0.28% 385:499 -2.008 *

Mean Mean Cumulative Cumulative Abnormal Positive: Patell Abnormal Positive: Patell Days N Return Negative Z N Return Negative Z

(-5,-1) 1287 0.21% 585:702 0.135 884 -0.76% 401:483 -1.49 (0,0) 1287 -5.81% 401:886<<< -27.650 *** 884 -5.76% 273:611<<< -20.650 *** (+1,+5) 1287 -1.47% 517:770<<< -4.589 *** 884 -1.29% 355:529<< -4.047 ***

The results on this page are based on using the CRSP value weighted index as a market proxy.

105 Table 11 The Daily Market Model Adjusted Returns of Reverse Stock Splits Around the Ex-date These panels show the average market model adjusted returns around the ex date of reverse stock splits from 1962 to 2003. The market model adjusted returns presented on this page use the CRSP equal-weighted index as a proxy for the market. The results on the following page use the CRSP value-weighted index as a proxy for the market. The estimation period for the market model parameters is t-160 to t-40 where t is the ex date. The positive versus negative returns is given followed by the standardized abnormal return test, Patell's Z (Patell, 1976). These tests are used to determine whether the mean is significantly different than zero. The symbols $,*,**, and *** denote statistical significance at the 0.10, 0.05, 0.01 and 0.001 levels, respectively, using a 1-tail test. The symbols (,< or ),> etc. correspond to $,* and show the significance and direction of the generalized sign test (not shown).

Panel D Panel E

All Ex Date with Ex Dates No Announcement Date Found

Mean Positive: Patell Mean Positive: Patell Day N Return Negative Z N Return Negative Z

-5 1811 0.28% 805:1006 0.989 403 1.17% 182:221 0.555 -4 1811 0.09% 852:959> 0.385 403 0.05% 192:211 -0.338 -3 1811 -0.29% 827:984 -1.093 403 -0.33% 193:210 0.59 -2 1811 0.52% 816:995 3.711 *** 403 1.68% 195:208 5.627 *** -1 1811 -0.31% 818:993 -0.921 403 -0.77% 177:226 -0.722 0 1811 -6.21% 553:1258<<< -32.724 *** 403 -5.97% 121:282<<< -20.267 *** +1 1811 0.01% 783:1028 0.57 403 0.18% 186:217 0.778 +2 1811 0.09% 797:1014 0.326 403 -0.59% 176:227 0.053 +3 1811 -0.68% 749:1062<< -3.430 *** 403 -0.60% 181:222 -2.084 * +4 1811 0.06% 794:1017 -0.391 403 -0.35% 177:226 -1.159 +5 1811 -0.44% 795:1016 -2.856 ** 403 -0.74% 182:221 -2.106 *

Mean Mean Cumulative Cumulative Abnormal Positive: Patell Abnormal Positive: Patell Days N Return Negative Z N Return Negative Z

(-5,-1) 1811 0.29% 857:954> 1.374 403 1.81% 183:220 2.554 * (0,0) 1811 -6.21% 553:1258<<< -32.724 *** 403 -5.97% 121:282<<< -20.267 *** (+1,+5) 1811 -0.95% 755:1056< -2.585 ** 403 -2.10% 168:235( -2.020 *

The results on this page are based on using the CRSP equally weighted index as a market proxy.

106 Table 11 (Continued)

Panel F Panel G

Ex Date is Not Announcement Date Ex Date is Not Announcement Date Missing Announcements Included Missing Announcements Not Included

Mean Positive: Patell Mean Positive: Patell Day N Return Negative Z N Return Negative Z

-5 1287 0.50% 579:708 1.392 884 0.20% 397:487 1.305 -4 1287 0.01% 611:676) -0.655 884 0.00% 419:465) -0.562 -3 1287 -0.35% 599:688 -1.074 884 -0.36% 406:478 -1.696 $ -2 1287 0.52% 578:709 4.046 *** 884 -0.02% 383:501 1.077 -1 1287 -0.85% 557:730 -3.168 ** 884 -0.89% 380:504 -3.337 *** 0 1287 -5.86% 385:902<<< -28.576 *** 884 -5.81% 264:620<<< -20.789 *** +1 1287 0.25% 564:723 1.155 884 0.28% 378:506 0.868 +2 1287 -0.53% 550:737( -2.167 * 884 -0.51% 374:510 -2.652 ** +3 1287 -0.79% 532:755<< -3.905 *** 884 -0.88% 351:533<< -3.305 *** +4 1287 -0.14% 560:727 -1.368 884 -0.04% 383:501 -0.868 +5 1287 -0.45% 571:716 -2.818 ** 884 -0.31% 389:495 -1.977 *

Mean Mean Cumulative Cumulative Abnormal Positive: Patell Abnormal Positive: Patell Days N Return Negative Z N Return Negative Z

(-5,-1) 1287 -0.17% 585:702 0.242 884 -1.07% 402:482 -1.437 (0,0) 1287 -5.86% 385:902<<< -28.576 *** 884 -5.81% 264:620<<< -20.789 *** (+1,+5) 1287 -1.66% 520:767<<< -4.071 *** 884 -1.46% 352:532<< -3.548 ***

The results on this page are based on using the CRSP equally weighted index as a market proxy.

107 Table 11 (Continued)

Panel D Panel E

All Ex Date with Ex Dates No Announcement Date Found

Mean Positive: Patell Mean Positive: Patell Day N Return Negative Z N Return Negative Z

-5 1811 0.23% 787:1024 0.777 403 1.16% 181:222 0.533 -4 1811 0.06% 839:972 0.103 403 0.02% 201:202 -0.48 -3 1811 -0.31% 799:1012 -1.237 403 -0.34% 184:219 0.481 -2 1811 0.59% 815:996 4.031 *** 403 1.83% 206:197) 5.951 *** -1 1811 -0.28% 811:1000 -0.72 403 -0.74% 185:218 -0.598 0 1811 -6.23% 562:1249<<< -32.608 *** 403 -5.99% 121:282<<< -20.157 *** +1 1811 -0.03% 783:1028 0.345 403 0.17% 196:207 0.689 +2 1811 0.09% 791:1020 0.276 403 -0.59% 176:227 0.105 +3 1811 -0.71% 748:1063<< -3.582 *** 403 -0.63% 178:225 -2.057 * +4 1811 0.03% 795:1016 -0.572 403 -0.39% 185:218 -1.194 +5 1811 -0.46% 789:1022 -3.008 ** 403 -0.75% 173:230 -2.037 *

Mean Mean Cumulative Cumulative Abnormal Positive: Patell Abnormal Positive: Patell Days N Return Negative Z N Return Negative Z

(-5,-1) 1811 0.30% 855:956) 1.321 403 1.93% 183:220 2.633 (0,0) 1811 -6.23% 562:1249<<< -32.608 *** 403 -5.99% 121:282<<< -20.157 (+1,+5) 1811 -1.08% 751:1060<< -2.926 ** 403 -2.19% 173:230 -2.010

The results on this page are based on using the CRSP value weighted index as a market proxy.

108 Table 11 (Continued)

Panel F Panel G

Ex Date is Not Announcement Date Ex Date is Not Announcement Date Missing Announcements Included Missing Announcements Not Included

Mean Positive: Patell Mean Positive: Patell Day N Return Negative Z N Return Negative Z

-5 1287 0.47% 561:726 1.251 884 0.15% 380:504 1.15 -4 1287 -0.02% 604:683 -0.929 884 -0.03% 403:481 -0.797 -3 1287 -0.36% 571:716 -1.195 884 -0.37% 387:497 -1.768 $ -2 1287 0.59% 580:707 4.300 *** 884 0.02% 374:510 1.165 -1 1287 -0.83% 554:733( -3.049 ** 884 -0.87% 369:515( -3.277 ** 0 1287 -5.88% 392:895<<< -28.477 *** 884 -5.83% 271:613<<< -20.743 *** +1 1287 0.23% 576:711 1.028 884 0.26% 380:504 0.774 +2 1287 -0.57% 548:739< -2.310 * 884 -0.57% 372:512 -2.860 ** +3 1287 -0.85% 528:759<< -4.054 *** 884 -0.94% 350:534<< -3.503 *** +4 1287 -0.20% 553:734( -1.618 884 -0.12% 368:516( -1.145 +5 1287 -0.49% 561:726 -2.950 ** 884 -0.37% 388:496 -2.183 *

Mean Mean Cumulative Cumulative Abnormal Positive: Patell Abnormal Positive: Patell Days N Return Negative Z N Return Negative Z

(-5,-1) 1287 -0.16% 583:704 0.169 884 -1.11% 400:484 -1.577 (0,0) 1287 -5.88% 392:895<<< -28.477 *** 884 -5.83% 271:613<<< -20.743 *** (+1,+5) 1287 -1.88% 517:770<<< -4.429 *** 884 -1.74% 344:540<<< -3.988 ***

The results on this page are based on using the CRSP value weighted index as a market proxy.

109 Table 12 Ex-date Returns vs. Announcement Returns This table shows the results of regressing ex-date returns on the announcement day returns. The announcement day is defined as the day of the announcement and the day after the announcement. Raw returns and abnormal returns are used in the regressions as indicated in the first column. The first column defines the ex-date and AD returns used in the regression. The numbers in bold Italics are t-values. Higher (absolute) t-values indicate that the coefficients are not equal to zero. CR is cumulative returns (raw) and CAR is cumulative abnormal return.

Ex Date Returns Intercept Coefficient AD Returns

Raw Ex = -0.0600 0.2114 x AD CR Returns -12.86 6.85

Market Adjusted Returns Ex = -0.0613 0.1367 x AD CAR Equally Weighted Index -12.93 4.51

Market Adjusted Returns Ex = -0.0611 0.1314 x AD CAR Value Weighted Index -12.90 4.35

Market Model Adjusted Returns Ex = -0.0621 0.1191 x AD CAR Equally Weighted Index -13.12 3.95

Market Model Adjusted Returns Ex = -0.0622 0.1198 x AD CAR Value Weighted Index -13.10 3.97

110 Table 13 Number of Reverse Stock Splits by Industry 1962-2003 Reverse splits are categorized according to the Fama and French industry classifications that are defined in Appendix C. These industry categories were obtained from Kenneth French's web site.

Five Industries Thirty Industries Seventeen Industries 1 Manuf 643 1 Food 34 16 Carry 6 1 Food 46 10 FabPr 8 2 Utils 28 2 Beer 5 17 Mines 100 2 Mines 102 11 Machn 194 3 Shops 213 3 Smoke 3 18 Coal 1 3 Oil 194 12 Cars 23 4 Money 216 4 Games 76 19 Oil 191 4 Clths 25 13 Trans 34 5 Other 918 5 Books 20 20 Util 11 5 Durbl 35 14 Utils 11 6 Hshld 24 21 Telcm 77 6 Chems 22 15 Rtail 104 7 Clths 16 22 Servs 328 7 Cnsum 87 16 Finan 207 8 Hlth 216 23 BusEq 139 8 Cnstr 60 17 Other 844 Ten Industries 9 Chems 22 24 Paper 12 9 Steel 13 1 NoDur 87 10 Txtls 2 25 Trans 25 2 Durbl 58 11 Cnstr 58 26 Whlsl 100 3 Oil 194 12 Steel 13 27 Rtail 67 Forty-eight Industries 4 Chems 104 13 FabPr 46 28 Meals 57 1 Agric 8 25 Ships 1 5 Manuf 402 14 ElcEq 85 29 Fin 211 2 Food 25 26 Guns 0 6 Telcm 77 15 Autos 18 30 Other 42 3 Soda 1 27 Gold 73 7Utils 12 4 Beer 5 28 Mines 27 8 Shops 555 Thirty-eight Industries 5Smoke 3 29Coal 1 8 Shops 555 1 Agric 12 20 MtlPr 27 6 Toys 11 30 Oil 190 9 Money 215 2 Mines 97 21 Machn 94 7 Fun 64 31 Util 11 10 Other 308 3 Oil 191 22 Elctr 151 8 Books 13 32 Telcm 77 4 Stone 5 23 Cars 21 9 Hshld 24 33 PerSv 30 5 Cnstr 23 24 Instr 97 10 Clths 16 34 BusSv 305 6 Food 31 25 Manuf 13 11 Hlth 72 35 Comps 57 Twelve Industries 7 Smoke 3 26 Trans 25 12 MedEq 69 36 Chips 61 1 NoDur 87 8 Txtls 2 27 Phone 51 13 Drugs 75 37 LabEq 21 2 Durbl 41 9 Apprl 16 28 TV 26 14 Chems 22 38 Paper 8 3 Manuf 153 10 Wood 4 29 Utils 12 15 Rubbr 12 39 Boxes 4 4 Enrgy 194 11 Chair 2 30 Garbg 15 16 Txtls 2 40 Trans 25 5 Chems 29 12 Paper 11 31 Steam 1 17 BldMt 35 41 Whlsl 99 6 BusEq 369 13 Print 20 32 Water 0 18 Cnstr 23 42 Rtail 67 7 Telcm 77 14 Chems 103 33 Whlsl 99 19 Steel 13 43 Meals 57 8 Utils 12 15 Ptrlm 2 34 Rtail 109 20 FabPr 4 44 Banks 47 9 Shops 218 16 Rubbr 14 35 Money 213 21 Mach 41 45 Insur 24 10 Hlth 216 17 Lethr 3 36 Srvc 485 22 ElcEq 84 46 RlEst 32 11 Money 215 18 Glass 6 37 Govt 1 23 Autos 17 47 Fin 103 12 Other 401 19 Metal 13 38 Other 1 24 Aero 5 48 Other 31

111 Table 14 Ex-date Returns by Industry The columns represent the number of firms, the average ex-date return, and the standard deviation. As a point of reference, the average ex-date return for the full sample is -0.06. The t-values shown represent the value of a two-sample test of equal means. The values tested are the significant high and low values that are highlighted. Reverse splits are categorized according to the Fama and French industry classifications that are defined in Appendix C. These industry categories were obtained from Kenneth French's web site.

Five Industries Seventeen Industries Thirty Industries No Name N Mean Std t-value No Name N Mean Std t-value No Name N Mean Std t-value 2 Utils 27 -4.77% 0.1802 -1.38 10 FabPr 7 2.23% 0.0886 0.66 3 Smoke 3 1.56% 0.0327 0.83 4 Money 212 -5.15% 0.1328 -5.65 * 9 Steel 13 -1.97% 0.1410 -0.50 10 Txtls 2 1.12% 0.1021 0.15 5 Other 898 -6.36% 0.1676 -11.37 6 Chems 22 -3.00% 0.1175 -1.20 1 Food 34 -1.15% 0.1030 -0.65 1 Manuf 624 -7.58% 0.1536 -12.33 14 Utils 10 -3.24% 0.1127 -0.91 12 Steel 13 -1.97% 0.1410 -0.50 3 Shops 208 -8.06% 0.1951 -5.96 * 1 Food 46 -3.25% 0.1332 -1.66 9 Chems 22 -3.18% 0.1173 -1.27 3 Oil 189 -4.38% 0.1419 -4.25 * 20 Util 10 -3.24% 0.1127 -0.91 2 Sample Test for Equal Means T = 1.78 16 Finan 206 -4.91% 0.1314 -5.37 5 Books 20 -3.51% 0.1634 -0.96 5 Durbl 34 -5.19% 0.1096 -2.76 30 Other 43 -3.91% 0.1489 -1.72 Ten Industries 12 Cars 24 -6.24% 0.1371 -2.23 19 Oil 188 -4.37% 0.1423 -4.21 No Name N Mean Std t-value 17 Other 825 -7.22% 0.1670 -12.42 24 Paper 12 -4.46% 0.1402 -1.10 7 Utils 11 -4.32% 0.1128 -1.27 11 Machn 189 -7.35% 0.1473 -6.86 15 Autos 20 -4.79% 0.1322 -1.62 3 Oil 189 -4.46% 0.1426 -4.30 * 7 Cnsum 86 -7.48% 0.1760 -3.94 29 Fin 209 -5.22% 0.1337 -5.64 1 NoDur 87 -5.04% 0.1508 -3.12 2 Mines 97 -8.25% 0.2115 -3.84 2 Beer 5 -5.57% 0.2149 -0.58 9 Money 212 -5.15% 0.1328 -5.65 13 Trans 34 -8.32% 0.1201 -4.04 4 Games 74 -6.15% 0.1599 -3.31 2 Durbl 58 -5.71% 0.1278 -3.40 8 Cnstr 60 -9.24% 0.1891 -3.78 25 Trans 24 -6.59% 0.1010 -3.20 10 Other 304 -6.85% 0.1924 -6.21 15 Rtail 102 -9.52% 0.2087 -4.61 26 Whlsl 99 -6.71% 0.1839 -3.63 6 Telcm 75 -7.11% 0.1828 -3.37 4 Clths 25 -12.80% 0.2291 -2.79 * 8 Hlth 211 -6.87% 0.1559 -6.40 8 Shops 544 -7.38% 0.1721 -10.00 21 Telcm 75 -7.11% 0.1828 -3.37 4 Chems 103 -7.62% 0.1631 -4.74 2 Sample Test for Equal Means T = 1.79 22 Servs 321 -7.14% 0.1621 -7.89 5 Manuf 386 -8.23% 0.1543 -10.47 * 28 Meals 56 -7.48% 0.1946 -2.88 11 Cnstr 57 -7.90% 0.1940 -3.08 2 Sample Test for Equal Means T = 2.89 23 BusEq 134 -8.08% 0.1467 -6.38 17 Mines 96 -8.10% 0.2121 -3.74 13 FabPr 43 -8.78% 0.1526 -3.77 Twelve Industries 14 ElcEq 85 -9.64% 0.1923 -4.62 No Name N Mean Std t-value 27 Rtail 67 -10.26% 0.1997 -4.21 8 Utils 11 -4.32% 0.1128 -1.27 6 Hshld 23 -13.15% 0.1746 -3.61 4 Enrgy 189 -4.46% 0.1426 -4.30 * 16 Carry 6 -14.10% 0.0913 -3.78 1 NoDur 87 -5.04% 0.1508 -3.12 7 Clths 16 -17.40% 0.2107 -3.30 11 Money 212 -5.15% 0.1328 -5.65 18 Coal 1 -22.22% 0.0000 3 Manuf 149 -6.25% 0.1439 -5.31 5 Chems 29 -6.41% 0.1462 -2.36 2 Sample Test for Equal Means T = 2.43 2 Durbl 41 -6.85% 0.1371 -3.20 10 Hlth 211 -6.87% 0.1559 -6.40 7 Telcm 75 -7.11% 0.1828 -3.37 12 Other 392 -7.34% 0.1808 -8.04 9 Shops 214 -7.87% 0.1931 -5.96 6 BusEq 359 -8.29% 0.1658 -9.47 *

2 Sample Test for Equal Means T = 2.82

112 Table 14 (Continued)

Thirty-eight Industries Forty-eight Industries No Name N Mean Std t-value No Name N Mean Std t-value 38 Other 1 16.67% 0.0000 1 Agric 8 4.02% 0.0898 1.27

15 Ptrlm 2 4.51% 0.0233 2.73 * 5 Smoke 3 1.56% 0.0327 0.83

7 Smoke 3 1.56% 0.0327 0.83 16 Txtls 2 1.12% 0.1021 0.15 8 Txtls 2 1.12% 0.1021 0.15 3 Soda 1 0.00% 0.0000 1 Agric 12 0.90% 0.0994 0.31 20 FabPr 4 -0.51% 0.0582 -0.18 25 Manuf 13 -1.84% 0.0897 -0.74 8 Books 13 -0.67% 0.1203 -0.20 17 Lethr 3 -1.89% 0.0974 -0.34 15 Rubbr 12 -0.72% 0.0930 -0.27 16 Rubbr 14 -1.97% 0.0957 -0.77 45 Insur 25 -1.79% 0.0717 -1.25

19 Metal 13 -1.97% 0.1410 -0.50 19 Steel 13 -1.97% 0.1410 -0.50

6 Food 31 -3.20% 0.1230 -1.45 2 Food 25 -2.85% 0.1051 -1.36 28 TV 25 -3.36% 0.1670 -1.01 14 Chems 22 -3.18% 0.1173 -1.27 13 Print 20 -3.51% 0.1634 -0.96 31 Util 10 -3.24% 0.1127 -0.91 20 MtlPr 26 -3.54% 0.1189 -1.52 11 Hlth 75 -3.49% 0.1561 -1.94 29 Utils 11 -4.32% 0.1128 -1.27 38 Paper 9 -4.07% 0.1532 -0.80 3 Oil 186 -4.46% 0.1428 -4.26 17 BldMt 35 -4.16% 0.1117 -2.20 * 30 Garbg 15 -4.47% 0.2249 -0.77 30 Oil 188 -4.37% 0.1423 -4.21 ** 35 Money 212 -5.15% 0.1328 -5.65 47 Fin 103 -4.48% 0.1116 -4.07 12 Paper 11 -6.18% 0.1307 -1.57 23 Autos 20 -4.79% 0.1322 -1.62 36 Srvc 481 -6.40% 0.1611 -8.71 48 Other 31 -5.15% 0.1652 -1.74 26 Trans 24 -6.59% 0.1010 -3.20 7 Fun 63 -5.56% 0.1606 -2.75 10 Wood 4 -6.63% 0.1075 -1.23 4 Beer 5 -5.57% 0.2149 -0.58 33 Whlsl 99 -6.71% 0.1839 -3.63 39 Boxes 3 -5.60% 0.1185 -0.82 23 Cars 24 -6.92% 0.1501 -2.26 46 RlEst 33 -6.23% 0.1510 -2.37 14 Chems 103 -7.62% 0.1631 -4.74 40 Trans 24 -6.59% 0.1010 -3.20 2 Mines 92 -7.79% 0.2132 -3.50 41 Whlsl 99 -6.71% 0.1839 -3.63 21 Machn 92 -8.15% 0.1496 -5.22 35 Comps 57 -6.94% 0.1594 -3.29 27 Phone 50 -8.99% 0.1890 -3.36 34 BusSv 299 -7.01% 0.1642 -7.39 18 Glass 8 -9.27% 0.1943 -1.35 32 Telcm 75 -7.11% 0.1828 -3.37 34 Rtail 109 -9.29% 0.2049 -4.74 27 Gold 72 -7.45% 0.1594 -3.97 22 Elctr 150 -9.30% 0.1658 -6.87 43 Meals 56 -7.48% 0.1946 -2.88 24 Instr 86 -9.36% 0.1395 -6.22 44 Banks 48 -7.90% 0.1798 -3.04 11 Chair 2 -12.37% 0.1632 -1.07 13 Drugs 74 -8.09% 0.1699 -4.10 5 Cnstr 22 -13.86% 0.2721 -2.39 33 PerSv 29 -8.82% 0.1547 -3.07 31 Steam 1 -14.29% 0.0000 36 Chips 57 -8.87% 0.1230 -5.44 37 Govt 1 -15.00% 0.0000 37 LabEq 20 -9.11% 0.1749 -2.33 4 Stone 5 -16.77% 0.1716 -2.19 12 MedEq 62 -9.49% 0.1315 -5.68 9 Apprl 16 -20.39% 0.2043 -3.99 * 6 Toys 11 -9.53% 0.1588 -1.99 32 Water 21 Mach 39 -9.63% 0.1571 -3.83 22 ElcEq 85 -9.64% 0.1923 -4.62 2 Sample Test for Equal Means T = 4.64 25 Ships 1 -10.00% 28 Mines 24 -10.07% 0.0000 42 Rtail 67 -10.26% 0.1997 -4.21 9 Hshld 23 -13.15% 0.1746 -3.61 18 Cnstr 22 -13.86% 0.2721 -2.39 24 Aero 5 -14.93% 0.0996 -3.35 10 Clths 16 -17.40% 0.2107 -3.30 * 29 Coal 1 -22.22% 0.0000 26 Guns

2 Sample Test for Equal Means T = 2.37 113 Table 15 Returns Based on Reason Given Managers of reverse split firms relate their motivations for reverse splitting the firm’s stock. This table shows the mean ex-date return for firms grouped according to reasons given for the reverse stock split. The symbols $, *, **, *** represent significance at the 10% level, 5% level, 1% level or 0.5% level respectively.

Day N Mean Patell Z Day N Mean Patell Z Delisting Attention -2 373 1.17% 2.27 * -2 96 1.65% 2.31 * -1 373 -1.52% -3.50 *** -1 96 -0.19% -0.22 0 373 -6.31% -11.28 *** 0 96 -3.35% -5.65 *** +1 373 0.80% 0.69 +1 96 2.15% 2.64 ** +2 373 1.59% 2.57 * +2 96 -0.16% -1.06

Reorgization Liquidity -2 147 2.93% 2.84 ** -2 34 0.02% 0.17 -1 147 -0.62% -0.24 -1 34 0.14% 0.30 0 147 -5.74% -8.20 *** 0 34 -2.96% -1.18 +1 147 -0.38% -1.43 +1 34 0.46% 0.57 +2 147 -0.71% -0.13 +2 34 -0.53% -1.14

Stay on the NMS (Nasdaq) Margin Requirements -2 34 0.67% 1.00 -2 5 0.01% -0.07 -1 34 -1.27% -0.51 -1 5 -2.14% -0.81 0 34 -5.51% -4.26 *** 0 5 -1.86% -0.41 +1 34 2.14% 0.65 +1 5 -0.51% -0.48 +2 34 -2.48% -2.47 * +2 5 0.01% 0.03

Merger or Acquisition Number of Shares -2 85 0.96% 1.08 -2 14 0.78% 0.57 -1 85 1.92% -0.05 -1 14 1.99% 0.32 0 85 -3.51% -2.22 * 0 14 -0.77% -0.97 +1 85 -0.95% -0.15 +1 14 -0.52% -0.69 +2 85 -1.64% -1.95 $ +2 14 -1.31% -1.04

Trading Range Privitization -2 69 -1.60% -2.20 * -2 3 2.22% 0.37 -1 69 0.24% 0.82 -1 3 3.79% 1.01 0 69 -2.46% -2.84 ** 0 3 10.19% 1.07 +1 69 0.54% 0.32 +1 3 -5.15% -0.74 +2 69 -0.13% -0.86 +2 3 0.35% 0.10

114 Table 16 The Returns of Reverse Stock Splits by Price This table shows the ex-date returns of reverse stock splits whose pre-spilt price was less than $1.00 versus firms whose stock price was more than $1.00 before the split. Also shown are the ex-date returns of firms that had a post split target price of less than $5.00 versus firms with a target price of more than $5.00. Target price was calculated based on pre-split price and the split factor. For example a firm with a pre-split price of $0.60 that reverse split 10 to 1 would have a post split target price of $6.00.

Ex date Pre split price less than $1.00 N Mean Std min max Raw Return 1386 -8.58% 18.10% -80.0% 60.0% Market Adjusted Return (equal weighted) 1386 -8.67% 18.07% -80.8% 60.1% Market Adjusted Return (value weighted) 1386 -8.63% 18.10% -80.6% 60.8% Market Model Adjusted Return (equal weighted) 1374 -8.77% 18.12% -80.1% 60.9% Market Model Adjusted Return (value weighted) 1374 -8.75% 18.11% -80.1% 60.9%

Pre split price greater than $1.00 N mean Std min max Raw Return 615 -2.49% 10.10% -92.3% 42.2% Market Adjusted Return (equal weighted) 615 -2.59% 10.09% -92.4% 42.7% Market Adjusted Return (value weighted) 615 -2.59% 10.14% -92.8% 42.9% Market Model Adjusted Return (equal weighted) 605 -2.65% 10.28% -95.6% 42.6% Market Model Adjusted Return (value weighted) 605 -2.56% 10.65% -94.5% 71.8% t-value to test difference in means Raw Return -1515.1 Market Adjusted Return (equal weighted) -1515.7 Market Adjusted Return (value weighted) -1495.1 Market Model Adjusted Return (equal weighted) -1479.0 Market Model Adjusted Return (value weighted) -1452.6

Post split target less than $5.00 N mean Std min max Raw Return 1235 -8.30% 18.32% -92.3% 60.0% Market Adjusted Return (equal weighted) 1235 -8.41% 18.29% -92.4% 60.1% Market Adjusted Return (value weighted) 1235 -8.37% 18.30% -92.8% 60.8% Market Model Adjusted Return (equal weighted) 1222 -8.47% 18.34% -92.6% 60.9% Market Model Adjusted Return (value weighted) 1222 -8.47% 18.32% -92.2% 60.9%

Post spilt target greater than $5.00 N mean Std min max Raw Return 781 -4.32% 12.17% -91.5% 50.0% Market Adjusted Return (equal weighted) 781 -4.38% 12.16% -91.5% 49.5% Market Adjusted Return (value weighted) 781 -4.37% 12.21% -91.6% 49.5% Market Model Adjusted Return (equal weighted) 771 -4.55% 12.37% -95.6% 49.4% Market Model Adjusted Return (value weighted) 771 -4.45% 12.63% -94.5% 71.8% t-value to test difference in means Raw Return -862.8 Market Adjusted Return (equal weighted) -876.1 Market Adjusted Return (value weighted) -864.0 Market Model Adjusted Return (equal weighted) -826.6 Market Model Adjusted Return (value weighted) -836.4

Post split taraget greater than $10.00 N mean Std min max Raw Return 472 -3.6% 11.31% -91.5% 40.6% Market Adjusted Return (equal weighted) 472 -3.7% 11.26% -91.5% 41.0% Market Adjusted Return (value weighted) 472 -3.6% 11.32% -91.6% 40.0% Market Model Adjusted Return (equal weighted) 464 -3.8% 11.57% -95.6% 40.2% Market Model Adjusted Return (value weighted) 464 -3.6% 12.03% -94.5% 71.8%

115

Table 17 The Returns of Reverse stock splits and the Size of the Firm The dependent variable is regressed on the size of the firm proxied by number of shares times price. Announcement day returns are the two day cumulative return on the announcement day and the day after the announcement. Raw returns, market model adjusted returns and market adjusted returns are used as the dependent varaible

Ex day returns regressed on firm size intercept market value (mv) Dependant estimate t value estimate t value Ex Date Raw Return 0.00499 0.43 0.00000 -0.42

Ex Date Market Adjusted Return 0.00395 0.36 0.00000 -0.51 (equally weighted index)

Ex Date Market Adjusted Return 0.00303 0.28 0.00000 -0.36 (value weighted index)

Ex Date Market Model Adjusted Return 0.00277 0.27 0.00000 -0.46 (equally weighted index)

Ex Date Market Model Adjusted Return 0.00186 0.18 0.00000 -0.34 (value weighted index)

116 Table 18 The Returns of Reverse Stock Splits and Consolidation The dependent variable is regressed on consolidation. Announcement day returns are the two day cumulative return from the day of the announcement to the day after the announcement. Raw returns, market model adjusted returns and market adjusted returns are used as the dependent varaible

Announcement Day Returns Regressed on Consolidation intercept shares consolidated Dependant estimate t value estimate t value AD Raw Return -0.0339 -4.73 -0.0007 -1.81

AD Market Adjusted Return -0.0357 -4.99 -0.0007 -1.85 (equally weighted index)

AD Market Adjusted Return -0.0351 -4.91 -0.0007 -1.81 (value weighted index)

AD Market Model Adjusted Return -0.0362 -5.09 -0.0008 -2.17 (equally weighted index)

AD Market Model Adjusted Return -0.0365 -5.11 -0.0008 -2.19 (value weighted index)

Ex-Date Returns Regressed on Consolidation intercept shares consolidated Dependant estimate t value estimate t value Ex Date Raw Return -0.0554 -12.83 -0.0013 -6.09

Ex Date Market Adjusted Return -0.0563 -13.05 -0.0013 -6.06 (equally weighted index)

Ex Date Market Adjusted Return -0.0561 -12.97 -0.0013 -6.01 (value weighted index)

Ex Date Market Model Adjusted Return -0.0559 -12.90 -0.0014 -6.41 (equally weighted index)

Ex Date Market Model Adjusted Return -0.0565 -13.10 -0.0014 -6.37 (value weighted index)

117 Table 19 Reverse Stock Split Delistings Within 250 Days of Ex-date (1962-2003) The delisting codes and delisting returns are taken from the Center for Research in Security Prices (CRSP). For firms that reverse split their stock, then number of firms delisted within 250 trading days (about 1 year) is reported. The number of each type of delisting is shown with its percentage of the total delistings. The average delisting return for each category is reported.

Percent of Average Delisting Delisting Code Code Explaination Number Total Return 231 Merged and shareholders received common stock 14 6% 1.3%

233 Issue exchanged primarily for cash 11 4% 1.2%

261 Issue exchanged for combination of cash and other securities 1 0% 0.0%

331 Issue exchanged for another class of stock 1 0% -42.3%

460 Issue liquidated - no further information is available 1 0% 0.0%

500 Issue stopped trading on exchange - reason unavailable 5 2% 8.1%

550 Delisted insufficient number of market makers 5 2% -34.4%

551 Delisted insufficient number of shareholders 1 0% -12.5%

552 Delisting - price fell below acceptable level 57 23% -12.8%

560 Delisted - insufficient capital, surplus, and/or equity 32 13% -10.9%

561 Delisted - insufficient (or non-compliance with rules of) float or assets 32 13% -13.2%

570 Delisted - company request (no reason given) 5 2% -24.3%

573 Delisted - company request, liquidation 2 1% -0.8%

574 Delisted - company request, bankruptcy, declared insolvent 11 4% -30.1%

580 Delisted - delinquent in filing or non-payment of fees 23 9% -14.2%

582 Delisted - failure to meet exception or equity requirements 16 6% -17.7%

584 Delisted - does not meet exchange's financial guidelines for continued listing 15 6% -25.5%

585 Delisted - protection of investors and the public interest 12 5% -16.3%

587 Delisted - corporate governance violation 2 1% -14.1%

591 Delisted - delisting required by Securities Exchange Commission 1 0% 0.0%

Total 247 100% -13.5%

118 Table 20 Average Buy and Hold Returns of Reverse stock splits from 1962 to 2003 Average BHRs are presented for holding periods of up to 250 days (about 1 calendar year). Returns are missing in the data and these missing returns are handled three ways. A missing return is replaced with the the daily average portfolio return, zero, or the market return.

Missing returns Missing returns replaced with daily average replaced with zero

Day n mean t Day n mean t 0 1218 -2.80% -5.88 0 1218 -2.80% -5.88 10 1218 -7.31% -9.25 10 1218 -7.31% -9.25 20 1218 -9.28% -9.56 20 1218 -9.27% -9.55 30 1218 -9.87% -9.04 30 1218 -9.88% -9.04 40 1218 -9.51% -7.67 40 1218 -9.54% -7.70 50 1218 -10.25% -7.86 50 1218 -10.30% -7.90 60 1218 -9.76% -6.62 60 1218 -9.85% -6.68 70 1218 -9.45% -6.18 70 1218 -9.63% -6.31 80 1218 -8.83% -5.32 80 1218 -9.07% -5.47 90 1218 -8.66% -4.79 90 1218 -8.94% -4.95 100 1218 -8.70% -4.52 100 1218 -9.05% -4.71 110 1218 -6.87% -3.25 110 1218 -7.35% -3.48 120 1218 -6.99% -3.38 120 1218 -7.51% -3.64 130 1218 -6.77% -3.25 130 1218 -7.39% -3.56 140 1218 -6.09% -2.79 140 1218 -6.83% -3.14 150 1218 -4.56% -1.89 150 1218 -5.46% -2.27 160 1218 -3.93% -1.60 160 1218 -5.03% -2.06 170 1218 -2.61% -0.98 170 1218 -3.88% -1.46 180 1218 0.03% 0.01 180 1218 -1.45% -0.42 190 1218 1.06% 0.30 190 1218 -0.59% -0.17 200 1218 1.81% 0.46 200 1218 0.05% 0.01 210 1218 3.36% 0.88 210 1218 1.38% 0.36 220 1218 6.74% 1.29 220 1218 4.48% 0.86 230 1218 7.67% 1.50 230 1218 5.15% 1.01 240 1218 8.70% 1.92 240 1218 5.87% 1.30 250 1218 8.80% 2.25 250 1218 5.60% 1.44

119 Table 20 (Continued)

Missing returns Missing returns replaced with equal weighted index replaced with value weighted index

Day n mean t Day n mean t 0 1218 -2.80% -5.88 0 1218 -2.80% -5.88 10 1218 -7.31% -9.25 10 1218 -7.31% -9.25 20 1218 -9.26% -9.56 20 1218 -9.27% -9.56 30 1218 -9.85% -9.03 30 1218 -9.85% -9.02 40 1218 -9.46% -7.62 40 1218 -9.49% -7.65 50 1218 -10.16% -7.77 50 1218 -10.22% -7.82 60 1218 -9.67% -6.53 60 1218 -9.77% -6.61 70 1218 -9.42% -6.14 70 1218 -9.54% -6.24 80 1218 -8.83% -5.30 80 1218 -8.97% -5.40 90 1218 -8.62% -4.75 90 1218 -8.82% -4.88 100 1218 -8.66% -4.47 100 1218 -8.88% -4.61 110 1218 -6.93% -3.27 110 1218 -7.18% -3.39 120 1218 -7.03% -3.39 120 1218 -7.31% -3.53 130 1218 -6.82% -3.27 130 1218 -7.13% -3.43 140 1218 -6.16% -2.81 140 1218 -6.52% -2.99 150 1218 -4.68% -1.94 150 1218 -5.12% -2.13 160 1218 -4.13% -1.68 160 1218 -4.63% -1.89 170 1218 -2.81% -1.05 170 1218 -3.40% -1.27 180 1218 -0.25% -0.07 180 1218 -0.94% -0.27 190 1218 0.74% 0.21 190 1218 -0.03% -0.01 200 1218 1.47% 0.37 200 1218 0.62% 0.16 210 1218 2.94% 0.77 210 1218 2.01% 0.53 220 1218 6.20% 1.19 220 1218 5.17% 0.99 230 1218 6.99% 1.37 230 1218 5.84% 1.15 240 1218 7.80% 1.73 240 1218 6.53% 1.45 250 1218 7.59% 1.95 250 1218 6.27% 1.61

120 Table 21 Average Market Adjusted Buy and Hold Abnormal Returns of Reverse stock splits from 1962 to 2003 The BHARs are calculated by subtracting a market index from the average BHR. Missing returns in the data are either replaced with the average return for that day (this page) or replaced with the daily market index return (next page).

Market Adjusted BHARs using the CRSP Equal Weighted Index Market Adjusted BHARs using the CRSP Value Weighted Index Missing returns replaced with daily average Missing returns replaced with daily average BHAR =-BHR ewBHR BHAR=- BHR vwBHR Days N mean t mean t mean t Days N mean t mean t mean t 0 1218 -2.92% -6.16 -2.80% -5.88 0.13% 5.54 0 1218 -2.87% -6.05 -2.80% -5.88 0.08% 2.54 10 1218 -8.34% -10.78 -7.31% -9.25 1.04% 9.69 10 1218 -7.76% -9.93 -7.31% -9.25 0.45% 4.56 20 1218 -11.42% -12.12 -9.28% -9.56 2.14% 13.10 20 1218 -10.22% -10.73 -9.28% -9.56 0.94% 6.93 30 1218 -12.99% -12.29 -9.87% -9.04 3.12% 15.27 30 1218 -11.22% -10.53 -9.87% -9.04 1.35% 8.56 40 1218 -13.65% -11.38 -9.51% -7.67 4.14% 17.08 40 1218 -11.22% -9.24 -9.51% -7.67 1.71% 9.14 50 1218 -15.28% -12.12 -10.25% -7.86 5.04% 18.03 50 1218 -12.27% -9.65 -10.25% -7.86 2.03% 9.72 60 1218 -15.86% -11.15 -9.76% -6.62 6.10% 19.58 60 1218 -12.29% -8.53 -9.76% -6.62 2.53% 11.20 70 1218 -16.61% -11.31 -9.45% -6.18 7.16% 21.06 70 1218 -12.41% -8.33 -9.45% -6.18 2.96% 12.08 80 1218 -16.97% -10.65 -8.83% -5.32 8.15% 22.12 80 1218 -12.12% -7.50 -8.83% -5.32 3.29% 12.48 90 1218 -17.98% -10.41 -8.66% -4.79 9.31% 23.81 90 1218 -12.46% -7.09 -8.66% -4.79 3.80% 13.58 100 1218 -19.30% -10.48 -8.70% -4.52 10.60% 26.21 100 1218 -13.08% -6.97 -8.70% -4.52 4.38% 15.23 110 1218 -18.65% -9.21 -6.87% -3.25 11.77% 27.80 110 1218 -11.70% -5.69 -6.87% -3.25 4.83% 16.20 120 1218 -20.24% -10.25 -6.99% -3.38 13.25% 30.59 120 1218 -12.56% -6.24 -6.99% -3.38 5.58% 18.29 130 1218 -21.45% -10.81 -6.77% -3.25 14.69% 32.61 130 1218 -12.94% -6.38 -6.77% -3.25 6.18% 19.62 140 1218 -22.31% -10.70 -6.09% -2.79 16.22% 34.24 140 1218 -12.88% -6.03 -6.09% -2.79 6.79% 20.78 150 1218 -22.11% -9.57 -4.56% -1.89 17.55% 35.23 150 1218 -11.72% -4.96 -4.56% -1.89 7.16% 20.97 160 1218 -22.98% -9.72 -3.93% -1.60 19.05% 36.21 160 1218 -11.66% -4.83 -3.93% -1.60 7.73% 21.80 170 1218 -23.18% -8.97 -2.61% -0.98 20.57% 37.69 170 1218 -10.97% -4.17 -2.61% -0.98 8.36% 22.80 180 1218 -21.95% -6.45 0.03% 0.01 21.98% 38.59 180 1218 -8.72% -2.53 0.03% 0.01 8.75% 23.21 190 1218 -22.33% -6.40 1.06% 0.30 23.39% 39.33 190 1218 -8.20% -2.32 1.06% 0.30 9.26% 23.64 200 1218 -22.96% -5.93 1.81% 0.46 24.78% 40.42 200 1218 -7.98% -2.03 1.81% 0.46 9.80% 24.24 210 1218 -22.76% -6.10 3.36% 0.88 26.12% 41.26 210 1218 -6.94% -1.83 3.36% 0.88 10.30% 24.90 220 1218 -20.79% -4.06 6.74% 1.29 27.54% 42.01 220 1218 -4.06% -0.78 6.74% 1.29 10.80% 25.11 230 1218 -21.48% -4.29 7.67% 1.50 29.15% 43.07 230 1218 -3.73% -0.74 7.67% 1.50 11.40% 25.79 240 1218 -21.97% -4.96 8.70% 1.92 30.66% 44.16 240 1218 -3.30% -0.73 8.70% 1.92 11.99% 26.43 250 1218 -23.19% -6.11 8.80% 2.25 31.98% 45.08 250 1218 -3.58% -0.93 8.80% 2.25 12.37% 26.56

121 Table 21 (Continued)

Market Adjusted BHARs using the CRSP Equal Weighted Index Market Adjusted BHARs using the CRSP Value Weighted Index Missing returns replaced with CRSP equally weighted index Missing returns replaced with CRSP value weighted index BHAR=- BHR ewBHR BHAR =-BHR vwBHR Days N mean t mean t mean t Days N mean t mean t mean t 0 1218 -2.92% -6.16 -2.80% -5.88 0.13% 5.54 0 1218 -2.92% -6.16 -2.80% -5.88 0.08% 2.54 10 1218 -8.34% -10.78 -7.31% -9.25 1.04% 9.69 10 1218 -7.76% -9.93 -7.31% -9.25 0.45% 4.56 20 1218 -11.40% -12.12 -9.26% -9.56 2.14% 13.10 20 1218 -10.21% -10.74 -9.27% -9.56 0.94% 6.93 30 1218 -12.97% -12.29 -9.85% -9.03 3.12% 15.27 30 1218 -11.20% -10.51 -9.85% -9.02 1.35% 8.56 40 1218 -13.60% -11.33 -9.46% -7.62 4.14% 17.08 40 1218 -11.20% -9.22 -9.49% -7.65 1.71% 9.14 50 1218 -15.20% -12.03 -10.16% -7.77 5.04% 18.03 50 1218 -12.25% -9.61 -10.22% -7.82 2.03% 9.72 60 1218 -15.76% -11.05 -9.67% -6.53 6.10% 19.58 60 1218 -12.29% -8.53 -9.77% -6.61 2.53% 11.20 70 1218 -16.58% -11.25 -9.42% -6.14 7.16% 21.06 70 1218 -12.51% -8.39 -9.54% -6.24 2.96% 12.08 80 1218 -16.98% -10.62 -8.83% -5.30 8.15% 22.12 80 1218 -12.27% -7.58 -8.97% -5.40 3.29% 12.48 90 1218 -17.94% -10.35 -8.62% -4.75 9.31% 23.81 90 1218 -12.62% -7.18 -8.82% -4.88 3.80% 13.58 100 1218 -19.26% -10.41 -8.66% -4.47 10.60% 26.21 100 1218 -13.26% -7.07 -8.88% -4.61 4.38% 15.23 110 1218 -18.71% -9.21 -6.93% -3.27 11.77% 27.80 110 1218 -12.00% -5.84 -7.18% -3.39 4.83% 16.20 120 1218 -20.29% -10.25 -7.03% -3.39 13.25% 30.59 120 1218 -12.88% -6.40 -7.31% -3.53 5.58% 18.29 130 1218 -21.51% -10.83 -6.82% -3.27 14.69% 32.61 130 1218 -13.31% -6.57 -7.13% -3.43 6.18% 19.62 140 1218 -22.37% -10.72 -6.16% -2.81 16.22% 34.24 140 1218 -13.31% -6.25 -6.52% -2.99 6.79% 20.78 150 1218 -22.24% -9.63 -4.68% -1.94 17.55% 35.23 150 1218 -12.28% -5.21 -5.12% -2.13 7.16% 20.97 160 1218 -23.18% -9.81 -4.13% -1.68 19.05% 36.21 160 1218 -12.37% -5.14 -4.63% -1.89 7.73% 21.80 170 1218 -23.38% -9.05 -2.81% -1.05 20.57% 37.69 170 1218 -11.76% -4.48 -3.40% -1.27 8.36% 22.80 180 1218 -22.22% -6.53 -0.25% -0.07 21.98% 38.59 180 1218 -9.69% -2.81 -0.94% -0.27 8.75% 23.21 190 1218 -22.65% -6.50 0.74% 0.21 23.39% 39.33 190 1218 -9.29% -2.63 -0.03% -0.01 9.26% 23.64 200 1218 -23.30% -6.02 1.47% 0.37 24.78% 40.42 200 1218 -9.18% -2.34 0.62% 0.16 9.80% 24.24 210 1218 -23.18% -6.22 2.94% 0.77 26.12% 41.26 210 1218 -8.30% -2.20 2.01% 0.53 10.30% 24.90 220 1218 -21.34% -4.16 6.20% 1.19 27.54% 42.01 220 1218 -5.63% -1.09 5.17% 0.99 10.80% 25.11 230 1218 -22.16% -4.42 6.99% 1.37 29.15% 43.07 230 1218 -5.56% -1.10 5.84% 1.15 11.40% 25.79 240 1218 -22.86% -5.18 7.80% 1.73 30.66% 44.16 240 1218 -5.46% -1.22 6.53% 1.45 11.99% 26.43 250 1218 -24.40% -6.46 7.59% 1.95 31.98% 45.08 250 1218 -6.10% -1.59 6.27% 1.61 12.37% 26.56

122 Table 22 Matched Firm Statistics Firms are matched on SIC, price, and size. Firm size is proxied by the number of shares times price per share. This table shows how close the sample was matched. Standard deviation is shown in parenthesis.

Matches Found - 697

Average Price – Sample Firms $3.73 (4.957) Average Price – Matched Firms $3.71 (4.961) Difference 0.02 % Difference 0.54%

Average Size – Sample Firms 47,618,470 (202,323,470) Average Size – Matched Firms 46,648,060 (192,322,130) Difference 970,410 % Difference 2.04%

123 Table 23 Average Abnormal Returns of Reverse stock splits from 1962 to 2003 Based on Matched Firm This table shows the abnormal buy-and-hold returns for reverse stock splits. The buy-and-hold abnormal returns shown are the average returns for holding a reverse split stock for a certain number of days minus the average return of a portfolio of firms matched on industry, size, and price. The two tailed t-stat tests mean not equal to zero.

Average Abnormal Holding Period Sample Matched Holding Period Return unpaired t 10 -3.81% 0.15% -3.96% -3.12 20 -6.88% 0.43% -7.31% -4.63 30 -8.28% -0.19% -8.09% -4.26 40 -7.95% 0.87% -8.82% -4.18 50 -8.34% 2.67% -11.01% -4.47 60 -7.57% 2.59% -10.16% -3.68 70 -6.00% 3.97% -9.97% -3.37 80 -5.55% 5.66% -11.21% -3.49 90 -3.76% 8.20% -11.97% -3.09 100 -3.39% 9.44% -12.83% -3.12 110 -1.32% 11.08% -12.40% -2.54 120 -2.12% 14.77% -16.88% -3.16 130 -1.66% 15.65% -17.31% -2.9 140 0.93% 18.10% -17.17% -2.68 150 4.46% 17.92% -13.46% -2.18 160 5.53% 18.52% -12.99% -2.22 170 6.57% 20.43% -13.87% -2.28 180 10.31% 20.95% -10.64% -1.64 190 12.67% 21.57% -8.91% -1.37 200 13.96% 24.85% -10.89% -1.57 210 17.14% 26.53% -9.40% -1.28 220 22.42% 28.65% -6.23% -0.7 230 21.14% 31.23% -10.08% -1.18 240 22.86% 31.62% -8.76% -1.04 250 23.27% 31.53% -8.26% -1.01

124 Table 24 Long Term Returns By Industry This table shows the 250-day raw returns by industry using the Fama French Industry Classifications. A 2-sample t-test for difference in means is performed on the highest and lowest average return. In addition, a 2-sample t-test is performed on the highest significant average return and the lowest significant average return. The asterisk denotes highest and lowest significant average return.

5 Industries N Mean t 2-sample T 2-sample T 2 Utils 14 -0.2822 -2.87 * High vs. Low Significant high 1 Manuf 337 0.1724 2.49 vs significant low 5 Other 471 0.2121 3.13 4 Money 116 0.2198 1.68 3 Shops 124 0.2722 2.01 * 19.78 19.78

10 Industries N Mean t 7 Utils 5 -0.1062 -0.56 2 Durbl 38 -0.0939 -0.91 1 NoDur 46 -0.0385 -0.33 3 Oil 107 0.0882 1.00 10 Other 153 0.0974 1.25 9 Money 116 0.2198 1.68 * 5 Manuf 201 0.2319 2.49 8 Shops 302 0.2764 2.85 * 4 Chems 56 0.3333 1.51 6 Telcm 38 0.5410 1.42 3.60 2.14

12 Industries N Mean t 2-sample T 2-sample T 8 Utils 5 -0.1062 -0.56 High vs. Low Significant high 2 Durbl 25 -0.0873 -0.61 vs significant low 1 NoDur 46 -0.0385 -0.33 3 Manuf 84 0.0559 0.75 12 Other 199 0.0592 0.67 4 Enrgy 107 0.0882 1.00 10 Hlth 114 0.0969 0.79 11 Money 116 0.2198 1.68 9 Shops 128 0.2591 1.97 * 6 BusEq 186 0.5074 3.71 5 Chems 14 0.5129 2.01 * 7 Telcm 38 0.5410 1.42 3.60 3.07

125 Table 24 (Continued)

17 Industries N Mean t 2-sample T 2-sample T 5 Durbl 21 -0.2165 -2.29 * High vs. Low Significant high 1 Food 25 -0.1505 -1.15 vs significant low 13 Trans 16 -0.1311 -0.80 2 Mines 32 -0.1216 -1.17 12 Cars 16 -0.0791 -0.48 14 Utils 4 -0.0045 -0.02 10 FabPr 5 0.0572 0.26 6 Chems 11 0.0575 0.37 3 Oil 109 0.0905 1.05 16 Finan 114 0.1088 1.66 9 Steel 8 0.2236 1.11 11 Machn 108 0.2243 1.94 4 Clths 14 0.2619 0.81 17 Other 439 0.2897 3.52 * 8 Cnstr 28 0.3019 1.31 15 Rtail 62 0.3405 1.51 7 Cnsum 50 0.3844 1.56 8.62 32.28

126 Table 24 (Continued)

30 Industries N Mean t 2-sample T 2-sample T 2 Beer 3 -0.3413 -1.70 High vs. Low Significant high 3 Smoke 1 -0.2650 vs significant low 25 Trans 12 -0.2537 -1.31 1 Food 18 -0.1463 -1.21 30 Other 28 -0.1288 -1.03 17 Mines 32 -0.1216 -1.17 15 Autos 14 -0.0906 -0.55 10 Txtls 2 -0.0241 -0.03 20 Util 4 -0.0045 -0.02 13 FabPr 25 -0.0025 -0.03 4 Games 42 0.0654 0.20 19 Oil 107 0.0882 1.00 8 Hlth 114 0.0969 0.79 9 Chems 10 0.1227 0.78 26 Whlsl 58 0.2027 1.27 29 Fin 115 0.2234 1.70 12 Steel 8 0.2236 1.11 6 Hshld 14 0.2248 0.74 28 Meals 33 0.2255 1.07 5 Books 9 0.2419 0.59 11 Cnstr 29 0.2512 1.16 23 BusEq 74 0.2992 1.85 24 Paper 7 0.3267 1.00 27 Rtail 41 0.3283 1.08 22 Servs 173 0.3352 2.82 * 16 Carry 2 0.3651 0.73 7 Clths 7 0.4004 0.73 21 Telcm 38 0.5410 1.42 14 Autos 42 0.6049 1.94 * 6.89 2.43

127 Table 24 (Continued)

38 Industries N Mean t 2-sample T 2-sample T 17 Lethr 1 -0.6471 High vs. Low Significant high 38 Other 1 -0.6143 vs significant low 30 Garbg 9 -0.3800 -3.58 * 7 Smoke 1 -0.2650 26 Trans 12 -0.2537 -1.31 6 Food 17 -0.2004 -1.69 * 16 Rubbr 12 -0.1488 -1.12 2 Mines 30 -0.1315 -1.19 29 Utils 5 -0.1062 -0.56 15 Ptrlm 1 -0.0989 20 MtlPr 14 -0.0977 -0.89 25 Manuf 8 -0.0369 -0.11 24 Instr 44 -0.0341 -0.21 8 Txtls 2 -0.0241 -0.03 4 Stone 2 0.0266 0.19 10 Wood 2 0.0388 0.11 23 Cars 15 0.0470 0.29 11 Chair 2 0.0667 0.29 3 Oil 106 0.0899 1.01 21 Machn 52 0.1178 1.12 33 Whlsl 58 0.2027 1.27 35 Money 116 0.2198 1.68 19 Metal 8 0.2236 1.11 13 Print 9 0.2419 0.59 36 Srvc 265 0.2577 2.66 27 Phone 25 0.3065 1.61 14 Chems 56 0.3333 1.51 34 Rtail 66 0.3334 1.56 1 Agric 5 0.3500 0.75 9 Apprl 9 0.3696 0.83 22 Elctr 75 0.4151 1.98 * 12 Paper 6 0.4836 1.42 5 Cnstr 13 0.5137 1.17 18 Glass 2 0.5420 0.87 10.64 28 TV 13 0.9921 0.93 1.20 14.42

128 Table 24 (Continued)

48 Industries N Mean t 2-sample T 2-sample T 4 Beer 3 -0.3413 -1.70 High vs. Low Significant high 5 Smoke 1 -0.2650 vs significant low 40 Trans 12 -0.2537 -1.31 6 Toys 5 -0.1952 -0.92 2 Food 14 -0.1703 -1.23 8 Books 7 -0.1525 -1.95 * 48 Other 18 -0.1443 -0.81 27 Gold 21 -0.1345 -0.87 12 MedEq 29 -0.1179 -0.95 15 Rubbr 10 -0.1011 -0.69 28 Mines 11 -0.0969 -1.25 23 Autos 14 -0.0906 -0.55 46 RlEst 18 -0.0776 -0.77 1 Agric 4 -0.0625 -0.23 16 Txtls 2 -0.0241 -0.03 31 Util 4 -0.0045 -0.02 21 Mach 23 -0.0037 -0.04 20 FabPr 2 0.0111 0.14 45 Insur 14 0.0253 0.14 17 BldMt 16 0.0380 0.23 11 Hlth 43 0.0694 0.46 30 Oil 107 0.0882 1.00 7 Fun 37 0.1006 0.27 14 Chems 10 0.1227 0.78 47 Fin 54 0.1906 2.05 38 Paper 4 0.1936 0.52 36 Chips 31 0.1947 0.70 41 Whlsl 58 0.2027 1.27 19 Steel 8 0.2236 1.11 9 Hshld 14 0.2248 0.74 43 Meals 33 0.2255 1.07 13 Drugs 42 0.2734 0.97 34 BusSv 157 0.3241 2.54 42 Rtail 41 0.3283 1.08 35 Comps 32 0.3632 1.97 24 Aero 2 0.3651 0.73 10 Clths 7 0.4004 0.73 37 LabEq 11 0.4076 0.74 39 Boxes 3 0.5042 0.75 18 Cnstr 13 0.5137 1.17 32 Telcm 38 0.5410 1.42 44 Banks 29 0.5669 1.18 33 PerSv 18 0.5751 1.88 22 ElcEq 42 0.6049 1.94 * 6.89 7.35

129 Table 25 Long Term Returns by Reason Given and Low-Priced Versus High Priced Managers give various reasons in their press releases as to why they are reverse splitting the firm’s stock. This table shows the mean 250-day return grouped by reason given. Buy and hold returns are denoted “BHR” and the buy and hold market adjusted returns using the equal-weighted CRSP index are denoted “BHAR.” It can be argued that low-priced firms are reverse split to avoid delisting – regardless of reason given, so the returns of low- priced versus high-priced firms are given. A low priced firm is priced below $1.00 and a high-priced firm is priced above $1.00.

Reason Given Return N Mean t-value Attract Attention BHR 68 5.01% 0.43 Delisting Threat BHR 195 51.92% 3.72 Increase Liquidity BHR 22 -7.09% -0.56 Margin Requirements BHR 3 -11.39% -0.43 Maintian NMS Requirements BHR 20 21.98% 0.68 No Reason Given BHR 537 16.96% 2.88 Taking the Company Private BHR 1 113.55% Reorginazation BHR 98 19.96% 1.38 Increase Number of Share BHR 9 -12.63% -1.56 Trading Range BHR 55 12.00% 0.97

Attract Attention BHAR 68 -0.42% -0.72 Delisting Threat BHAR 195 0.53% 1.05 Increase Liquidity BHAR 22 -0.91% -0.64 Margin Requirements BHAR 3 -2.52% -1.37 Maintian NMS Requirements BHAR 20 1.85% 1.79 No Reason Given BHAR 537 0.04% 0.13 Taking the Company Private BHAR 1 0.57% Reorginazation BHAR 98 0.18% 0.45 Increase Number of Share BHAR 9 -0.73% -1.02 Trading Range BHAR 55 0.21% 0.33

Low Price BHR 771 27.07% 4.82 High Price BHR 71 2.74% 0.41 2 sample test of equal means t-value 32.11

Low Price BHAR 771 0.13% 0.56 High Price BHAR 71 -0.11% -0.25 2 sample test of equal means t-value 94.39

130 Table 26 Long Term Returns By Size, Consolidation, and Post Split Price The 250-day ex-date BHRs are regressed on the size of the firm proxied by the number of shares times price on the ex-date. The returns are also regressed on consolidation and the post split share price. Consolidation is the number of shares that the shareholder has to convert to one share at the reverse stock split. The regressions are repeated using the 250-day market adjusted BHAR as the dependent variable. The t-values are in italics under the regression coefficients.

Size Vs. 250-Day Returns

Regression Dependent Intercept Coefficient Variable t-value t-value BHR = 0.0015 + 0.0000 X Shares Outstanding X Price 0.92 0.01

BHAR = 0.0002 + 0.0000 X Shares Outstanding X Price 0.12 -0.01

Consolidation Versus 250-Day Returns

Regression Dependent Intercept Coefficient Variable t-value t-value BHR = 0.0027 + -0.0001 X Shares Consolidated 1.41 -1.21

BHAR = 0.0015 + -0.0013 X Shares Consolidated 0.80 -1.34

Post Split Price Versus 250-Day Returns

Regression Dependent Intercept Coefficient Variable t-value t-value BHR = 0.00133 + 0.00006 X Post Split Price 0.77 0.36

BHAR = 0.00003 + 0.00006 X Post Split Price 0.02 0.35

131 Table 27 Monthly Stock Splits Related to Market Movements The number of stock splits per month (as a percentage of total firms in the market) is regressed on recent market returns. The market index is the CRSP equally weighted index. The independent variable is the buy and hold return of the market index for the previous 12 months or the the previous 24 months.

Intercept Parameter Dependent Estimate Estimate Independent Variable t-value t-value Variable

% RSS per Month = 0.00088 + -0.00026 X 12 Month EW Index 23.01 -1.87

% RSS per Month = 0.00096 + -0.00037 X 24 Month EW Index * 21.38 -3.67

132 Table 28 Long Run Returns Regressed on Lagged Market Returns The dependent variable is either the the buy-and-hold return or the market adjusted buy- and-hold market adjusted return. The lagged market is the buy-and-hold market return of the CRSP equal weighted index from 12 or 24 months prior to the ex-date.

Dependent Interc ept Par a meter Independent Variable Es ti mate Es ti mate Variable (long run returns) t-value t-value (lagged mkt)

250 Day BHRs = 0.155 + -0.922 X EW Index * 3.04 -4.80 12 month

250 Day BHRs = 0.168 + -0.520 X EW Index * 2.74 -3.53 24 month

250 Day BHARs = -0.202 + -0.458 X EW Index * EW Index Adjusted -4.47 -2.69 12 month

250 Day BHARs = -0.185 + -0.290 X EW Index * EW Index Adjusted -3.46 -2.25 24 month

250 Day BHARs = -0.012 + -0.666 X EW Index * VW Index Adjusted -0.25 -3.60 12 month

250 Day BHARs = 0.004 + -0.399 X EW Index * VW Index Adjusted 0.07 -2.84 24 month

133 Table 29 Regression Analysis of Financial Distress Variables – Definitions Researchers, such as Theodossiou, Kahya, Saidi, and Philippatos (1996), have identified variables that may be predictive of financial distress. Regression analysis of the full sample has shown that reverse stock splits in general have negative returns. Other research has shown an unfavorable survival rate for firms that reverse split. The purpose of this table is to show correlation between reverse stock splits and some of the accounting values that are used in analyzing financial distress.

Variable Description COMPUSTAT VALUES dta debt to assets data54/data44 opincta operating income to assets data21/data44 invtsls inventories to sales data38/data2 lgsales log of sales log(data2) lgassets log of assets log(data44) nwcta net working capital to assets (data40-data49)/data44 cacl current assets to current liabilibies data40/data49 qacl quick assets to current liabilibies (data40-data38)/data49 ebita EBIT to assets (data23+data22)/data44 reta retained earnings to assets data58/data44 ltdta long term debt to assets data51/data40 mveltd market value of equity to long term debt (data61*abs(pn1))/data51 fata fixed assets to total assets data42/data44

COMPUSTAT data2 sales data21 operating income before depreciation data22 interest expense data23 pretax income data38 inventories data40 current assets data42 property, plant, and equipment data44 assets total data49 current liabilities data51 long term debt data54 liabilities total data58 retained earnings data61 common shares outstanding pn1 price one day prior to split

Returns r0 ex-date raw return ma0e market adjusted ex-date return using CRSP equal-weighted index ma0v market adjusted ex-date return using CRSP value-weighted index mm0e market model adjusted ex-date return using CRSP equal-weighted index mm0v market model adjusted ex-date return using CRSP value-weighted index r30 30 day holding period return r60 60 day holding period return r90 90 day holding period return r125 125 day holding period return r250 250 day holding period return

134 Table 30 Correlation Matrix: The Returns of Reverse Stock Splits and Financial Distress Variables The table presents the correlations between ex-date returns, 250-day BHRs and various financial distress variables.

Summary Information Variable N Mean Std Dev r0 885 -0.06909 0.16193 r250 891 -0.09464 1.35869 dta 860 0.1942 0.25952 opincta 739 -0.05694 0.41406 invtsls 789 1.00226 6.31133 lgsales 826 1.83512 2.20676 lgassets 869 3.61029 1.86927 nwcta 788 0.08858 0.97682 cacl 788 2.66551 6.91255 qacl 771 2.17809 5.54878 ebita 645 -0.10826 0.61856 reta 772 -35.6595 615.6494 ltdta 781 1.00218 2.79656 mveltd 648 32.95378 207.4896 fata 850 0.27389 0.26425

Number of Observations

r0 r250 dta opincta invtsls lgsales lgassets nwcta cacl qacl ebita reta ltdta mveltd fata r0 885 883 848 728 778 815 857 777 777 760 637 762 770 637 838 r250 883 891 853 732 783 820 862 782 782 765 640 767 776 642 844 dta 848 853 860 733 781 805 860 781 781 764 642 766 781 648 843 opincta 728 732 733 739 678 700 739 679 679 663 570 668 673 560 733 invtsls 778 783 781 678 789 789 789 720 720 720 593 708 715 607 774 lgsales 815 820 805 700 789 826 814 736 736 720 612 722 731 621 795 lgassets 857 862 860 739 789 814 869 788 788 771 645 772 781 648 850 nwcta 777 782 781 679 720 736 788 788 788 771 590 770 779 589 784 cacl 777 782 781 679 720 736 788 788 788 771 590 770 779 589 784 qacl 760 765 764 663 720 720 771 771 771 771 575 755 762 580 768 ebita 637 640 642 570 593 612 645 590 590 575 645 580 587 499 628 reta 762 767 766 668 708 722 772 770 770 755 580 772 766 580 769 ltdta 770 776 781 673 715 731 781 779 779 762 587 766 781 590 777 mveltd 637 642 648 560 607 621 648 589 589 580 499 580 590 648 633 fata 838 844 843 733 774 795 850 784 784 768 628 769 777 633 850

135 Table 30 (Continued)

Correlation Matrix r0 r250 dta opincta invtsls lgsales lgassets nwcta cacl qacl ebita reta ltdta mveltd fata r0 1.0000 0.0111 0.0293 0.0616 -0.0174 0.1233 0.1358 0.0388 -0.0307 -0.0279 0.1021 0.0014 0.0586 0.0641 0.0101 p-value 0.7415 0.3938 0.0970 0.6284 0.0004 <.0001 0.2799 0.3930 0.4417 0.0100 0.9690 0.1040 0.1062 0.7694 r250 1.0000 -0.0780 0.0704 -0.0075 0.0189 0.0197 0.0661 0.0504 0.0430 0.0796 -0.1038 -0.0632 -0.0284 -0.0244 p-value 0.0227 0.0570 0.8334 0.5892 0.5636 0.0647 0.1592 0.2354 0.0440 0.0040 0.0785 0.4726 0.4789 dta 1.0000 0.0048 -0.0244 0.2781 0.3026 -0.0691 -0.0823 -0.0992 -0.0261 0.0291 0.4001 -0.1418 0.2095 p-value 0.8965 0.4961 <.0001 <.0001 0.0537 0.0215 0.0060 0.5089 0.4217 <.0001 0.0003 <.0001 opincta 1.0000 -0.0734 0.3177 0.2880 0.8691 0.0218 0.0183 0.9508 0.0149 0.0341 0.0060 -0.0881 p-value 0.0560 <.0001 <.0001 <.0001 0.5709 0.6384 <.0001 0.7002 0.3767 0.8882 0.0171 invtsls 1.0000 -0.1810 -0.0482 0.0831 0.0347 0.0367 -0.0132 0.0062 -0.0324 -0.0189 -0.0675 p-value <.0001 0.1765 0.0258 0.3527 0.3249 0.7487 0.8701 0.3870 0.6421 0.0605 lgsales 1.0000 0.8035 -0.0466 -0.0994 -0.1504 0.1865 0.1459 0.0051 -0.0471 0.0277 p-value <.0001 0.2069 0.0069 <.0001 <.0001 <.0001 0.8905 0.2408 0.4349 lgassets 1.0000 0.1645 -0.0673 -0.0655 0.2565 0.0443 0.0966 -0.0859 0.0603 p-value <.0001 0.0591 0.0693 <.0001 0.2185 0.0069 0.0288 0.0789 nwcta 1.0000 0.1541 0.1619 0.6932 0.0473 -0.2046 0.0764 -0.2412 p-value <.0001 <.0001 <.0001 0.1901 <.0001 0.0640 <.0001 cacl 1.0000 0.9461 0.0370 0.0198 -0.0868 0.0120 -0.1597 p-value <.0001 0.3699 0.5839 0.0154 0.7722 <.0001 The numbers shown here represent the Pearson's Correlation Coefficient which is qacl usually signified by r (rho). It can take on the values from -1.0 to 1.0. Where -1.0 is a 1.0000 0.0286 0.0201 -0.0877 0.0291 -0.1696 p-value perfect negative (inverse) correlation, 0.0 is no correlation, and 1.0 is a perfect 0.4936 0.5817 0.0155 0.4843 <.0001 positive correlation. ebita 1.0000 0.1350 0.0280 0.0050 -0.0348 p-value The value r0 represents the raw ex-date return and the value r250 is raw buy and hold 0.0011 0.4980 0.9113 0.3840 return 250 trading days after the ex-date. The other values are defined in Table 29. reta 1.0000 -0.0275 0.0129 -0.1021 p-value 0.4478 0.7570 0.0046 A low p-value for this test (less than 0.10 for example) means that there is a statistically significant ltdta relationship between the two variables. The p-value is in italics below the correlation coefficient. 1.0000 -0.0659 0.3640 p-value 0.1097 <.0001 mveltd 1.0000 -0.0386 p-value 0.3328 fata 1.0000 136 Table 31 Regression Analysis of Financial Distress Variables – Single Independent Variable This table shows the results of regressing raw ex-date returns on various accounting variables. Each column represents a regression with the dependent variable being the column heading and the accounting variables in the far left column being the independent variables. The regression coefficient of each independent variable is followed by its p-value in italics.

Multi ple Variable Regression Results - Full Model Fitted Ex-Date Returns Long Run Buy and Hold Returns Independent r0 ma0e ma0v mm0e mm0v r30 r60 r90 r125 r250 Variables p-value p-value p-value p-value p-value p-value p-value p-value p-value p-value dta 0.0148 0.0197 0.0197 0.0073 0.0163 (0.1680) (0.0938) (0.1366) (0.1414) (0.1269) 0.7254 0.6396 0.6388 0.8663 0.6985 0.3972 0.5563 0.4849 0.5147 0.6713 opincta 0.0172 0.0230 0.0203 0.0517 0.0229 (0.1583) 0.0632 0.1945 0.1810 0.6566 0.8593 0.8121 0.8335 0.6035 0.8125 0.7287 0.8632 0.6656 0.7170 0.3401 invtsls 0.0002 0.0002 0.0001 0.0003 0.0003 (0.0002) (0.0105) (0.0131) (0.0097) (0.0056) 0.9059 0.9423 0.9473 0.8814 0.8830 0.9838 0.1898 0.1810 0.3694 0.7096 lgsales 0.0226 0.0210 0.0206 0.0213 0.0216 0.0460 (0.0342) (0.0474) 0.0076 0.0292 0.0192 0.0291 0.0325 0.0325 0.0249 0.3090 0.3470 0.2878 0.8772 0.6681 lgassets (0.0158) (0.0145) (0.0139) (0.0136) (0.0143) (0.0552) 0.0499 0.0529 (0.0008) (0.0408) 0.1478 0.1833 0.2033 0.2296 0.1922 0.2826 0.2274 0.2967 0.9887 0.5980 nwcta 0.0490 0.0473 0.0470 (0.0067) 0.0405 (0.8785) 0.1732 0.4139 0.4623 0.3988 0.2911 0.3071 0.3099 0.8884 0.3807 0.0001 0.3242 0.0553 0.0538 0.2263 cacl (0.0112) (0.0108) (0.0104) (0.0028) (0.0087) 0.0605 (0.1013) (0.1421) (0.1669) (0.1373) 0.5228 0.5395 0.5512 0.8752 0.6197 0.4651 0.1282 0.0822 0.0659 0.2714 qacl 0.0208 0.0201 0.0197 0.0168 0.0189 0.0264 0.0900 0.0964 0.1273 0.0896 0.2351 0.2514 0.2601 0.3516 0.2798 0.7488 0.1755 0.2366 0.1594 0.4715 ebita 0.0130 0.0116 0.0124 0.0091 0.0073 0.4780 0.0151 (0.1563) (0.1575) (0.4219) 0.8350 0.8527 0.8425 0.8866 0.9061 0.1045 0.9490 0.5896 0.6244 0.3411 reta (0.0002) (0.0002) (0.0002) (0.0001) (0.0001) 0.0033 0.0038 0.0036 0.0030 0.0032 0.7984 0.7691 0.7945 0.8396 0.8586 0.2911 0.1360 0.2384 0.3823 0.4910 ltdta 0.0040 0.0037 0.0037 0.0031 0.0040 (0.0098) (0.0062) (0.0117) (0.0081) (0.0090) 0.3252 0.3666 0.3654 0.4658 0.3288 0.6105 0.6872 0.5394 0.7014 0.7567 mveltd 0.0000 0.0000 0.0000 0.0000 0.0000 (0.0002) (0.0002) (0.0002) (0.0003) (0.0002) 0.7223 0.6951 0.6871 0.7256 0.7036 0.2487 0.2456 0.2582 0.2428 0.4302 fata 0.0450 0.0411 0.0416 0.0347 0.0427 0.0489 0.1784 0.3141 0.2975 0.3580 0.2415 0.2839 0.2783 0.3802 0.2655 0.7867 0.2200 0.0786 0.1332 0.1894

137

Table 32 Multivariate Regression Reduced Model This table shows the results of regressing ex-date returns on various accounting variables. It also shows the results of regressing long run buy and hold returns on various accounting variables. Each column represents a regression with the dependent variable being the column heading and the accounting variables in the far left column being the independent variables. Column heading definitions are given in Table 28. The variable bhar205ma is the 250-day buy and hold market adjusted return and bhar250mm is the 250-day buy and hold abnormal return using the market model. Each regression coefficient estimate has its p-value below in italics.

Multiple Variable Regression Results - Reduced Model Ex-Date Returns Long Run Buy and Hold Returns Independent r0 ma0e ma0v mm0e mm0v r125 r250 bhar250ma bhar250mm Variables p-value p-value p-value p-value p-value p-value p-value p-value p-value dta 0.00981 0.01258 0.01223 0.00721 0.01365 -0.11072 -0.38094 -0.54777 -0.54563 0.7268 0.6524 0.6613 0.8013 0.6286 0.4615 0.1127 0.0325 0.035 opincta 0.08169 0.08146 0.08002 0.07818 0.07589 0.29247 0.45576 0.36658 0.36330 0.0114 0.0112 0.0127 0.0169 0.0186 0.0914 0.0993 0.2013 0.2067 invtsls 0.00533 0.00512 0.00515 0.00601 0.00555 0.01078 0.00373 0.01671 0.01678 0.1003 0.1123 0.1102 0.0673 0.0861 0.5341 0.8928 0.5704 0.5699 lgsales 0.00007 0.00010 0.00012 0.00012 0.00011 -0.00194 -0.00057 -0.00199 -0.00194 0.9401 0.918 0.8985 0.898 0.9039 0.7035 0.9445 0.8122 0.817

138 Table 33 NASDAQ Requirements September 2001 The listing requirements are different for small firms and larger firms. The National Market is for larger firms. The charts show initial and continued listing requirements. To be listed on a market a firm must meet the initial listing requirements. To stay listed on a market, the firm must meet the continued listing requirements.

NASDAQ Small Cap Market Requirements

Requirements Initial Listing Continued Listing Net Tangible Assets $4 million $2 million Or or Market Capitalization $50 million $35 million Or or Net Income (in latest fiscal year or 2 of last 3 fiscal years) $750,000 $500,000 Public Float (shares) 1 million 500,000 Market Value of Public Float $5 million $1 million Minimum Bid Price $4 $1 Market Makers 3 2 Shareholders (round lot holders) 300 300 Operating History 1 year N/A or Market Capitalization $50 million Corporate Governance Yes Yes

NASDAQ National Market Requirements

Requirements Initial Listing Initial Listing Initial Listing Continued Continued Listing Listing Plan 1 2 3 1 & 2 3 Net Tangible Assets $6 million $18 million N/A $4 million N/A Market Capitalization N/A N/A $75 million N/A $50 million or or Total Assets $75 million $50 million and and Total Revenue $75 million $50 million Pretax Income (in latest fiscal $1 million N/A N/A N/A N/A year or 2 of last 3 fiscal years) Public Float (shares) 1.1 million 1.1 million 1.1 million 750,000 1.1 million Operating History N/A 2 years N/A N/A N/A Market Value of $8 million $18 million $20 million $5 million $15 million Public Float Minimum Bid Price $5 $5 $5 $1 $5 Shareholders (round lot 400 400 400 400 400 holders) Market Makers 3 3 4 2 4 Corporate Governance Yes Yes Yes Yes Yes

139 Table 34 Low-priced NASDAQ Stocks and Reverse Stock Splits 1998 - 2005 This table shows how many stocks are trading for less than $1.00 each month from 1998 to 2005. The monthly figure shown is the average daily low-priced firms. The number of low priced firms per day is totaled for the month and divided by 30. The number of reverse stock splits per month and the reverse stock splits as a percentage of low-priced firms is shown. Rolling Averages Year/Month Low Priced Reverse Percent Reverse 6 month 12 month 18 month 24 month Stocks Stock Splits Stock Splits rolling rolling rolling rolling average average average average

199801 184 13 7.1% 199802 166 19 11.4% 199803 176 14 7.9% 199804 165 10 6.1% 199805 144 33 23.0% 199806 174 21 12.1% 11.3% 199807 186 10 5.4% 11.0% 199808 226 5 2.2% 9.4% 199809 300 11 3.7% 8.7% 199810 341 12 3.5% 8.3% 199811 243 16 6.6% 5.6% 199812 255 18 7.1% 4.7% 8.0% 199901 166 16 9.6% 5.4% 8.2% 199902 140 13 9.3% 6.6% 8.0% 199903 153 14 9.2% 7.5% 8.1% 199904 126 8 6.4% 8.0% 8.2% 199905 92 8 8.7% 8.4% 7.0% 199906 92 11 11.9% 9.2% 7.0% 8.4% 199907 79 9 11.5% 9.5% 7.5% 8.6% 199908 93 12 12.9% 10.1% 8.4% 8.7% 199909 88 4 4.5% 9.3% 8.4% 8.5% 199910 95 5 5.2% 9.1% 8.6% 8.5% 199911 90 6 6.7% 8.8% 8.6% 7.6% 199912 82 8 9.8% 8.4% 8.8% 7.4% 8.4% 200001 44 7 15.8% 9.2% 9.3% 8.0% 8.8% 200002 16 3 18.8% 10.1% 10.1% 8.9% 9.1% 200003 7 3 42.5% 16.5% 12.9% 11.1% 10.5% 200004 32 4 12.6% 17.7% 13.4% 11.6% 10.8% 200005 63 12 19.0% 19.7% 14.3% 12.3% 10.6% 200006 69 5 7.3% 19.3% 13.9% 12.3% 10.4% 200007 76 7 9.2% 18.2% 13.7% 12.3% 10.6% 200008 114 4 3.5% 15.7% 12.9% 12.0% 10.6% 200009 94 3 3.2% 9.1% 12.8% 11.6% 10.6% 200010 166 6 3.6% 7.6% 12.7% 11.5% 10.6% 200011 194 8 4.1% 5.1% 12.4% 11.2% 10.5% 200012 308 11 3.6% 4.5% 11.9% 10.8% 10.4% 200101 243 4 1.6% 3.3% 10.7% 10.2% 10.0% 200102 187 6 3.2% 3.2% 9.4% 9.7% 9.8% 200103 286 8 2.8% 3.2% 6.1% 9.6% 9.5% 200104 298 10 3.4% 3.1% 5.4% 9.5% 9.4% 200105 238 15 6.3% 3.5% 4.3% 9.5% 9.3% 200106 215 20 9.3% 4.4% 4.5% 9.4% 9.2% 200107 231 17 7.4% 5.4% 4.3% 9.0% 9.0% 200108 273 15 5.5% 5.8% 4.5% 8.2% 8.7% 200109 232 10 4.3% 6.0% 4.6% 6.1% 8.7% 200110 369 11 3.0% 6.0% 4.5% 5.6% 8.6%

140

Table 34 (Continued) Year/Month Low Priced Reverse Percent Reverse 6 month 12 month 18 month 24 month Stocks Stock Splits Stock Splits rolling rolling rolling rolling average average average average

200111 299 1 0.3% 5.0% 4.2% 4.5% 8.3% 200112 261 7 2.7% 3.9% 4.2% 4.3% 8.0% 200201 240 4 1.7% 2.9% 4.2% 3.9% 7.4% 200202 230 4 1.7% 2.3% 4.0% 3.8% 6.7% 200203 238 6 2.5% 2.0% 4.0% 3.7% 5.1% 200204 265 5 1.9% 1.8% 3.9% 3.6% 4.6% 200205 279 12 4.3% 2.5% 3.7% 3.6% 4.0% 200206 286 8 2.8% 2.5% 3.2% 3.6% 3.8% 200207 375 21 5.6% 3.1% 3.0% 3.8% 3.7% 200208 381 11 2.9% 3.3% 2.8% 3.8% 3.7% 200209 351 13 3.7% 3.5% 2.8% 3.8% 3.7% 200210 435 19 4.4% 3.9% 2.9% 3.9% 3.7% 200211 311 13 4.2% 3.9% 3.2% 3.8% 3.7% 200212 312 9 2.9% 3.9% 3.2% 3.4% 3.7% 200301 286 36 12.6% 5.1% 4.1% 3.7% 4.1% 200302 263 16 6.1% 5.6% 4.5% 3.8% 4.3% 200303 288 8 2.8% 5.5% 4.5% 3.7% 4.3% 200304 255 18 7.1% 5.9% 4.9% 3.9% 4.4% 200305 184 18 9.8% 6.9% 5.4% 4.4% 4.6% 200306 121 36 29.8% 11.3% 7.6% 5.9% 5.4% 200307 101 5 4.9% 10.1% 7.6% 6.1% 5.3% 200308 82 20 24.4% 13.1% 9.4% 7.4% 6.1% 200309 58 8 13.8% 14.9% 10.2% 8.0% 6.5% 200310 54 7 13.1% 15.9% 10.9% 8.6% 6.9% 200311 36 4 11.2% 16.2% 11.5% 9.0% 7.4% 200312 43 7 16.3% 13.9% 12.6% 9.7% 7.9% 200401 25 6 23.7% 17.1% 13.6% 10.7% 8.8% 200402 21 3 14.6% 15.4% 14.3% 11.4% 9.4% 200403 26 4 15.4% 15.7% 15.3% 12.0% 9.9% 200404 25 2 7.9% 14.8% 15.4% 12.2% 10.2% 200405 38 5 13.1% 15.2% 15.7% 12.7% 10.5% 200406 43 2 4.6% 13.2% 13.6% 12.8% 10.6% 200407 54 4 7.4% 10.5% 13.8% 12.5% 10.7% 200408 71 0 0.0% 8.0% 11.7% 12.2% 10.6% 200409 62 4 6.5% 6.6% 11.1% 12.4% 10.7% 200410 68 5 7.4% 6.5% 10.7% 12.4% 10.8% 200411 61 1 1.6% 4.6% 9.9% 12.0% 10.7% 200412 44 7 16.0% 6.5% 9.8% 11.2% 11.2% 200501 40 2 4.9% 6.1% 8.3% 11.2% 10.9% 200502 42 4 9.5% 7.7% 7.9% 10.4% 11.1% 200503 59 4 6.8% 7.7% 7.1% 10.0% 11.2% 200504 72 2 2.8% 6.9% 6.7% 9.4% 11.1% 200505 80 8 10.0% 8.4% 6.5% 9.4% 11.1% 200506 76 4 5.2% 6.6% 6.5% 8.7% 10.0% 200507 66 2 3.0% 6.2% 6.2% 7.6% 10.0% 200508 71 5 7.1% 5.8% 6.7% 7.2% 9.2% 200509 61 4 6.6% 5.8% 6.8% 6.7% 8.9% 200510 62 7 11.2% 7.2% 7.1% 6.9% 8.9% 200511 62 6 9.7% 7.2% 7.8% 6.7% 8.8% 200512 57 9 15.9% 8.9% 7.7% 7.3% 8.8%

141

Table 35 Delistings Around the NASDAQ 2001 Moratorium Delistings for various periods around the NASDAQ moratorium on the minimum bid price September 2001. Column A is the actual number of delistings due to low price for the period. Column B is the firms that should have been delisted under rules prior to September 2001. Under the old rules, a firm was issued a warning after 30 days of non-compliance and then delisted 90 days later if it did not meet continued listing standards. Month t=0 is September 2001.

(A) (B) (C) Firms that Should Have Been Percentage Period Months begin end Delistings Delisted (A)/(B) 36 months prior to rule change* t-42 to t-7 199803 200102 283 352 80% 36 months after rule change t+1 to t+36 200110 200409 133 709 19%

Six month increments t-42 to t-37 199803 199808 20 98 20% t-36 to t-31 199809 199902 76 215 35% t-30 to t-25 199903 199908 61 182 34% t-24 to t-19 199909 200002 39 127 31% t-18 to t-13 200003 200008 11 102 11% t-12 to t-7 200009 200102 76 119 64% t-6 to t-1 200103 200108 123 259 47% t+1 to t+6 200110 200203 3 371 1% t+7 to t+12 200204 200209 60 416 14% t+13 to t+18 200210 200303 48 513 9% t+19 to t+24 200304 200309 22 423 5% t+25 to t+30 200310 200403 0 333 0% t+31 to t+36 200404 200409 0 301 0% t+37 to t+42 200410 200503 2 299 1%

pre-moratorium and post-mortorium average delistings compared two sample t-test of equal means T= 2035 (means not equal)

142

APPENDIX B: FIGURES

143 200

180

160

140

120 Number 100 of Splits 80

60

40

20

0 1963 1966 1969 1972 1975 1978 1981 1984 1987 1990 1993 1996 1999 2002

Year

Figure 1 Reverse Stock Splits 1962-2002 Source: Data from the Center for Research in Security Prices

144 450 400 350 300 Number of 250 reverse splits 200 150 100 50 0 1 for 2 or less 1 for 2.01 to3.49 1 for 3.5 to 4.99 1 for 5 1 for 5.01 to 9.99 1 for 10 1 for 10.0 to 24.99 1 for 25 to 34.99 1 for 35 to 49.99 1 for 50 or more

Split ratio

8000 7000 6000 5000 Number of 4000 forward splits 3000 2000 1000 0 Less than 1.25 for 1 1.25 for 1 1.251 to 1.5 for 1 1.5 for 1 B22 1.51 to 1.99 for 1 2 for 1 2.01 to 2.99 for 1 3 for 1 3.01 to 4.99 for 1 5 or more for 1

Split ratio

Figure 2 Split Ratios of Stock Splits

145 17. 5

15. 0

12. 5

10. 0 P e r c e n t 7. 5

5. 0

2. 5

0 - 1 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 bhr

Figure 3 Distribution of 250-Day Buy and Hold Returns

146

Buy and Hold Returns

15.00%

10.00%

5.00%

0.00% 0 20 40 60 80 100 120 140 160 180 200 220 240 -5.00%

BHR -10.00%

-15.00%

-20.00%

-25.00%

-30.00% Days after Ex-date

Legend: top to BHR missing = daily average bottom on day 250 BHR mis s ing = 0 Figure 4 The Average Buy and Hold Return of a Reverse Stock Split After the Ex-date The average buy and hold return for days after the ex-date is plotted. Missing returns are replaced with the average daily return or with zero.

147 Buy and Hold Abnormal Returns

15.00%

10.00%

5.00%

0.00% 0 20 40 60 80 100 120 140 160 180 200 220 240 -5.00%

BHAR -10.00%

-15.00%

-20.00%

-25.00%

-30.00% Days after Ex-date

Legend: top to 1. VW BHAR missing = daily average 2. VW BHAR missing = VW index bottom on day 250 3. EW BHAR missing = daily average 4. EW BHAR missing = EW index Figure 5 Market Adjusted Returns of Reverse Stock Splits The market adjusted average buy and hold abnormal return for days after the ex-date is plotted. Market is either the CRSP value weighted index (VW) or the CRSP equal weighted index (EW). Missing returns are replaced with the average daily return, with zero, or with a market index.

148 Matched Firm Buy and Hold Abnormal Returns

15.00%

10.00%

5.00%

0.00% 10 30 50 70 90 110 130 150 170 190 210 230 250 -5.00%

BHAR -10.00%

-15.00%

-20.00%

-25.00%

-30.00% Days After Ex-date

Matched Firm BHARs

Figure 6 The Average Matched Firm BHAR of a Reverse Stock Split After the Ex-date The matched firm adjusted average buy and hold abnormal return for days after the ex-date is plotted. The abnormal return is calculated by subtracting the return of an out of sample firm with similar size, price and industry.

149 APPENDIX C: FAMA AND FRENCH INDUSTRY CLASSIFICATIONS

This Appendix contains the Standard Industry Classification (SIC) codes that are used in determining the industry classification of a firm. These industry classifications were found on Kenneth French’s website. Some researchers use the first two digits of the SIC code to determine which industries have similar characteristics. This type of classification is an attempt to group like industries in a more thoughtful way. In the first panel, all industries are divided into five broad categories and in the last panel all industries are divided into 48 more narrow categories.

5 INDUSTRIES 1 Manuf Manufacturing 4 Money Finance 2000-3999 6000-6999 2 Utils Utilities 5 Other Agric, Mines, Oil, Construct, Transport, Telecom 4900-4999 Health and Legal Services 3 Shops Wholesale, Retail, and Some Services 5000-5999 7000-7999

10 INDUSTRIES

1 NoDur Consumer NonDurables 0100-0999 5 Manuf Manufacturing 2000-2399 2440-2499 2700-2749 2520-2589 2770-2799 2600-2699 3100-3199 2750-2769 3940-3989 3200-3629 3660-3709 2 Durbl Consumer Durables 3712-3713 2400-2439 3715-3715 2500-2519 3717-3749 2590-2599 3752-3791 3000-3099 3793-3909 3630-3659 3710-3711 6 Telcm Telephones and Television 3714-3714 4800-4899 3716-3716 3750-3751 7 Utils Utilities 3792-3792 4900-4949 3910-3939 3990-3999 8 Shops Wholesale, Retail, and Some Services 5000-5999 3 Oil Oil, Gas, and Coal Extraction and Products 7000-7999 1200-1399 2900-2999 9 Money Finance 6000-6999 4 Chems Chemicals and Allied Products 2800-2899 10 Other Everything Else

150

12 INDUSTRIES

1 NoDur Consumer NonDurables -- Food, Tobacco, Textiles, Apparel, Leather, Toys 5 Chems Chemicals and Allied Products

0100-0999 2800-2829 2000-2399 2840-2899 2700-2749 2770-2799 3100-3199 6 BusEq Business Equipment -- Computers, Software, 3940-3989 and Electronic Equipment

3570-3579 2 Durbl Consumer Durables -- Cars, TV's, 3660-3692 Furniture, Household Appliances 3694-3699 3810-3829 2500-2519 7370-7379 2590-2599 3630-3659 3710-3711 7 Telcm Telephone and Television Transmission 3714-3714 3716-3716 4800-4899 3750-3751 3792-3792 3900-3939 8 Utils Utilities 3990-3999 4900-4949

3 Manuf Manufacturing -- Machinery, Trucks, Planes, Off Furn, Paper, Com Printing 9 Shops Wholesale, Retail, and Some Services (Laundries, Repair Shops) 2520-2589 2600-2699 5000-5999 2750-2769 7200-7299 3000-3099 7600-7699 3200-3569 3580-3629 10 Hlth Healthcare, Medical Equipment, and Drugs 3700-3709 3712-3713 2830-2839 3715-3715 3693-3693 3717-3749 3840-3859 3752-3791 8000-8099 3793-3799 3830-3839 3860-3899 11 Money Finance

6000-6999 4 Enrgy Oil, Gas, and Coal Extraction and Products

1200-1399 12 Other Everything Else -- Mines, Constr, BldMt, Trans, 2900-2999 Hotels, Bus Serv, Entertainment

151

17 INDUSTRIES 1 Food Food 2299-2299 Misc textile products 0100-0199 Agric production - crops 2300-2390 Apparel and other finished products 0200-0299 Agric production - livestock 2391-2392 Curtains, home furnishings 0700-0799 Agricultural services 2393-2395 Textile bags, canvas products 0900-0999 Fishing, hunting & trapping 2396-2396 Auto trim 2000-2009 Food and kindred products 2397-2399 Misc textile products 2010-2019 Meat products 3020-3021 Rubber and plastics footwear 2020-2029 Dairy products 3100-3111 Leather tanning and finishing 2030-2039 Canned-preserved fruits-vegs 3130-3131 Boot, shoe cut stock, findings 2040-2046 Flour and other grain mill products 3140-3149 Footwear except rubber 2047-2047 Dog and cat food 3150-3151 Leather gloves and mittens 2048-2048 Prepared feeds for animals 3963-3965 Fasteners, buttons, needles, pins 2050-2059 Bakery products 5130-5139 Wholesale - apparel 2060-2063 Sugar and confectionery products 2064-2068 Candy and other confectionery 5 Durbl Consumer Durables 2070-2079 Fats and oils 2510-2519 Household furniture 2080-2080 Beverages 2590-2599 Misc furniture and fixtures 2082-2082 Malt beverages 3060-3069 Fabricated rubber products 2083-2083 Malt 3070-3079 Misc rubber products 2084-2084 Wine 3080-3089 Misc plastic products 2085-2085 Distilled and blended liquors 3090-3099 Misc rubber and plastic products 2086-2086 Bottled-canned soft drinks 3630-3639 Household appliances 2087-2087 Flavoring syrup 3650-3651 Household audio visual equip 2090-2092 Misc food preps 3652-3652 Phonographic records 2095-2095 Roasted coffee 3860-3861 Photographic equip (Kodak etc, but also 2096-2096 Potato chips Xerox) 2097-2097 Manufactured ice 3870-3873 Watches clocks and parts 2098-2099 Misc food preparations 3910-3911 Jewelry-precious metals 5140-5149 Wholesale - groceries & related prods 3914-3914 Silverware 5150-5159 Wholesale - farm products 3915-3915 Jewelers' findings, materials 5180-5182 Wholesale - beer, wine 3930-3931 Musical instruments 5191-5191 Wholesale - farm supplies 3940-3949 Toys 3960-3962 Costume jewelry and notions 2 Mines Mining and Minerals 5020-5023 Wholesale - furniture and home furnishings 1000-1009 Metal mining 5064-5064 Wholesale - electrical appliance TV and radio 1010-1019 Iron ores 5094-5094 Wholesale - jewelry and watches 1020-1029 Copper ores 5099-5099 Wholesale - durable goods 1030-1039 Lead and zinc ores 1040-1049 Gold & silver ores 6 Chems Chemicals 1060-1069 Ferroalloy ores 2800-2809 Chemicals and allied products 1080-1089 Mining services 2810-2819 Industrial inorganical chems 1090-1099 Misc metal ores 2820-2829 Plastic material & synthetic resin 1200-1299 Bituminous coal 2860-2869 Industrial organic chems 1400-1499 Mining and quarrying non-metallic minerals 2870-2879 Agriculture chemicals 5050-5052 Wholesale - metals and minerals 2890-2899 Misc chemical products 5160-5169 Wholesale - chemicals & allied prods 3 Oil Oil and Petroleum Products 1300-1300 Oil and gas extraction 7 Cnsum Drugs, Soap, Prfums, Tobacco 1310-1319 Crude petroleum & natural gas 2100-2199 Tobacco products 1320-1329 Natural gas liquids 2830-2830 Drugs 1380-1380 Oil and gas field services 2831-2831 Biological products 1381-1381 Drilling oil & gas wells 2833-2833 Medicinal chemicals 1382-1382 Oil-gas field exploration 2834-2834 Pharmaceutical preparations 1389-1389 Oil and gas field services 2840-2843 Soap & other detergents 2900-2912 Petroleum refining 2844-2844 Perfumes cosmetics 5170-5172 Wholesale - petroleum and petro prods 5120-5122 Wholesale - drugs & propietary 5194-5194 Wholesale - tobacco and tobacco products 4 Clths Textiles, Apparel & Footwear 2200-2269 Textile mill products 8 Cnstr Construction and Construction Materials 2270-2279 Floor covering mills 0800-0899 Forestry 2280-2284 Yarn and thread mills 1500-1511 Build construction - general contractors 2290-2295 Misc textile goods 1520-1529 Gen building contractors - residential 2296-2296 Tire cord and fabric 1530-1539 Operative builders 2297-2297 Nonwoven fabrics 1540-1549 Gen building contractors - non-residential 2298-2298 Cordage and twine

152

17 INDUSTRIES (continued)

1600-1699 Heavy Construction - not building contractors 3580-3580 Refrig & service ind machines 1700-1799 Construction - special contractors 3581-3581 Automatic vending machines 2400-2439 Lumber and wood products 3582-3582 Commercial laundry and dry cleaning 2440-2449 Wood containers machines 2450-2459 Wood buildings-mobile homes 3585-3585 Air conditioning, heating, refrid eq 2490-2499 Misc wood products 3586-3586 Measuring and dispensing pumps 2850-2859 Paints 3589-3589 Service industry machinery 2950-2952 Paving & roofing materials 3590-3599 Misc industrial and commercial equipment and 3200-3200 Stone, clay, glass, concrete etc mach 3210-3211 Flat glass 3600-3600 Elec mach eq & supply 3240-3241 Cement hydraulic 3610-3613 Elec transmission 3250-3259 Structural clay prods 3620-3621 Electrical industrial appar 3261-3261 Vitreous china plumbing fixtures 3622-3622 Industrial controls 3264-3264 Porcelain electrical supply 3623-3629 Electrical industrial appar 3270-3275 Concrete gypsum & plaster 3670-3679 Electronic components 3280-3281 Cut stone and stone products 3680-3680 Computers 3290-3293 Abrasive and asbestos products 3681-3681 Computers - mini 3420-3429 Hand tools and hardware 3682-3682 Computers - mainframe 3430-3433 Heating equip & plumbing fix 3683-3683 Computers - terminals 3440-3441 Fabricated struct metal products 3684-3684 Computers - disk & tape drives 3442-3442 Metal doors, frames 3685-3685 Computers - optical scanners 3446-3446 Architectural or ornamental metal work 3686-3686 Computers - graphics 3448-3448 Pre-fab metal buildings 3687-3687 Computers - office automation systems 3449-3449 Misc structural metal work 3688-3688 Computers - peripherals 3450-3451 Screw machine products 3689-3689 Computers - equipment 3452-3452 Bolts, nuts screws 3690-3690 Miscellaneous electrical machinery and equip 5030-5039 Wholesale - lumber and construction materials 3691-3692 Storage batteries 5070-5078 Wholesale - hardware, plumbing, heating equip 3693-3693 X-ray, electromedical app 5198-5198 Wholesale - Paints, varnishes, and supplies 3694-3694 Elec eq, internal combustion engines 5210-5211 Retail - lumber & other building mat 3695-3695 Magnetic and optical recording media 5230-5231 Retail - paint, glass, wallpaper 3699-3699 Electrical machinery and equip 5250-5251 Retail - hardward stores 3810-3810 Search, detection, navigation, guidance 3811-3811 Engr lab and research equipment 9 Steel Steel Works Etc 3812-3812 Search, detection, navigation, guidance 3300-3300 Primary metal industries 3820-3820 Measuring and controlling equipment 3310-3317 Blast furnaces & steel works 3821-3821 Lab apparatus and furniture 3320-3325 Iron & steel foundries 3822-3822 Automatic controls - Envir and applic 3330-3339 Prim smelt-refin nonfer metals 3823-3823 Industrial measurement instru 3340-3341 Secondary smelt-refin nonfer metals 3824-3824 Totalizing fluid meters 3350-3357 Rolling & drawing nonferous metals 3825-3825 Elec meas & test instr 3360-3369 Non-ferrous foundries and casting 3826-3826 Lab analytical instruments 3390-3399 Misc primary metal products 3827-3827 Optical instr and lenses 3829-3829 Meas and control devices 10 FabPr Fabricated Products 3830-3839 Optical instr and lenses 3410-3412 Metal cans and shipping containers 3950-3955 Pens pencils and office supplies 3443-3443 Fabricated plate work 5060-5060 Wholesale - electrical goods 3444-3444 Sheet metal work 5063-5063 Wholesale - electrical apparatus and 3460-3469 Metal forgings and stampings equipment 3470-3479 Coating and engraving 5065-5065 Wholesale - electronic parts 3480-3489 Ordnance & accessories 5080-5080 Wholesale - machinery and equipment 3490-3499 Misc fabricated metal products 5081-5081 Wholesale - machinery and equipment (?)

11 Machn Machinery and Business Equipment 12 Cars Automobiles 3510-3519 Engines & turbines 3710-3710 Motor vehicles and motor vehicle equip 3520-3529 Farm and garden machinery 3711-3711 Motor vehicles & car bodies 3530-3530 Constr, mining material handling machinery 3714-3714 Motor vehicle parts 3531-3531 Construction machinery 3716-3716 Motor homes 3532-3532 Mining machinery, except oil field 3750-3751 Motorcycles, bicycles and parts (Harley & 3533-3533 Oil field machinery Huffy) 3534-3534 Elevators 3792-3792 Travel trailers and campers 3535-3535 Conveyors 5010-5015 Wholesale - autos and parts 3536-3536 Cranes, hoists 5510-5521 Retail - auto dealers 3540-3549 Metalworking machinery 5530-5531 Retail - auto and home supply stores 3550-3559 Special industry machinery 5560-5561 Retail - recreational vehicle dealers 3560-3569 General industrial machinery 3570-3579 Office computers

153

17 INDUSTRIES (continued)

5570-5571 Retail - motorcycle dealers 5430-5431 Retail - fruite and vegetable markets 5590-5599 Retail - automotive dealers 5440-5441 Retail - candy, nut, confectionary stores 5450-5451 Retail - dairy product stores 13 Trans Transportation 5460-5461 Retail - bakeries 3713-3713 Truck & bus bodies 5490-5499 Retail - miscellaneous food stores 3715-3715 Truck trailers 5540-5541 Retail - gasoline service stations 3720-3720 Aircraft & parts 5550-5551 Retail - boat dealers 3721-3721 Aircraft 5600-5699 Retail - apparel & acces 3724-3724 Aircraft engines, engine parts 5700-5700 Retail - home furniture and equipment stores 3725-3725 Aircraft parts 5710-5719 Retail - home furnishings stores 3728-3728 Aircraft parts 5720-5722 Retail - household appliance stores 3730-3731 Ship building and repair 5730-5733 Retail - radio, TV and consumer electronic stores 3732-3732 Boat building and repair 5734-5734 Retail - computer and computer software stores 3740-3743 Railroad Equipment 5735-5735 Retail - record and tape stores 3760-3769 Guided missiles and space vehicles 5736-5736 Retail - musical instrument stores 3790-3790 Misc trans equip 5750-5750 Retail - (?) 3795-3795 Tanks and tank components 5800-5813 Retail - eating places 3799-3799 Misc trans equip 5890-5890 Eating and drinking places 4000-4013 Railroads-line haul 5900-5900 Retail - misc 4100-4100 Transit and passenger trans 5910-5912 Retail - drug & proprietary stores 4110-4119 Local passenger trans 5920-5921 Retail - liquor stores 4120-4121 Taxicabs 5930-5932 Retail - used merchandise stores 4130-4131 Intercity bus trans (Greyhound) 5940-5940 Retail - misc 4140-4142 Bus charter 5941-5941 Retail - sporting goods stores, bike shops 4150-4151 School buses 5942-5942 Retail - book stores 4170-4173 Motor vehicle terminals, service facilities 5943-5943 Retail - stationery stores 4190-4199 Misc transit and passenger transportation 5944-5944 Retail - jewelry stores 4200-4200 Motor freight trans, warehousing 5945-5945 Retail - hobby, toy and game shops 4210-4219 Trucking 5946-5946 Retail - camera and photo shop 4220-4229 Warehousing and storage 5947-5947 Retail - gift, novelty 4230-4231 Terminal facilities - motor freight 5948-5948 Retail - luggage 4400-4499 Water transport 5949-5949 Retail - sewing & needlework stores 4500-4599 Air transportation 5960-5963 Retail - non-store retailers (catalogs, etc) 4600-4699 Pipelines, except natural gas 5980-5989 Retail - fuel & ice stores (Penn Central Co) 4700-4700 Transportation services 5990-5990 Retail - retail stores 4710-4712 Freight forwarding 5992-5992 Retail - florists 4720-4729 Travel agencies, etc 5993-5993 Retail - tobacco stores 4730-4739 Arrange trans - freight and cargo 5994-5994 Retail - news dealers 4740-4742 Rental of railroad cars 5995-5995 Retail - computer stores 4780-4780 Misc services incidental to trans 5999-5999 Retail stores 4783-4783 Packing and crating 4785-4785 Motor vehicle inspection 4789-4789 Transportation services 16 Finan Banks, Insurance Companies, and Other Financials 6010-6019 Federal reserve banks 14 Utils Utilities 6020-6020 Commercial banks 4900-4900 Electric, gas, sanitary services 6021-6021 National commercial banks 4910-4911 Electric services 6022-6022 State banks - Fed Res System 4920-4922 Natural gas transmission 6023-6023 State banks - not Fed Res System 4923-4923 Natural gas transmission-distr 6025-6025 National banks - Fed Res System 4924-4925 Natural gas distribution 6026-6026 National banks - not Fed Res System 4930-4931 Electric and other services combined 6028-6029 Banks 4932-4932 Gas and other services combined 6030-6036 Savings institutions 4939-4939 Combination utilities 6040-6049 Trust companies, nondeposit 4940-4942 Water supply 6050-6059 Functions closely related to banking 6060-6062 Credit unions 15 Rtail Retail Stores 6080-6082 Foreign banks 5260-5261 Retail - nurseries, lawn, garden stores 6090-6099 Functions related to deposit banking 5270-5271 Retail - mobile home dealers 6100-6100 Nondepository credit institutions 5300-5300 Retail - general merchandise stores 6110-6111 Federal credit agencies 5310-5311 Retail - department stores 6112-6112 FNMA 5320-5320 Retail - general merchandise stores (?) 6120-6129 S&Ls 5330-5331 Retail - variety stores 6140-6149 Personal credit institutions (Beneficial) 5334-5334 Retail - catalog showroom 6150-6159 Business credit institutions 5390-5399 Retail - Misc general merchandise stores 6160-6163 Mortgage bankers 5400-5400 Retail - food stores 5410-5411 Retail - grocery stores 5412-5412 Retail - convenience stores 5420-5421 Retail - meat, fish mkt

154

17 INDUSTRIES (continued)

6172-6172 Finance lessors 3640-3644 Electric lighting, wiring 6199-6199 Financial services 3645-3645 Residential lighting fixtures 6200-6299 Security and commodity brokers 3646-3646 Commercial lighting 6300-6300 Insurance 3647-3647 Vehicular lighting 6310-6312 Life insurance 3648-3649 Lighting equipment 6320-6324 Accident and health insurance 3660-3660 Communication equip 6330-6331 Fire, marine, property-casualty ins 3661-3661 Telephone and telegraph apparatus 6350-6351 Surety insurance 3662-3662 Communications equipment 6360-6361 Title insurance 3663-3663 Radio TV comm equip & apparatus 6370-6371 Pension, health, welfare funds 3664-3664 Search, navigation, guidance systems 6390-6399 Insurance carriers 3665-3665 Training equipment & simulators 6400-6411 Insurance agents 3666-3666 Alarm & signaling products 6500-6500 Real estate 3669-3669 Communication equipment 6510-6510 Real estate operators 3840-3849 Surg & med instru 6512-6512 Operators - non-resident buildings 3850-3851 Ophthalmic goods 6513-6513 Operators - apartment buildings 3991-3991 Brooms and brushes 6514-6514 Operators - other than apartment 3993-3993 Signs, advertising specialty 6515-6515 Operators - residential mobile home 3995-3995 Burial caskets 6517-6519 Lessors of real property 3996-3996 Hard surface floor cover 6530-6531 Real estate agents and managers 4810-4813 Telephone communications 6532-6532 Real estate dealers 4820-4822 Telegraph and other message communication 6540-6541 Title abstract offices 4830-4839 Radio-TV Broadcasters 6550-6553 Real estate developers 4840-4841 Cable and other pay TV services 6611-6611 Combined real estate, insurance, etc 4890-4890 Communication services (Comsat) 6700-6700 Holding, other investment offices 4891-4891 Cable TV operators 6710-6719 Holding offices 4892-4892 Telephone interconnect 6720-6722 Investment offices 4899-4899 Communication services 6723-6723 Management investment, closed-end 4950-4959 Sanitary services 6730-6733 Trusts 4960-4961 Steam, air conditioning supplies 6790-6790 Miscellaneous investing 4970-4971 Irrigation systems 6792-6792 Oil royalty traders 4991-4991 Cogeneration - SM power producer 6794-6794 Patent owners & lessors 5040-5042 Wholesale - professional and commercial 6795-6795 Mineral royalty traders equipment and supplies 6798-6798 REIT 5043-5043 Wholesale - photographic equipment 6799-6799 Investors, NEC 5044-5044 Wholesale - office equipment 5045-5045 Wholesale - computers 17 Other Everything Else 5046-5046 Wholesale - commercial equip 2520-2549 Office furniture and fixtures 5047-5047 Wholesale - medical, dental equip 2600-2639 Paper and allied products 5048-5048 Wholesale - ophthalmic goods 2640-2659 Paperboard containers, boxes, drums, tubs 5049-5049 Wholesale - professional equip and supplies 2661-2661 Building paper and board mills 5082-5082 Wholesale - construction and mining equipment 2670-2699 Paper and allied products 5083-5083 Wholesale - farm and garden machinery 2700-2709 Printing publishing and allied 5084-5084 Wholesale - industrial machinery and equipment 2710-2719 Newspapers: publishing-printing 5085-5085 Wholesale - industrial supplies 2720-2729 Periodicals: publishing-printing 5086-5087 Wholesale - machinery and equipment (?) 2730-2739 Books: publishing-printing 5088-5088 Wholesale - trans eq except motor vehicles 2740-2749 Misc publishing 5090-5090 Wholesale - misc durable goods 2750-2759 Commercial printing 5091-5092 Wholesale - sporting goods, toys 2760-2761 Manifold business forms 5093-5093 Wholesale - scrap and waste materials 2770-2771 Greeting card publishing 5100-5100 Wholesale - nondurable goods 2780-2789 Book binding 5110-5113 Wholesale - paper and paper products 2790-2799 Service industries for print trade 5199-5199 Wholesale - non-durable goods 2835-2835 In vitro, in vivo diagnostics 7000-7000 Hotels, other lodging places 2836-2836 Biological products, except diagnostics 7010-7011 Hotels motels 2990-2999 Misc petroleum products 7020-7021 Rooming and boarding houses 3000-3000 Rubber & misc plastic products 7030-7033 Camps and recreational vehicle parks 3010-3011 Tires and inner tubes 7040-7041 Membership hotels and lodging 3041-3041 Rubber & plastics hose and belting 3050-3053 Gaskets, hoses, etc 3160-3161 Luggage 3170-3171 Handbags and purses 3172-3172 Personal leather goods, except handbags 3190-3199 Leather goods 3220-3221 Glass containers 3229-3229 Pressed and blown glass 3230-3231 Glass products 3260-3260 Pottery and related products 3262-3263 China and earthenware table articles 3269-3269 Pottery products 3295-3299 Non-metalic mineral products 3537-3537 Trucks, tractors, trailers 155

17 INDUSTRIES (continued

7200-7200 Services - personal 7397-7397 Services - commercial testing labs 7210-7212 Services - laundry, cleaners 7399-7399 Services - business services 7213-7213 Services - linen 7500-7500 Services - auto repair, services 7215-7216 Services - coin-op cleaners, dry cleaners 7510-7519 Services - truck, auto, trailer rental and leasing 7217-7217 Services - carpet, upholstery cleaning 7520-7523 Services - automobile parking 7218-7218 Services - industrial launderers 7530-7539 Services - auto repair shops 7219-7219 Services - laundry, cleaners 7540-7549 Services - auto services, except repair (car 7220-7221 Services - photo studios, portrait washes) 7230-7231 Services - beauty shops 7600-7600 Services - Misc repair services 7240-7241 Services - barber shops 7620-7620 Services - Electrical repair shops 7250-7251 Services - shoe repair 7622-7622 Services - Radio and TV repair shops 7260-7269 Services - funeral 7623-7623 Services - Refridg and air conditioner repair 7290-7290 Services - misc 7629-7629 Services - Electrical repair shops 7291-7291 Services - tax return 7630-7631 Services - Watch, clock and jewelry repair 7299-7299 Services - misc 7640-7641 Services - Reupholster, furniture repair 7300-7300 Services - business services 7690-7699 Services - Misc repair shops 7310-7319 Services - advertising 7800-7829 Services - motion picture production and 7320-7323 Services - credit reporting agencies, collection distribution services 7830-7833 Services - motion picture theatres 7330-7338 Services - mailing, reproduction, commercial art 7840-7841 Services - video rental 7340-7342 Services - services to dwellings, other buildings 7900-7900 Services - amusement and recreation 7349-7349 Services - cleaning and builging maint 7910-7911 Services - dance studios 7350-7351 Services - misc equip rental and leasing 7920-7929 Services - bands, entertainers 7352-7352 Services - medical equip rental 7930-7933 Services - bowling centers 7353-7353 Services - heavy construction equip rental 7940-7949 Services - professional sports 7359-7359 Services - equip rental and leasing 7980-7980 Amusement and recreation services 7360-7369 Services - personnel supply services 7990-7999 Services - misc entertainment 7370-7372 Services - computer programming and data 8000-8099 Services - health processing 8100-8199 Services - legal 7373-7373 Computer integrated systems design 8200-8299 Services - educational 7374-7374 Services - computer processing, data prep 8300-8399 Services - social services 7375-7375 Services - information retrieval services 8400-8499 Services - museums, galleries, botanic gardens 7376-7376 Services - computer facilities management service 8600-8699 Services - membership organizations 7377-7377 Services - computer rental and leasing 8700-8700 Services - engineering, accounting, research, 7378-7378 Services - computer maintenance and repair management 7379-7379 Services - computer related services 8710-8713 Services - engineering, accounting, surveying 7380-7380 Services - misc business services 8720-8721 Services - accounting, auditing, bookkeeping 7381-7382 Services - security 8730-8734 Services - research, development, testing labs 7383-7383 Services - news syndicates 8740-8748 Services - management, public relations, 7384-7384 Services - photofinishing labs consulting 7385-7385 Services - telephone interconnections 8800-8899 Services - private households 7389-7390 Services - misc business services 8900-8910 Services - misc 7391-7391 Services - R&D labs 8911-8911 Services - engineering & architect 7392-7392 Services - management consulting & P.R. 8920-8999 Services - misc 7393-7393 Services - detective and protective (ADT) 7394-7394 Services - equipment rental & leasing 7395-7395 Services - photofinishing labs (School pictures)

156 30 INDUSTRIES

1 Food Food Products 2840-2843 Soap & other detergents 0100-0199 Agric production - crops 2844-2844 Perfumes cosmetics 0200-0299 Agric production - livestock 3160-3161 Luggage 0700-0799 Agricultural services 3170-3171 Handbags and purses 0910-0919 Commercial fishing 3172-3172 Personal leather goods, except handbags 2000-2009 Food and kindred products 3190-3199 Leather goods 2010-2019 Meat products 3229-3229 Pressed and blown glass 2020-2029 Dairy products 3260-3260 Pottery and related products 2030-2039 Canned-preserved fruits-vegs 3262-3263 China and earthenware table articles 2040-2046 Flour and other grain mill products 3269-3269 Pottery products 2048-2048 Prepared feeds for animals 3230-3231 Glass products 2050-2059 Bakery products 3630-3639 Household appliances 2060-2063 Sugar and confectionery products 3750-3751 Motorcycles, bicycles and parts (Harley & Huffy) 2064-2068 Candy and other confectionery 3800-3800 Misc inst, photo goods, watches 2070-2079 Fats and oils 3860-3861 Photographic equip (Kodak etc, but also Xerox) 2086-2086 Bottled-canned soft drinks 3870-3873 Watches clocks and parts 2087-2087 Flavoring syrup 3910-3911 Jewelry-precious metals 2090-2092 Misc food preps 3914-3914 Silverware 2095-2095 Roasted coffee 3915-3915 Jewelers' findings, materials 2096-2096 Potato chips 3960-3962 Costume jewelry and notions 2097-2097 Manufactured ice 3991-3991 Brooms and brushes 2098-2099 Misc food preparations 3995-3995 Burial caskets

2 Beer Beer & Liquor 7 Clths Apparel 2080-2080 Beverages 2300-2390 Apparel and other finished products 2082-2082 Malt beverages 3020-3021 Rubber and plastics footwear 2083-2083 Malt 3100-3111 Leather tanning and finishing 2084-2084 Wine 3130-3131 Boot, shoe cut stock, findings 2085-2085 Distilled and blended liquors 3140-3149 Footwear except rubber 3150-3151 Leather gloves and mittens 3 Smoke Tobacco Products 3963-3965 Fasteners, buttons, needles, pins 2100-2199 Tobacco products 8 Hlth Healthcare, Medical Equipment, Pharmaceutical Products 4 Games Recreation 2830-2830 Drugs 0920-0999 Fishing, hunting & trapping 2831-2831 Biological products 3650-3651 Household audio visual equip 2833-2833 Medicinal chemicals 3652-3652 Phonographic records 2834-2834 Pharmaceutical preparations 3732-3732 Boat building and repair 2835-2835 In vitro, in vivo diagnostics 3930-3931 Musical instruments 2836-2836 Biological products, except diagnostics 3940-3949 Toys 3693-3693 X-ray, electromedical app 7800-7829 Services - motion picture production and distribution 3840-3849 Surg & med instru 7830-7833 Services - motion picture theatres 3850-3851 Ophthalmic goods 7840-7841 Services - video rental 8000-8099 Services - health 7900-7900 Services - amusement and recreation 7910-7911 Services - dance studios 9 Chems Chemicals 7920-7929 Services - bands, entertainers 2800-2809 Chemicals and allied products 7930-7933 Services - bowling centers 2810-2819 Industrial inorganical chems 7940-7949 Services - professional sports 2820-2829 Plastic material & synthetic resin 7980-7980 Amusement and recreation services (?) 2850-2859 Paints 7990-7999 Services - misc entertainment 2860-2869 Industrial organic chems 2870-2879 Agriculture chemicals 5 Books Printing and Publishing 2890-2899 Misc chemical products 2700-2709 Printing publishing and allied 2710-2719 Newspapers: publishing-printing 10 Txtls Textiles 2720-2729 Periodicals: publishing-printing 2200-2269 Textile mill products 2730-2739 Books: publishing-printing 2270-2279 Floor covering mills 2740-2749 Misc publishing 2280-2284 Yarn and thread mills 2750-2759 Commercial printing 2290-2295 Misc textile goods 2770-2771 Greeting card publishing 2297-2297 Nonwoven fabrics 2780-2789 Book binding 2298-2298 Cordage and twine 2790-2799 Service industries for print trade 2299-2299 Misc textile products 3993-3993 Signs, advertising specialty

6 Hshld Consumer Goods 2047-2047 Dog and cat food 2391-2392 Curtains, home furnishings 2510-2519 Household furniture 2590-2599 Misc furniture and fixtures

157

30 INDUSTRIES (continued) 3580-3580 Refrig & service ind machines 3581-3581 Automatic vending machines 3582-3582 Commercial laundry and dry cleaning machines 2393-2395 Textile bags, canvas products 3585-3585 Air conditioning, heating, refrid eq 2397-2399 Misc textile products 3586-3586 Measuring and dispensing pumps

3589-3589 Service industry machinery 11 Cnstr Construction and Construction Materials 3590-3599 Misc industrial and commercial equipment and mach 0800-0899 Forestry

1500-1511 Build construction - general contractors 14 ElcEq Electrical Equipment 1520-1529 Gen building contractors - residential 3600-3600 Elec mach eq & supply 1530-1539 Operative builders 3610-3613 Elec transmission 1540-1549 Gen building contractors - non-residential 3620-3621 Electrical industrial appar 1600-1699 Heavy Construction - not building contractors 3623-3629 Electrical industrial appar 1700-1799 Construction - special contractors 3640-3644 Electric lighting, wiring 2400-2439 Lumber and wood products 3645-3645 Residential lighting fixtures 2450-2459 Wood buildings-mobile homes 3646-3646 Commercial lighting 2490-2499 Misc wood products 3648-3649 Lighting equipment 2660-2661 Building paper and board mills 3660-3660 Communication equip 2950-2952 Paving & roofing materials 3690-3690 Miscellaneous electrical machinery and equip 3200-3200 Stone, clay, glass, concrete etc 3691-3692 Storage batteries 3210-3211 Flat glass 3699-3699 Electrical machinery and equip 3240-3241 Cement hydraulic

3250-3259 Structural clay prods 15 Autos Automobiles and Trucks 3261-3261 Vitreous china plumbing fixtures 2296-2296 Tire cord and fabric 3264-3264 Porcelain electrical supply 2396-2396 Auto trim 3270-3275 Concrete gypsum & plaster 3010-3011 Tires and inner tubes 3280-3281 Cut stone and stone products 3537-3537 Trucks, tractors, trailers 3290-3293 Abrasive and asbestos products 3647-3647 Vehicular lighting 3295-3299 Non-metallic mineral products 3694-3694 Elec eq, internal combustion engines 3420-3429 Hand tools and hardware 3700-3700 Transportation equipment 3430-3433 Heating equip & plumbing fix 3710-3710 Motor vehicles and motor vehicle equip 3440-3441 Fabricated struct metal products 3711-3711 Motor vehicles & car bodies 3442-3442 Metal doors, frames 3713-3713 Truck & bus bodies 3446-3446 Architectural or ornamental metal work 3714-3714 Motor vehicle parts 3448-3448 Pre-fab metal buildings 3715-3715 Truck trailers 3449-3449 Misc structural metal work 3716-3716 Motor homes 3450-3451 Screw machine products 3792-3792 Travel trailers and campers 3452-3452 Bolts, nuts screws 3790-3791 Misc trans equip 3490-3499 Misc fabricated metal products 3799-3799 Misc trans equip 3996-3996 Hard surface floor cover

16 Carry Aircraft, ships, and railroad equipment 12 Steel Steel Works Etc 3720-3720 Aircraft & parts 3300-3300 Primary metal industries 3721-3721 Aircraft 3310-3317 Blast furnaces & steel works 3723-3724 Aircraft engines, engine parts 3320-3325 Iron & steel foundries 3725-3725 Aircraft parts 3330-3339 Prim smelt-refin nonfer metals 3728-3729 Aircraft parts 3340-3341 Secondary smelt-refin nonfer metals 3730-3731 Ship building and repair 3350-3357 Rolling & drawing nonferous metals 3740-3743 Railroad Equipment 3360-3369 Non-ferrous foundries and casting

3370-3379 Steel works etc 17 Mines Precious Metals, Non-Metallic, and Industrial Metal Mining 3390-3399 Misc primary metal products 1000-1009 Metal mining

1010-1019 Iron ores 13 FabPr Fabricated Products and Machinery 1020-1029 Copper ores 3400-3400 Fabricated metal, except machinery and trans eq 1030-1039 Lead and zinc ores 3443-3443 Fabricated plate work 1040-1049 Gold & silver ores 3444-3444 Sheet metal work 1050-1059 Bauxite and other aluminum ores 3460-3469 Metal forgings and stampings 1060-1069 Ferroalloy ores 3470-3479 Coating and engraving 1070-1079 Mining 3510-3519 Engines & turbines 1080-1089 Mining services 3520-3529 Farm and garden machinery 1090-1099 Misc metal ores 3530-3530 Constr, mining material handling machinery 1100-1119 Anthracite mining 3531-3531 Construction machinery 3532-3532 Mining machinery, except oil field 3533-3533 Oil field machinery 3534-3534 Elevators 3535-3535 Conveyors 3536-3536 Cranes, hoists 3538-3538 Machinery 3540-3549 Metalworking machinery 3550-3559 Special industry machinery 3560-3569 General industrial machinery

158

30 INDUSTRIES (continued) 7353-7353 Services - heavy construction equip rental 7359-7359 Services - equip rental and leasing 1400-1499 Mining and quarrying non-metallic minerals 7360-7369 Services - personnel supply services 7370-7372 Services - computer programming and data processing 18 Coal Coal 7374-7374 Services - computer processing, data prep 1200-1299 Bituminous coal 7375-7375 Services - information retrieval services 7376-7376 Services - computer facilities management service 19 Oil Petroleum and Natural Gas 7377-7377 Services - computer rental and leasing 1300-1300 Oil and gas extraction 7378-7378 Services - computer maintenance and repair 1310-1319 Crude petroleum & natural gas 7379-7379 Services - computer related services 1320-1329 Natural gas liquids 7380-7380 Services - misc business services 1330-1339 Petroleum and natural gas 7381-7382 Services - security 1370-1379 Petroleum and natural gas 7383-7383 Services - news syndicates 1380-1380 Oil and gas field services 7384-7384 Services - photofinishing labs 1381-1381 Drilling oil & gas wells 7385-7385 Services - telephone interconnections 1382-1382 Oil-gas field exploration 7389-7390 Services - misc business services 1389-1389 Oil and gas field services 7391-7391 Services - R&D labs 2900-2912 Petroleum refining 7392-7392 Services - management consulting & P.R. 2990-2999 Misc petroleum products 7393-7393 Services - detective and protective (ADT) 7394-7394 Services - equipment rental & leasing 20 Util Utilities 7395-7395 Services - photofinishing labs (School pictures) 4900-4900 Electric, gas, sanitary services 7396-7396 Services - trading stamp services 4910-4911 Electric services 7397-7397 Services - commercial testing labs 4920-4922 Natural gas transmission 7399-7399 Services - business services 4923-4923 Natural gas transmission-distr 7500-7500 Services - auto repair, services 4924-4925 Natural gas distribution 7510-7519 Services - truck, auto, trailer rental and leasing 4930-4931 Electric and other services combined 7520-7529 Services - automobile parking 4932-4932 Gas and other services combined 7530-7539 Services - auto repair shops 4939-4939 Combination utilities 7540-7549 Services - auto services, except repair (car washes) 4940-4942 Water supply 7600-7600 Services - Misc repair services 7620-7620 Services - Electrical repair shops 21 Telcm Communication 7622-7622 Services - Radio and TV repair shops 4800-4800 Communications 7623-7623 Services - Refridg and air conditioner repair 4810-4813 Telephone communications 7629-7629 Services - Electrical repair shops 4820-4822 Telegraph and other message communication 7630-7631 Services - Watch, clock and jewelry repair 4830-4839 Radio-TV Broadcasters 7640-7641 Services - Reupholster, furniture repair 4840-4841 Cable and other pay TV services 7690-7699 Services - Misc repair shops 4880-4889 Communications 8100-8199 Services - legal 4890-4890 Communication services (Comsat) 8200-8299 Services - educational 4891-4891 Cable TV operators 8300-8399 Services - social services 4892-4892 Telephone interconnect 8400-8499 Services - museums, galleries, botanic gardens 4899-4899 Communication services 8600-8699 Services - membership organizations 8700-8700 Services - engineering, accounting, research, 22 Servs Personal and Business Services management 7020-7021 Rooming and boarding houses 8710-8713 Services - engineering, accounting, surveying 7030-7033 Camps and recreational vehicle parks 8720-8721 Services - accounting, auditing, bookkeeping 7200-7200 Services - personal 8730-8734 Services - research, development, testing labs 7210-7212 Services - laundry, cleaners 7214-7214 Services - diaper service 7215-7216 Services - coin-op cleaners, dry cleaners 7217-7217 Services - carpet, upholstery cleaning 7218-7218 Services - industrial launderers 7219-7219 Services - laundry, cleaners 7220-7221 Services - photo studios, portrait 7230-7231 Services - beauty shops 7240-7241 Services - barber shops 7250-7251 Services - shoe repair 7260-7269 Services - funeral 7270-7290 Services - misc 7291-7291 Services - tax return 7292-7299 Services - misc 7300-7300 Services - business services 7310-7319 Services - advertising 7320-7329 Services - credit reporting agencies, collection services 7330-7339 Services - mailing, reproduction, commercial art 7340-7342 Services - services to dwellings, other buildings 7349-7349 Services - cleaning and building maint 7350-7351 Services - misc equip rental and leasing 7352-7352 Services - medical equip rental

159

30 INDUSTRIES (continued) 4220-4229 Warehousing and storage 4230-4231 Terminal facilities - motor freight 4240-4249 Transportation 8740-8748 Services - management, public relations, consulting 4400-4499 Water transport 8800-8899 Services - private households 4500-4599 Air transportation 8900-8910 Services - misc 4600-4699 Pipelines, except natural gas 8911-8911 Services - engineering & architect 4700-4700 Transportation services 8920-8999 Services - misc 4710-4712 Freight forwarding

4720-4729 Travel agencies, etc 23 BusEq Business Equipment 4730-4739 Arrange trans - freight and cargo 3570-3579 Office computers 4740-4749 Rental of railroad cars 3622-3622 Industrial controls 4780-4780 Misc services incidental to trans 3661-3661 Telephone and telegraph apparatus 4782-4782 Inspection and weighing services 3662-3662 Communications equipment 4783-4783 Packing and crating 3663-3663 Radio TV comm equip & apparatus 4784-4784 Fixed facilities for vehicles, not elsewhere classified 3664-3664 Search, navigation, guidance systems 4785-4785 Motor vehicle inspection 3665-3665 Training equipment & simulators 4789-4789 Transportation services 3666-3666 Alarm & signaling products

3669-3669 Communication equipment 26 Whlsl Wholesale 3670-3679 Electronic components 5000-5000 Wholesale - durable goods 3680-3680 Computers 5010-5015 Wholesale - autos and parts 3681-3681 Computers - mini 5020-5023 Wholesale - furniture and home furnishings 3682-3682 Computers - mainframe 5030-5039 Wholesale - lumber and construction materials 3683-3683 Computers - terminals 5040-5042 Wholesale - professional and commercial equipment and 3684-3684 Computers - disk & tape drives supplies 3685-3685 Computers - optical scanners 5043-5043 Wholesale - photographic equipment 3686-3686 Computers - graphics 5044-5044 Wholesale - office equipment 3687-3687 Computers - office automation systems 5045-5045 Wholesale - computers 3688-3688 Computers - peripherals 5046-5046 Wholesale - commercial equip 3689-3689 Computers - equipment 5047-5047 Wholesale - medical, dental equip 3695-3695 Magnetic and optical recording media 5048-5048 Wholesale - ophthalmic goods 3810-3810 Search, detection, navigation, guidance 5049-5049 Wholesale - professional equip and supplies 3811-3811 Engr lab and research equipment 5050-5059 Wholesale - metals and minerals 3812-3812 Search, detection, navigation, guidance 5060-5060 Wholesale - electrical goods 3820-3820 Measuring and controlling equipment 5063-5063 Wholesale - electrical apparatus and equipment 3821-3821 Lab apparatus and furniture 5064-5064 Wholesale - electrical appliance TV and radio 3822-3822 Automatic controls - Envir and applic 5065-5065 Wholesale - electronic parts 3823-3823 Industrial measurement instru 5070-5078 Wholesale - hardware, plumbing, heating equip 3824-3824 Totalizing fluid meters 5080-5080 Wholesale - machinery and equipment 3825-3825 Elec meas & test instr 5081-5081 Wholesale - machinery and equipment (?) 3826-3826 Lab analytical instruments 5082-5082 Wholesale - construction and mining equipment 3827-3827 Optical instr and lenses 5083-5083 Wholesale - farm and garden machinery 3829-3829 Meas and control devices 5084-5084 Wholesale - industrial machinery and equipment 3830-3839 Optical instr and lenses 5085-5085 Wholesale - industrial supplies 7373-7373 Computer integrated systems design 5086-5087 Wholesale - machinery and equipment (?)

5088-5088 Wholesale - trans eq except motor vehicles 24 Paper Business Supplies and Shipping Containers 5090-5090 Wholesale - misc durable goods 2440-2449 Wood containers 5091-5092 Wholesale - sporting goods, toys 2520-2549 Office furniture and fixtures 5093-5093 Wholesale - scrap and waste materials 2600-2639 Paper and allied products 5094-5094 Wholesale - jewelry and watches 2640-2659 Paperboard containers, boxes, drums, tubs 5099-5099 Wholesale - durable goods 2670-2699 Paper and allied products 5100-5100 Wholesale - nondurable goods 2760-2761 Manifold business forms 5110-5113 Wholesale - paper and paper products 3220-3221 Glass containers 5120-5122 Wholesale - drugs & proprietary 3410-3412 Metal cans and shipping containers 3950-3955 Pens pencils and office supplies

25 Trans Transportation 4000-4013 Railroads-line haul 4040-4049 Railway express service 4100-4100 Transit and passenger trans 4110-4119 Local passenger trans 4120-4121 Taxicabs 4130-4131 Intercity bus trans (Greyhound) 4140-4142 Bus charter 4150-4151 School buses 4170-4173 Motor vehicle terminals, service facilities 4190-4199 Misc transit and passenger transportation 4200-4200 Motor freight trans, warehousing 4210-4219 Trucking

160

30 INDUSTRIES (continued) 5950-5959 Retail 5960-5969 Retail - non-store retailers (catalogs, etc) 5970-5979 Retail 5130-5139 Wholesale - apparel 5980-5989 Retail - fuel & ice stores (Penn Central Co) 5140-5149 Wholesale - groceries & related prods 5990-5990 Retail - retail stores 5150-5159 Wholesale - farm products 5992-5992 Retail - florists 5160-5169 Wholesale - chemicals & allied prods 5993-5993 Retail - tobacco stores 5170-5172 Wholesale - petroleum and petro prods 5994-5994 Retail - news dealers 5180-5182 Wholesale - beer, wine 5995-5995 Retail - computer stores 5190-5199 Wholesale - non-durable goods 5999-5999 Retail stores

27 Rtail Retail 28 Meals Restaurants, Hotels, Motels 5200-5200 Retail - bldg material, hardware, garden 5800-5819 Retail - eating places 5210-5219 Retail - lumber & other building mat 5820-5829 Restaurants, hotels, motels 5220-5229 Retail 5890-5899 Eating and drinking places 5230-5231 Retail - paint, glass, wallpaper 7000-7000 Hotels, other lodging places 5250-5251 Retail - hardware stores 7010-7019 Hotels motels 5260-5261 Retail - nurseries, lawn, garden stores 7040-7049 Membership hotels and lodging 5270-5271 Retail - mobile home dealers 7213-7213 Services - linen 5300-5300 Retail - general merchandise stores

5310-5311 Retail - department stores 29 Fin Banking, Insurance, Real Estate, Trading 5320-5320 Retail - general merchandise stores (?) 6000-6000 Depository institutions 5330-5331 Retail - variety stores 6010-6019 Federal reserve banks 5334-5334 Retail - catalog showroom 6020-6020 Commercial banks 5340-5349 Retail 6021-6021 National commercial banks 5390-5399 Retail - Misc general merchandise stores 6022-6022 State banks - Fed Res System 5400-5400 Retail - food stores 6023-6024 State banks - not Fed Res System 5410-5411 Retail - grocery stores 6025-6025 National banks - Fed Res System 5412-5412 Retail - convenience stores 6026-6026 National banks - not Fed Res System 5420-5429 Retail - meat, fish mkt 6027-6027 National banks, not FDIC 5430-5439 Retail - fruit and vegetable markets 6028-6029 Banks 5440-5449 Retail - candy, nut, confectionary stores 6030-6036 Savings institutions 5450-5459 Retail - dairy product stores 6040-6059 Banks (?) 5460-5469 Retail - bakeries 6060-6062 Credit unions 5490-5499 Retail - miscellaneous food stores 6080-6082 Foreign banks 5500-5500 Retail - auto dealers and gas stations 6090-6099 Functions related to deposit banking 5510-5529 Retail - auto dealers 6100-6100 Nondepository credit institutions 5530-5539 Retail - auto and home supply stores 6110-6111 Federal credit agencies 5540-5549 Retail - gasoline service stations 6112-6113 FNMA 5550-5559 Retail - boat dealers 6120-6129 S&Ls 5560-5569 Retail - recreational vehicle dealers 6130-6139 Agricultural credit institutions 5570-5579 Retail - motorcycle dealers 6140-6149 Personal credit institutions (Beneficial) 5590-5599 Retail - automotive dealers 6150-6159 Business credit institutions 5600-5699 Retail - apparel & acces 6160-6169 Mortgage bankers 5700-5700 Retail - home furniture and equipment stores 6170-6179 Finance lessors 5710-5719 Retail - home furnishings stores 6190-6199 Financial services 5720-5722 Retail - household appliance stores 6200-6299 Security and commodity brokers 5730-5733 Retail - radio, TV and consumer electronic stores 6300-6300 Insurance 5734-5734 Retail - computer and computer software stores 6310-6319 Life insurance 5735-5735 Retail - record and tape stores 6320-6329 Accident and health insurance 5736-5736 Retail - musical instrument stores 6330-6331 Fire, marine, property-casualty ins 5750-5799 Retail 6350-6351 Surety insurance 5900-5900 Retail - misc 6360-6361 Title insurance 5910-5912 Retail - drug & proprietary stores 6370-6379 Pension, health, welfare funds 5920-5929 Retail - liquor stores 6390-6399 Insurance carriers 5930-5932 Retail - used merchandise stores 6400-6411 Insurance agents 5940-5940 Retail - misc 6500-6500 Real estate 5941-5941 Retail - sporting goods stores, bike shops 6510-6510 Real estate operators 5942-5942 Retail - book stores 6512-6512 Operators - non-resident buildings 5943-5943 Retail - stationery stores 6513-6513 Operators - apartment buildings 5944-5944 Retail - jewelry stores 6514-6514 Operators - other than apartment 5945-5945 Retail - hobby, toy and game shops 6515-6515 Operators - residential mobile home 5946-5946 Retail - camera and photo shop 6517-6519 Lessors of real property 5947-5947 Retail - gift, novelty 6520-6529 Real estate 5948-5948 Retail - luggage 6530-6531 Real estate agents and managers 5949-5949 Retail - sewing & needlework stores

161

30 INDUSTRIES (continued) 6790-6791 Miscellaneous investing 6792-6792 Oil royalty traders 6793-6793 Commodity traders 6532-6532 Real estate dealers 6794-6794 Patent owners & lessors 6540-6541 Title abstract offices 6795-6795 Mineral royalty traders 6550-6553 Real estate developers 6798-6798 REIT 6590-6599 Real estate 6799-6799 Investors, NEC 6610-6611 Combined real estate, insurance, etc

6700-6700 Holding, other investment offices 30 Other Everything Else 6710-6719 Holding offices 4950-4959 Sanitary services 6720-6722 Investment offices 4960-4961 Steam, air conditioning supplies 6723-6723 Management investment, closed-end 4970-4971 Irrigation systems 6724-6724 Unit investment trusts 4990-4991 Cogeneration - SM power producer 6725-6725 Face-amount certificate offices

6730-6733 Trusts 6740-6779 Investment offices

38 INDUSTRIES

1 Agric Agriculture, forestry, and fishing 20 MtlPr Fabricated Metal Products 0100-0999 3400-3499 2 Mines Mining 21 Machn Machinery, Except Electrical 1000-1299 3500-3599 3 Oil Oil and Gas Extraction 22 Elctr Electrical and Electronic Equipment 1300-1399 3600-3699 4 Stone Nonmetallic Minerals Except Fuels 23 Cars Transportation Equipment 1400-1499 3700-3799 5 Cnstr Construction 24 Instr Instruments and Related Products 1500-1799 3800-3879 6 Food Food and Kindred Products 25 Manuf Miscellaneous Manufacturing Industries 2000-2099 3900-3999 7 Smoke Tobacco Products 26 Trans Transportation 2100-2199 4000-4799 8 Txtls Textile Mill Products 27 Phone Telephone and Telegraph Communication 2200-2299 4800-4829 9 Apprl Apparel and other Textile Products 28 TV Radio and Television Broadcasting 2300-2399 4830-4899 10 Wood Lumber and Wood Products 29 Utils Electric, Gas, and Water Supply 2400-2499 4900-4949 11 Chair Furniture and Fixtures 30 Garbg Sanitary Services 2500-2599 4950-4959 12 Paper Paper and Allied Products 31 Steam Steam Supply 2600-2661 4960-4969 13 Print Printing and Publishing 32 Water Irrigation Systems 2700-2799 4970-4979 14 Chems Chemicals and Allied Products 33 Whlsl Wholesale 2800-2899 5000-5199 15 Ptrlm Petroleum and Coal Products 34 Rtail Retail Stores 2900-2999 5200-5999 16 Rubbr Rubber and Miscellaneous Plastics Products 35 Money Finance, Insurance, and Real Estate 3000-3099 6000-6999 17 Lethr Leather and Leather Products 36 Srvc Services 3100-3199 7000-8999 18 Glass Stone, Clay and Glass Products 37 Govt Public Administration 3200-3299 9000-9999 19 Metal Primary Metal Industries 38 Other Everything Else 3300-3399

162

48 INDUSTRIES

1 Agric Agriculture 2391-2392 Curtains, home furnishings 0100-0199 Agric production - crops 2510-2519 Household furniture 0200-0299 Agric production - livestock 2590-2599 Misc furniture and fixtures 0700-0799 Agricultural services 2840-2843 Soap & other detergents 0910-0919 Commercial fishing 2844-2844 Perfumes cosmetics 2048-2048 Prepared feeds for animals 3160-3161 Luggage 3170-3171 Handbags and purses 2 Food Food Products 3172-3172 Personal leather goods, except handbags 2000-2009 Food and kindred products 3190-3199 Leather goods 2010-2019 Meat products 3229-3229 Pressed and blown glass 2020-2029 Dairy products 3260-3260 Pottery and related products 2030-2039 Canned-preserved fruits-vegs 3262-3263 China and earthenware table articles 2040-2046 Flour and other grain mill products 3269-3269 Pottery products 2050-2059 Bakery products 3230-3231 Glass products 2060-2063 Sugar and confectionery products 3630-3639 Household appliances 2070-2079 Fats and oils 3750-3751 Motorcycles, bicycles and parts (Harley & Huffy) 2090-2092 Misc food preps 3800-3800 Misc inst, photo goods, watches 2095-2095 Roasted coffee 3860-3861 Photographic equip (Kodak etc, but also Xerox) 2098-2099 Misc food preparations 3870-3873 Watches clocks and parts 3910-3911 Jewelry-precious metals 3 Soda Candy & Soda 3914-3914 Silverware 2064-2068 Candy and other confectionery 3915-3915 Jewelers' findings, materials 2086-2086 Bottled-canned soft drinks 3960-3962 Costume jewelry and notions 2087-2087 Flavoring syrup 3991-3991 Brooms and brushes 2096-2096 Potato chips 3995-3995 Burial caskets 2097-2097 Manufactured ice 10 Clths Apparel 4 Beer Beer & Liquor 2300-2390 Apparel and other finished products 2080-2080 Beverages 3020-3021 Rubber and plastics footwear 2082-2082 Malt beverages 3100-3111 Leather tanning and finishing 2083-2083 Malt 3130-3131 Boot, shoe cut stock, findings 2084-2084 Wine 3140-3149 Footware except rubber 2085-2085 Distilled and blended liquors 3150-3151 Leather gloves and mittens 3963-3965 Fasteners, buttons, needles, pins 5 Smoke Tobacco Products 2100-2199 Tobacco products 11 Hlth Healthcare 8000-8099 Services - health 6 Toys Recreation 0920-0999 Fishing, hunting & trapping 12 MedEq Medical Equipment 3650-3651 Household audio visual equip 3693-3693 X-ray, electromedical app 3652-3652 Phonographic records 3840-3849 Surg & med instru 3732-3732 Boat building and repair 3850-3851 Ophthalmic goods 3930-3931 Musical instruments 3940-3949 Toys 13 Drugs Pharmaceutical Products 2830-2830 Drugs 7 Fun Entertainment 2831-2831 Biological products 7800-7829 Services - motion picture production and 2833-2833 Medicinal chemicals distribution 2834-2834 Pharmaceutical preparations 7830-7833 Services - motion picture theatres 2835-2835 In vitro, in vivo diagnostics 7840-7841 Services - video rental 2836-2836 Biological products, except diagnostics 7900-7900 Services - amusement and recreation 7910-7911 Services - dance studios 14 Chems Chemicals 7920-7929 Services - bands, entertainers 2800-2809 Chemicals and allied products 7930-7933 Services - bowling centers 2810-2819 Industrial inorganical chems 7940-7949 Services - professional sports 2820-2829 Plastic material & synthetic resin 7980-7980 Amusement and recreation services (?) 2850-2859 Paints 7990-7999 Services - misc entertainment 2860-2869 Industrial organic chems 2870-2879 Agriculture chemicals 8 Books Printing and Publishing 2890-2899 Misc chemical products 2700-2709 Printing publishing and allied 2710-2719 Newspapers: publishing-printing 2720-2729 Periodicals: publishing-printing 2730-2739 Books: publishing-printing 2740-2749 Misc publishing 2770-2771 Greeting card publishing 2780-2789 Book binding 2790-2799 Service industries for print trade

9 Hshld Consumer Goods 2047-2047 Dog and cat food

163 48 INDUSTRIES (continued) 3444-3444 Sheet metal work 3460-3469 Metal forgings and stampings 3470-3479 Coating and engraving 15 Rubbr Rubber and Plastic Products

3031-3031 Reclaimed rubber 21 Mach Machinery 3041-3041 Rubber & plastic hose and belting 3510-3519 Engines & turbines 3050-3053 Gaskets, hoses, etc 3520-3529 Farm and garden machinery 3060-3069 Fabricated rubber products 3530-3530 Constr, mining material handling machinery 3070-3079 Misc rubber products (?) 3531-3531 Construction machinery 3080-3089 Misc plastic products 3532-3532 Mining machinery, except oil field 3090-3099 Misc rubber and plastic products (?) 3533-3533 Oil field machinery

3534-3534 Elevators 16 Txtls Textiles 3535-3535 Conveyors 2200-2269 Textile mill products 3536-3536 Cranes, hoists 2270-2279 Floor covering mills 3538-3538 Machinery 2280-2284 Yarn and thread mills 3540-3549 Metalworking machinery 2290-2295 Misc textile goods 3550-3559 Special industry machinery 2297-2297 Nonwoven fabrics 3560-3569 General industrial machinery 2298-2298 Cordage and twine 3580-3580 Refrig & service ind machines 2299-2299 Misc textile products 3581-3581 Automatic vending machines 2393-2395 Textile bags, canvas products 3582-3582 Commercial laundry and dry cleaning machines 2397-2399 Misc textile products 3585-3585 Air conditioning, heating, refrid eq

3586-3586 Measuring and dispensing pumps 17 BldMt Construction Materials 3589-3589 Service industry machinery 0800-0899 Forestry 3590-3599 Misc industrial and commercial equipment and mach 2400-2439 Lumber and wood products

2450-2459 Wood buildings-mobile homes 22 ElcEq Electrical Equipment 2490-2499 Misc wood products 3600-3600 Elec mach eq & supply 2660-2661 Building paper and board mills 3610-3613 Elec transmission 2950-2952 Paving & roofing materials 3620-3621 Electrical industrial appar 3200-3200 Stone, clay, glass, concrete etc 3623-3629 Electrical industrial appar 3210-3211 Flat glass 3640-3644 Electric lighting, wiring 3240-3241 Cement hydraulic 3645-3645 Residential lighting fixtures 3250-3259 Structural clay prods 3646-3646 Commercial lighting 3261-3261 Vitreous china plumbing fixtures 3648-3649 Lighting equipment 3264-3264 Porcelain electrical supply 3660-3660 Communication equip 3270-3275 Concrete gypsum & plaster 3690-3690 Miscellaneous electrical machinery and equip 3280-3281 Cut stone and stone products 3691-3692 Storage batteries 3290-3293 Abrasive and asbestos products 3699-3699 Electrical machinery and equip 3295-3299 Non-metallic mineral products

3420-3429 Hand tools and hardware 23 Autos Automobiles and Trucks 3430-3433 Heating equip & plumbing fix 2296-2296 Tire cord and fabric 3440-3441 Fabricated struct metal products 2396-2396 Auto trim 3442-3442 Metal doors, frames 3010-3011 Tires and inner tubes 3446-3446 Architectural or ornamental metal work 3537-3537 Trucks, tractors, trailers 3448-3448 Pre-fab metal buildings 3647-3647 Vehicular lighting 3449-3449 Misc structural metal work 3694-3694 Elec eq, internal combustion engines 3450-3451 Screw machine products 3700-3700 Transportation equipment 3452-3452 Bolts, nuts screws 3710-3710 Motor vehicles and motor vehicle equip 3490-3499 Misc fabricated metal products 3711-3711 Motor vehicles & car bodies 3996-3996 Hard surface floor cover 3713-3713 Truck & bus bodies

3714-3714 Motor vehicle parts 18 Cnstr Construction 3715-3715 Truck trailers 1500-1511 Build construction - general contractors 3716-3716 Motor homes 1520-1529 Gen building contractors - residential 3792-3792 Travel trailers and campers 1530-1539 Operative builders 3790-3791 Misc trans equip 1540-1549 Gen building contractors - non-residential 3799-3799 Misc trans equip 1600-1699 Heavy Construction - not building

contractors 1700-1799 Construction - special contractors

19 Steel Steel Works Etc 3300-3300 Primary metal industries 3310-3317 Blast furnaces & steel works 3320-3325 Iron & steel foundries 3330-3339 Prim smelt-refin nonfer metals 3340-3341 Secondary smelt-refin nonfer metals 3350-3357 Rolling & drawing nonferous metals 3360-3369 Non-ferrous foundries and casting 3370-3379 Steel works etc 3390-3399 Misc primary metal products

20 FabPr Fabricated Products 3400-3400 Fabricated metal, except machinery and trans eq 3443-3443 Fabricated plate work

164 48 INDUSTRIES (continued) 33 PerSv Personal Services 7020-7021 Rooming and boarding houses 24 Aero Aircraft 7030-7033 Camps and recreational vehicle parks 3720-3720 Aircraft & parts 7200-7200 Services - personal 3721-3721 Aircraft 7210-7212 Services - laundry, cleaners 3723-3724 Aircraft engines, engine parts 7214-7214 Services - diaper service 3725-3725 Aircraft parts 7215-7216 Services - coin-op cleaners, dry cleaners 3728-3729 Aircraft parts 7217-7217 Services - carpet, upholstery cleaning 7219-7219 Services - laundry, cleaners 25 Ships Shipbuilding, Railroad Equipment 7220-7221 Services - photo studios, portrait 3730-3731 Ship building and repair 7230-7231 Services - beauty shops 3740-3743 Railroad Equipment 7240-7241 Services - barber shops 7250-7251 Services - shoe repair 26 Guns Defense 7260-7269 Services - funeral 3760-3769 Guided missiles and space vehicles 7270-7290 Services - misc 3795-3795 Tanks and tank components 7291-7291 Services - tax return 3480-3489 Ordnance & accessories 7292-7299 Services - misc 7395-7395 Services - photofinishing labs (School pictures) 27 Gold Precious Metals 7500-7500 Services - auto repair, services 1040-1049 Gold & silver ores 7520-7529 Services - automobile parking 7530-7539 Services - auto repair shops 28 Mines Non-Metallic and Industrial Metal Mining 7540-7549 Services - auto services, except repair (car washes) 1000-1009 Metal mining 7600-7600 Services - Misc repair services 1010-1019 Iron ores 7620-7620 Services - Electrical repair shops 1020-1029 Copper ores 7622-7622 Services - Radio and TV repair shops 1030-1039 Lead and zinc ores 7623-7623 Services - Refridg and air conditioner repair 1050-1059 Bauxite and other aluminum ores 7629-7629 Services - Electrical repair shops 1060-1069 Ferroalloy ores 7630-7631 Services - Watch, clock and jewelry repair 1070-1079 Mining 7640-7641 Services - Reupholster, furniture repair 1080-1089 Mining services 7690-7699 Services - Misc repair shops 1090-1099 Misc metal ores 8100-8199 Services - legal 1100-1119 Anthracite mining 8200-8299 Services - educational 1400-1499 Mining and quarrying non-metallic minerals 8300-8399 Services - social services 8400-8499 Services - museums, galleries, botanic gardens 29 Coal Coal 8600-8699 Services - membership organizations 1200-1299 Bituminous coal 8800-8899 Services - private households

30 Oil Petroleum and Natural Gas 34 BusSv Business Services 1300-1300 Oil and gas extraction 2750-2759 Commercial printing 1310-1319 Crude petroleum & natural gas 3993-3993 Signs, advertising specialty 1320-1329 Natural gas liquids 7218-7218 Services - industrial launderers 1330-1339 Petroleum and natural gas 7300-7300 Services - business services 1370-1379 Petroleum and natural gas 7310-7319 Services - advertising 1380-1380 Oil and gas field services 7320-7329 Services - credit reporting agencies, collection 1381-1381 Drilling oil & gas wells services 1382-1382 Oil-gas field exploration 7330-7339 Services - mailing, reproduction, commercial art 1389-1389 Oil and gas field services 7340-7342 Services - services to dwellings, other buildings 2900-2912 Petroleum refining 7349-7349 Services - cleaning and building maint 2990-2999 Misc petroleum products 7350-7351 Services - misc equip rental and leasing 7352-7352 Services - medical equip rental 31 Util Utilities 7353-7353 Services - heavy construction equip rental 4900-4900 Electric, gas, sanitary services 7359-7359 Services - equip rental and leasing 4910-4911 Electric services 7360-7369 Services - personnel supply services 4920-4922 Natural gas transmission 7370-7372 Services - computer programming and data 4923-4923 Natural gas transmission-distr processing 4924-4925 Natural gas distribution 4930-4931 Electric and other services combined 4932-4932 Gas and other services combined 4939-4939 Combination utilities 4940-4942 Water supply

32 Telcm Communication 4800-4800 Communications 4810-4813 Telephone communications 4820-4822 Telegraph and other message communication 4830-4839 Radio-TV Broadcasters 4840-4841 Cable and other pay TV services 4880-4889 Communications 4890-4890 Communication services (Comsat) 4891-4891 Cable TV operators 4892-4892 Telephone interconnect 4899-4899 Communication services

165 48 INDUSTRIES (continued) 3826-3826 Lab analytical instruments 3827-3827 Optical instr and lenses 7374-7374 Services - computer processing, data prep 3829-3829 Meas and control devices 7375-7375 Services - information retrieval services 3830-3839 Optical instr and lenses 7376-7376 Services - computer facilities management service 38 Paper Business Supplies 7377-7377 Services - computer rental and leasing 2520-2549 Office furniture and fixtures 7378-7378 Services - computer maintenance and repair 2600-2639 Paper and allied products 7379-7379 Services - computer related services 2670-2699 Paper and allied products 7380-7380 Services - misc business services 2760-2761 Manifold business forms 7381-7382 Services - security 3950-3955 Pens pencils and office supplies 7383-7383 Services - news syndicates 7384-7384 Services - photofinishing labs 39 Boxes Shipping Containers 7385-7385 Services - telephone interconnections 2440-2449 Wood containers 7389-7390 Services - misc business services 2640-2659 Paperboard containers, boxes, drums, tubs 7391-7391 Services - R&D labs 3220-3221 Glass containers 7392-7392 Services - management consulting & P.R. 3410-3412 Metal cans and shipping containers 7393-7393 Services - detective and protective (ADT) 7394-7394 Services - equipment rental & leasing 40 Trans Transportation 7396-7396 Services - trading stamp services 4000-4013 Railroads-line haul 7397-7397 Services - commercial testing labs 4040-4049 Railway express service 7399-7399 Services - business services 4100-4100 Transit and passenger trans 7510-7519 Services - truck, auto, trailer rental and 4110-4119 Local passenger trans leasing 4120-4121 Taxicabs 8700-8700 Services - engineering, accounting, 4130-4131 Intercity bus trans (Greyhound) research, management 4140-4142 Bus charter 8710-8713 Services - engineering, accounting, 4150-4151 School buses surveying 4170-4173 Motor vehicle terminals, service facilities 8720-8721 Services - accounting, auditing, 4190-4199 Misc transit and passenger transportation bookkeeping 4200-4200 Motor freight trans, warehousing 8730-8734 Services - research, development, testing 4210-4219 Trucking labs 4220-4229 Warehousing and storage 8740-8748 Services - management, public relations, 4230-4231 Terminal facilities - motor freight consulting 4240-4249 Transportation 8900-8910 Services - misc 4400-4499 Water transport 8911-8911 Services - engineering & architect 4500-4599 Air transportation 8920-8999 Services - misc 4600-4699 Pipelines, except natural gas 4700-4700 Transportation services 35 Comps Computers 4710-4712 Freight forwarding 3570-3579 Office computers 4720-4729 Travel agencies, etc 3680-3680 Computers 4730-4739 Arrange trans - freight and cargo 3681-3681 Computers - mini 4740-4749 Rental of railroad cars 3682-3682 Computers - mainframe 4780-4780 Misc services incidental to trans 3683-3683 Computers - terminals 4782-4782 Inspection and weighing services 3684-3684 Computers - disk & tape drives 4783-4783 Packing and crating 3685-3685 Computers - optical scanners 4784-4784 Fixed facilities for vehicles, not elsewhere classified 3686-3686 Computers - graphics 4785-4785 Motor vehicle inspection 3687-3687 Computers - office automation systems 4789-4789 Transportation services 3688-3688 Computers - peripherals 3689-3689 Computers - equipment 41 Whlsl Wholesale 3695-3695 Magnetic and optical recording media 5000-5000 Wholesale - durable goods 7373-7373 Computer integrated systems design 5010-5015 Wholesale - autos and parts 5020-5023 Wholesale - furniture and home furnishings 36 Chips Electronic Equipment 5030-5039 Wholesale - lumber and construction materials 3622-3622 Industrial controls 5040-5042 Wholesale - professional and commercial equipment 3661-3661 Telephone and telegraph apparatus and supplies 3662-3662 Communications equipment 3663-3663 Radio TV comm equip & apparatus 3664-3664 Search, navigation, guidance systems 3665-3665 Training equipment & simulators 3666-3666 Alarm & signaling products 3669-3669 Communication equipment 3670-3679 Electronic components 3810-3810 Search, detection, navigation, guidance 3812-3812 Search, detection, navigation, guidance

37 LabEq Measuring and Control Equipment 3811-3811 Engr lab and research equipment 3820-3820 Measuring and controlling equipment 3821-3821 Lab apparatus and furniture 3822-3822 Automatic controls - Envir and applic 3823-3823 Industrial measurement instru 3824-3824 Totalizing fluid meters 3825-3825 Elec meas & test instr

166 48 INDUSTRIES (continued) 5570-5579 Retail - motorcycle dealers 5590-5599 Retail - automotive dealers 5043-5043 Wholesale - photographic equipment 5600-5699 Retail - apparel & acces 5044-5044 Wholesale - office equipment 5700-5700 Retail - home furniture and equipment stores 5045-5045 Wholesale - computers 5710-5719 Retail - home furnishings stores 5046-5046 Wholesale - commercial equip 5720-5722 Retail - household appliance stores 5047-5047 Wholesale - medical, dental equip 5730-5733 Retail - radio, TV and consumer electronic stores 5048-5048 Wholesale - ophthalmic goods 5734-5734 Retail - computer and computer software stores 5049-5049 Wholesale - professional equip and supplies 5735-5735 Retail - record and tape stores 5050-5059 Wholesale - metals and minerals 5736-5736 Retail - musical instrument stores 5060-5060 Wholesale - electrical goods 5750-5799 Retail 5063-5063 Wholesale - electrical apparatus and 5900-5900 Retail - misc equipment 5910-5912 Retail - drug & proprietary stores 5064-5064 Wholesale - electrical appliance TV and 5920-5929 Retail - liquor stores radio 5930-5932 Retail - used merchandise stores 5065-5065 Wholesale - electronic parts 5940-5940 Retail - misc 5070-5078 Wholesale - hardware, plumbing, heating 5941-5941 Retail - sporting goods stores, bike shops equip 5942-5942 Retail - book stores 5080-5080 Wholesale - machinery and equipment 5943-5943 Retail - stationery stores 5081-5081 Wholesale - machinery and equipment (?) 5944-5944 Retail - jewelry stores 5082-5082 Wholesale - construction and mining 5945-5945 Retail - hobby, toy and game shops equipment 5946-5946 Retail - camera and photo shop 5083-5083 Wholesale - farm and garden machinery 5947-5947 Retail - gift, novelty 5084-5084 Wholesale - industrial machinery and 5948-5948 Retail - luggage equipment 5949-5949 Retail - sewing & needlework stores 5085-5085 Wholesale - industrial supplies 5950-5959 Retail 5086-5087 Wholesale - machinery and equipment (?) 5960-5969 Retail - non-store retailers (catalogs, etc) 5088-5088 Wholesale - trans eq except motor vehicles 5970-5979 Retail 5090-5090 Wholesale - misc durable goods 5980-5989 Retail - fuel & ice stores (Penn Central Co) 5091-5092 Wholesale - sporting goods, toys 5990-5990 Retail - retail stores 5093-5093 Wholesale - scrap and waste materials 5992-5992 Retail - florists 5094-5094 Wholesale - jewelry and watches 5993-5993 Retail - tobacco stores 5099-5099 Wholesale - durable goods 5994-5994 Retail - news dealers 5100-5100 Wholesale - nondurable goods 5995-5995 Retail - computer stores 5110-5113 Wholesale - paper and paper products 5999-5999 Retail stores 5120-5122 Wholesale - drugs & proprietary 5130-5139 Wholesale - apparel 43 Meals Restaurants, Hotels, Motels 5140-5149 Wholesale - groceries & related prods 5800-5819 Retail - eating places 5150-5159 Wholesale - farm products 5820-5829 Restaurants, hotels, motels 5160-5169 Wholesale - chemicals & allied prods 5890-5899 Eating and drinking places 5170-5172 Wholesale - petroleum and petro prods 7000-7000 Hotels, other lodging places 5180-5182 Wholesale - beer, wine 7010-7019 Hotels motels 5190-5199 Wholesale - non-durable goods 7040-7049 Membership hotels and lodging 7213-7213 Services - linen 42 Rtail Retail 5200-5200 Retail - bldg material, hardware, garden 44 Banks Banking 5210-5219 Retail - lumber & other building mat 6000-6000 Depository institutions 5220-5229 Retail 6010-6019 Federal reserve banks 5230-5231 Retail - paint, glass, wallpaper 6020-6020 Commercial banks 5250-5251 Retail - hardware stores 6021-6021 National commercial banks 5260-5261 Retail - nurseries, lawn, garden stores 6022-6022 State banks - Fed Res System 5270-5271 Retail - mobile home dealers 6023-6024 State banks - not Fed Res System 5300-5300 Retail - general merchandise stores 6025-6025 National banks - Fed Res System 5310-5311 Retail - department stores 6026-6026 National banks - not Fed Res System 5320-5320 Retail - general merchandise stores 5330-5331 Retail - variety stores 5334-5334 Retail - catalog showroom 5340-5349 Retail 5390-5399 Retail - Misc general merchandise stores 5400-5400 Retail - food stores 5410-5411 Retail - grocery stores 5412-5412 Retail - convenience stores 5420-5429 Retail - meat, fish mkt 5430-5439 Retail - fruit and vegetable markets 5440-5449 Retail - candy, nut, confectionary stores 5450-5459 Retail - dairy product stores 5460-5469 Retail - bakeries 5490-5499 Retail - miscellaneous food stores 5500-5500 Retail - auto dealers and gas stations 5510-5529 Retail - auto dealers 5530-5539 Retail - auto and home supply stores 5540-5549 Retail - gasoline service stations 5550-5559 Retail - boat dealers 5560-5569 Retail - recreational vehicle dealers

167 48 INDUSTRIES (continued) 9900-9999 Not classifiable

6027-6027 National banks, not FDIC 6028-6029 Banks 6030-6036 Savings institutions 6040-6059 Banks (?) 6060-6062 Credit unions 6080-6082 Foreign banks 6090-6099 Functions related to deposit banking 6100-6100 Nondepository credit institutions 6110-6111 Federal credit agencies 6112-6113 FNMA 6120-6129 S&Ls 6130-6139 Agricultural credit institutions 6140-6149 Personal credit institutions (Beneficial) 6150-6159 Business credit institutions 6160-6169 Mortgage bankers 6170-6179 Finance lessors 6190-6199 Financial services

45 Insur Insurance 6300-6300 Insurance 6310-6319 Life insurance 6320-6329 Accident and health insurance 6330-6331 Fire, marine, property-casualty ins 6350-6351 Surety insurance 6360-6361 Title insurance 6370-6379 Pension, health, welfare funds 6390-6399 Insurance carriers 6400-6411 Insurance agents

46 RlEst Real Estate 6500-6500 Real estate 6510-6510 Real estate operators 6512-6512 Operators - non-resident buildings 6513-6513 Operators - apartment buildings 6514-6514 Operators - other than apartment 6515-6515 Operators - residential mobile home 6517-6519 Lessors of real property 6520-6529 Real estate 6530-6531 Real estate agents and managers 6532-6532 Real estate dealers 6540-6541 Title abstract offices 6550-6553 Real estate developers 6590-6599 Real estate 6610-6611 Combined real estate, insurance, etc

47 Fin Trading 6200-6299 Security and commodity brokers 6700-6700 Holding, other investment offices 6710-6719 Holding offices 6720-6722 Investment offices 6723-6723 Management investment, closed-end 6724-6724 Unit investment trusts 6725-6725 Face-amount certificate offices 6730-6733 Trusts 6740-6779 Investment offices 6790-6791 Miscellaneous investing 6792-6792 Oil royalty traders 6793-6793 Commodity traders 6794-6794 Patent owners & lessors 6795-6795 Mineral royalty traders 6798-6798 REIT 6799-6799 Investors, NEC

48 Other Miscellaneous (also includes other SIC codes not in a specific industry) 3990-3990 Misc manufacturing industries 3992-3994 Misc manufacturing industries 3997-3999 Misc manufacturing industries 4950-4959 Sanitary services 4960-4961 Steam, air conditioning supplies 4970-4971 Irrigation systems 4990-4991 Cogeneration - SM power producer

168 BIBLIOGRAPHY

Admati, A. & Pfleiderer, P. (1988). A theory of intraday patterns: Volume and price variability. Review of Financial Studies, 1, 3-40.

Altman, E. I. (1968) Financial Ratios, Discriminant Analysis and the Prediction of Corporate Bankruptcy. Journal of Finance, 23, 4, 589-609.

Angel, J. (1997). Tick size, share prices, and stock splits. Journal of Finance (June), 655-581.

Angel, J., Brooks, R., & Matthew, P.G. (2004). When-issued shares, small trades, and the variance of returns around stock splits. Journal of Financial Research 27.3

Anshuman, V.R. & Kalay, A. (2002). Can splits create market liquidity? Journal of Financial Markets, 5, 83-125.

Arbanell, J., Lanen W., & Verrecchia, R. (1995). Analysts' forecasts as proxies for investor beliefs in empirical research. Journal of Accounting and Economics (July), 31-60.

Asquith, P., Healy, P.M. & Palepu, K.G. (1989) Earnings and stock splits. The Accounting Review 64, 3, 387-403

Baker, H.K. & Gallagher, P. (1980). Management view of stock splits. Financial Management 9, 73-77. Baker, H.K., & Phillips, A. (1994). Why companies issue stock dividends. Financial Practice and Education, 3, 29-27

Baker, H.K., Phillips, A.L., & Powell, G.E. (1995). The stock distribution puzzle: A synthesis of the literature on stock splits and stock dividends. Financial Practice and Education (Spring/Summer), 24-37.

Baker, H.K. & Powell, G.E. (1993). Further evidence on managerial motives for stock splits. Quarterly Journal of Business and Economics 32, 20-31.

Baker, M. & Wurgler J. (2000). The equity share in new issues and aggregate stock returns. Journal of Finance, 55, 2219-2257.

Barber, B. & Lyon, J. (1997). Detecting long-run abnormal stock returns: The empirical power and specification of test statistics. Journal of Financial Economics, 43, 341-372

Barberis, N., Schleifer, A. & Vishny, R. (1998). A model of investor sentiment. Journal of Financial Economics, 43, 307-343.

Barker, C.A., (1956). Effective stock splits, Harvard Business Review, 34, 101-106.

Beatty, R. & Kadiyala, P. (2003). Impact of the penny stock reform act of 1990 on the initial public offering market. Journal of Law and Economics, 47, 517-541.

Bikhchandani, S., Hirshleifer, D. & Welch, I. (1992). A theory of fads, fashion, custom, and cultural change as informational cascades. Journal of Political Economy, 100, 992.

169 Boehme, R.D. Danielsen, B.R. & Sorescu, S.M. (2003). Stock split post-announcement underreaction or market friction? Working paper available online at http://www.rdboehme.com/Papers/splits_sept_2004.pdf.

Boehme, R.D. & Sorescu, S.M. (2002). The long-run performance following dividend initiations and resumptions: Underreaction or product of chance? Journal of Finance, 57, 871-900

Bothwell, J.C., Jr. (1950). Periodic stock dividends. Harvard Business Review (January), 89- 100.

Brennan, M.J. (1995). The individual investor. Journal of Financial research, 18, 59-74.

Brennan, M.J. & Copeland, T.E. (1988). Stock splits, stock prices, and transaction costs. Journal of Financial Economics (October), 83-101.

Brennan, M.J. & Hughes, P.J. (1991). Stock prices and the supply of information, Journal of Finance (December), 1665-1691.

Brown, L.D. (1993). Earnings forecasting research: Its implications for capital markets research. International Journal of Forecasting (November), 295-320.

Byun, J. & Rozeff, M.S. (2003), Long-run performance after stock splits: 1927 to 1996, Journal of Finance 58, 1063-1085.

Chen, N., Roll R., & Ross, S. (1986). Economic forces and the stock market, Journal of Business, 59, 383-403.

Conrad, J.S. & Conroy, R.M. (1994). Market microstructure and the ex-date return. Journal of Finance, 49, 1507-1519.

Conroy, R.M. & Harris, R.S. (1999). Stock splits and information: The role of share price. Financial Management, 28, 28-40.

Conroy, R.M., Harris R.S., & Benet, B.A. (1995). The effects of stock splits on bid-ask spreads. Journal of Finance, 45, 1285-1295.

Constantinides, G. (1984). Optimal stock trading with personal taxes. Journal of Finance, 13, 65-89.

Copeland, T.E. (1979). Liquidity changes following stock splits. Journal of Finance, 34, 115-141.

Corwin, S. & Lipson, M. (2000). Order flow and liquidity around NYSE trading halts, Journal of Finance, 55, 1771-1801.

Crum, W.L. (1953). Analysis of stock ownership, Harvard Business Review (May-June), 36-54.

Daniel, K.D., Hirshleifer, D. & Subrahmanyam, A. (2001). Overconfidence, arbitrage, and equilibrium asset pricing, Journal of Finance, 56, 921-956.

Dennis, P.J. & Strickland, D. (1998). The effect of stock splits on liquidity: Evidence from shareholder ownership composition. Working Papers Series, 98-1. Available at SSRN: http://ssrn.com/abstract=93658 or DOI: 10.2139/ssrn.93658

170 Desai, A.S., Nimalendran, M., & Venkatraman, S. (1998). Changes in trading activity following stock splits and their effect on volatility and the adverse information component of the bid- ask spread. Journal of Financial Research, 21, 159-183.

Desai, H. & Jain, P.C. (1997). Long-run common stock returns following stock splits and reverse splits. Journal of Business, 70, 409-433.

Dewing, A.S. (1953). The financial policy of corporations (5th ed.). New York: Ronald Press.

Dolley, J.C. (1933). Common-stock split-ups: Motives and effects. Harvard Business Review (October), 70-81.

Dolley, J.C. (1934). The price-effect of stock right issues. Journal of Business (April), 133-160.

Dravid, A.R. (1987). A note on the behavior of stock returns around ex-dates of stock distributions, Journal of Finance, 42, 163-168.

Dubofsky, D.A. & French, D.W. (1986). Stock splits and implied stock-price volatility. Journal of Portfolio Management, 12 (4), 55-59.

Dubofsky, D.A. (1991). Volatility increases subsequent to NYSE and AMEX stock splits. Journal of Finance, 46, 421-431.

Easley, D., O'Hara, M., & Saar, G. (2001). How stock splits affect trading; A microstructure approach. Journal of Financial and Quantitative Analysis 36, 25-52.

Elton, N., Gruber, M., & Rentlzler, J. (1989). New public offering, information, and investor rationality: The case of publicly offered commodity funds. Journal of Business, 62, 1-15.

Fama, E. & French K. (1992). The cross-section of expected stock returns. Journal of Finance, 46, 427-465.

Fama, E. (1998). Market efficiency, long-term returns, and behavioral finance. Journal of Financial Economics. 49, 283–306

Fama, E., Fisher, L., Jensen, M. & Roll, R. (1969). The adjustment of stock prices to new information. International Economic Review (February), 1-21.

Fernando, C.S., Krishnamurthy, S. & SpindtIs, P.A. (1999). Share price related to marketability? Evidence from mutual fund share splits. Financial Management 28 (3), 15-31.

Gillespie, W. and Seitz, N. (1977), Returns for and Results of Reverse Stock Splits, Paper presented at regional AIDS meeting 1979.

Gleason, K.C., Rosenthal, L., Lee, C. (2004). When is a reverse split good news? Wealth effects and performance implications of reverse stock splits. Presented at the Financial Management Association Conference in New Orleans. Available at http://207.36.165.114/NewOrleans/Papers/1801768.pdf

Grinblatt, M.S., Masulis, R.W. & Titman, S. (1984). The valuation effects of stock splits and stock dividends. Journal of Financial Economics (December), 461-490.

Han, K.C. (1995). The effects of reverse splits on the liquidity of the stock. Journal of Financial and Quantitative Analysis (March), 159-169.

171 Hulbert, M. (1996). A strong signal. Forbes (April), 435.

Ikenberry, D.L. & Ramnath, S. (2002). Underreaction to self selected news events: The case of stock splits. The Review of Financial Studies 15, 489-526.

Ikenberry, D.L., Lakonishok, J., & Vermaelen, T. (1995). Market underreaction to open market share repurchases. Journal of Financial Economics (October), 181-208.

Ikenberry, D.L., Lakonishok, J., & Vermaelen, T. (2000). Stock repurchases in Canada: Performance and strategic trading. Journal of Finance, (55) 5, 2373-2398.

Ikenberry, D.L., Rankine, G. & Stice, E.K. (1996) What do stock splits really signal? Journal of Financial and Quantitative Analysis (September), 357-375.

Jing, L. (2003). An event study of reverse stock splits in Hong Kong market. EFMA 2003 Helsinki Meetings. Available at SSRN: http://ssrn.com/abstract=393222 or DOI: 10.2139/ssrn.393222

Kadiyala, P. & Vetsuypens, M.R. (2002). Are stock splits credible signals? Evidence from short- interest data. Financial Management, (Spring),

Kahneman, D. & Tversky, A. (1979). Prospect theory: An analysis of decision under risk. Econometrica (March), 263-291.

Kahya, E., Theodossiu, P. (1999). Predicting corporate financial distress: A time-series CUSUM methodology. Review of Quantitative Finance and Accounting, 13, 323-345

Kiang, M.Y., Fisher, D.M., Fisher, S.A., Chi, R.T. Understand corporate rationales for engaging in reverse stock splits - a data mining application. hicss, p. 75a, Proceedings of the 38th Annual Hawaii International Conference on System Sciences (HICSS'05) - Track 3, 2005

Kim, J., Jain, P., Jiang, C., McInish, T., & Wood, R.A. (2003). Reverse stock splits, Working Paper, North Dakota State University and University of Memphis.

Kimball, P. & Papera, D.R. (1964). Effect of stock splits on short-term market prices. Financial Analysts Journal (May-June), 75-80.

Klein, A., Rosenfeld, J., & Tucker, J. (2005). Return performance surrounding reverse stock splits: Can investors profit? Available at SSRN: http://ssrn.com/abstract=679028

Kryzanowski, L. & Zhang, H. (1996). Trading patterns of small and large traders around stock split ex-dates. The Journal of Financial Research 19, 75-90.

Lakonishok, J. & Lev, B. (1987). Stock splits and stock dividends: Why, who, and when, Journal of Finance (September), 913-932.

Lakonishok, J. & Vermaelen, T. (1990), Anomalous price behavior around repurchase tender offers. The Journal of Finance 45, 455-477.

Lakonishok, J., Shleifer, A. & Vishny, H. (1992). The structure and performance of the money management industry. Brookings Papers on Economic Activity, Microeconomics, 339-391.

Lamoureux, C.G. & Poon, P. (1987). The Market Reaction to Stock Splits. The Journal of Finance. 42, 1347-1370.

172 Leland, H. & Pyle, D.H. (1977). Informational asymmetries, financial structure, and financial intermediation. The Journal of Finance, 32, 371-387.

Lewin, (1996). Economics and psychology, lessons for our own day. The Journal of Economic Literature, 34, 3

Linn, S. & McConnell, J.J. (1983). An empirical investigation of the impact of anti-takeover amendments on common stock prices. Journal of Financial Economics, 11, 361-399.

Loughran, T. & Ritter, J. (1995). The new issues puzzle, The Journal of Finance, 50, 23-51.

Lyon, J., Barber, B., & Tsai, C. (1999). Improved methods for tests of long-run abnormal stock returns. The Journal of Finance 54, 165-201.

Maloney, M. & J.H. Mulherin, (1992). The effects of splitting on the ex: A Microstructure reconciliation. Financial Management, 21, 44-59.

McDonald, R.F., (1960). Are high prices discouraging small investors? Financial Analysts Journal (May-June), 21-30.

McNichols, M. & Dravid, A. (1990). Stock dividends, stock splits, and signaling. Journal of Finance (July), 857-879.

Merton, Robert C. (1987). A simple model of capital market equilibrium with incomplete information. The Journal of Finance, 42, 483-510.

Mitchell, M & Stafford, E. (2000). Managerial decisions and long-term stock-price performance. Journal of Business, July.

Muscarella, C. & Vetsuypens, M.R. (1996). Stock splits: signaling or liquidity? The case of ADR solo-splits. Journal of Financial Economics, 42, 2-27.

Myers, J.H. & Bakay, A.J. (1948). Influence of stock split-ups on market price. Harvard Business Review (March), 251-255.

Nayak, S. & Prabhala, N.R. (2001). Disentangling the dividend information in splits: A decomposition using conditional event-study methods. The Review of Financial Studies, 14, 1083-1116.

Nayar, N. & Rozeff, M.S. (2001). Record date, when-issued, and ex-date effects in stock splits. Journal of Financial and Quantitative Analysis, 36: 119-139.

Niederhoffer, V., (1966). A new look at clustering of stock prices. Journal of Business (April), 309-313.

Ohlson, J. & Penman, S.H. (1985). Volatility increases subsequent to stock splits: An empirical aberration. Journal of Financial Economics, 14, 251-266.

Patell, J. (1976). Corporate forecasts of earnings per share and stock price behavior: empirical tests. Journal of Accounting Research (Autumn): 246-276.

Peterson, D.R. & Peterson, P.P. (1992). A further understanding of stock distributions: The case of reverse stock splits. Journal of Financial Research, 15, 189-205

173 Pilotte, E. & Manuel, T. (1996). The market's response to recurring events: The case of stock splits. Journal of Financial Economics (May), 111-127.

Porter, D.C. & Weaver, D.G. (1997). Tick size and market quality. Financial Management (Winter), 5-26.

Radcliffe, R. C. & W. Gillespie, (1979). The price impact of reverse splits. Financial Analysts Journal, 35, 63-67.

Reason, T., (2002). Reverse psychology today. CFO, Magazine for Senior Financial Executives, (December).

Ritter, J.R. (1991). The long-run performance of initial public offerings. The Journal of Finance, 46, 3-27.

Ross, S.A. (1977). The determination of financial structure: The incentive-signaling approach. The Bell Journal of Economics, 8, 23-40.

Rozeff, M.S., (1998). Stock splits: Evidence from mutual funds. Journal of Finance (February), 335-349.

Schipper, K. & Smith, A. (1986). A comparison of equity carve-outs and seasoned equity offerings: Share price effects and corporate restructuring. Journal of Financial Economics, 15, 153-186.

Schultz, P., (1997). Why do stocks split? The optimal size and preferred price range hypotheses. University of Chicago Graduate School of Business CRSP Working Paper 458 (November).

Schultz, P., (2000). Stock splits, tick size, and sponsorship. The Journal of Finance, 55, 429-450.

Shefrin, H. & Statman, M. (1984). Explaining investor preference for cash dividends. Journal of Financial Economics (June), 253-282.

Shefrin, H. & Statman, M. (1993). Behavioral Portfolio Theory. Journal of Financial and Quantitative Analysis 35, 127-151. Sheikh, A.M. (1989), Stock splits, volatility increases, and implied volatilities. The Journal of Finance, 44, 5, 1361-1372.

Shiler, R.J. (2005). Irrational Exuberance (2nd Ed). Princeton University Press: Princeton, NJ.

Spiess, D. & Affleck-Graves, J. (1995). Underperformance in long-run returns following seasoned equity offerings. Journal of Financial Economics, 38, 234-267.

Spudeck, R. & Moyer, C. (1985). Reverse splits and shareholder wealth: The impact of commissions. Financial Management, 14, 52-56.

Szewczyk, S. & Tsetsekos, G.P. (1993). The effect of managerial ownership stock split-induced abnormal returns. The Financial Review, 28, 351-370.

Tawatnuntachai, O. & D'Mello, R. (2002). Intra-industry reactions to stock split announcements. The Journal of Financial Research, 25, 39-57.

174 Theodossiou, P., Kahya, E., Saidi, R., & Philippatos, G. (1996). Financial distress and corporate acquisitions: Further empirical evidence. Journal of Business Finance & Accounting 23, 699–719.

Vafeas, N., (2001). Reverse stock splits and earnings performance. Accounting and Business Research, 31 (3), 191-202.

West, R. & Brouileter, A. (1970). Reverse stock splits…harbinger of bad times or valid management technique, Financial Executive, 38, 12-17.

White, H., (1980). A heteroskedasticity-consistent covariance matrix estimator and a direct test for heteroskedasticity. Econometrica (May), 817-829.

Woolridge, J.R. & Chambers, D.R. (1983). Reverse splits and shareholder wealth. Financial Management 12, 5-15

175 BIBLIOGRAPHICAL SKETCH

Barry Marchman is originally from Macon, Georgia. He received his undergraduate degree in Electrical Engineering from Georgia Institute of Technology and worked as an engineer for several multinational companies before beginning his doctoral work at Florida State University. He has been married to Evelyn Steen since March 1995 and together they have had six children (four while doing doctoral work). He currently resides in Tallahassee, Florida.

176