Estimating the Growth Rate to Use in Terminal Value Calculations

HCVI

Henley Centre for Value Improvement

Testing IPO Pricing

Two Contrasting Telecomm Illustrations

Roger W. Mills*

Henley Discussion Paper Series

(HCVI HDP No. 7, June 2005)

* The author wishes to express his thanks to Mr Teo Cocca, University of Zurich, for his assistance with the analysis of Swisscom

ISBN 1 86181 236 1

Henley Centre for Value Improvement (HCVI) Henley Management College Greenlands Henley-on-Thames Oxon RG9 3AU

Tel: + 44 1491 418762 Fax: + 44 1491 571574 Email: [email protected] Website: www.henleymc.ac.uk/hcvi

©Roger Mills 2 Henley Discussion Paper Series HCVI HDP No. 7

Abstract

In this paper a framework that can be used for challenging the assumptions used in IPO pricing is described and illustrated using IPO discounted cash flow valuations of Jordan

Telecom and Swisscom that were produced by analysts. The framework makes use of information about comparable businesses that is applied in challenging key assumptions using

Market Implied Competitive Advantage Period (MICAP) analysis.

Typically, IPOs are valued using a variety of methods and within the valuations there will often be a discounted cash flow (DCF) estimate in some form, about which great care has to be exercised by the reader in terms of understanding the implicit and explicit assumptions used. Conventional wisdom in estimating the value of businesses with long-term growth potential is to use a two stage model, where the first stage represents a time period in which growth is assumed to be captured in the explicit growth in individual cash flows over a finite time period. The second stage is assumed to be a perpetuity thereby avoiding the need to forecast the growth in individual cash flows; instead a composite growth rate, ‘g’, for free cash flows can be assumed. The selection of a growth rate has a significant impact upon the value and the higher the growth rate the higher the value .

A major concern typically arises about the impact of the g on the resulting value and, therefore, how it might be estimated in any meaningful way. At the end of the day, the user of investment research needs to be able to establish confidence with the assumptions used and the problem is that within corporate finance there is limited evidence of any readily available framework

©Roger Mills 3 Henley Discussion Paper Series HCVI HDP No. 7

In this paper a framework for challenging perpetuity with growth assumptions is proposed with reference to the case of Jordan Telecom. The use of this company has the advantage that it is a relatively recent issue in a sector often beset by forecasting challenges and, most importantly, it draws upon information predominately available in the investment research report itself. Thereafter, the outcome is contrasted with that for Swisscom, for which the same framework was applied. These two IPO valuations had markedly contrasting results, but the analysis provided here illustrates how the framework can be applied to help ensure that the right questions about IPO pricing are being asked.

©Roger Mills 4 Henley Discussion Paper Series HCVI HDP No. 7

Introduction

As we have just seen with Google, the valuation and pricing of an in Initial Public Offer (IPO) is always a difficult and contentious issue1. Trying to gauge market sentiment and setting a price that does not spell disaster in terms of the desired objectives is a real challenge2.

Typically, IPOs will be valued using a variety of methods and within the valuations there will often be a discounted cash flow (DCF) estimate in some form, about which great care has to be exercised by the reader in terms of understanding the implicit and explicit assumptions used. In very simple terms, the key components within a DCF valuation typically include assumptions about:

the generation of the cash flows, both in the period for which explicit and detailed

estimates are made and the period beyond (often referred to as the continuing,

terminal, or residual value);

the discount rate to use, i.e. the cost of capital;

the method to use for valuing the time period beyond the explicit cash flow forecast;

and

the time horizon to use for the explicit forecast

A number of academics and practitioners have identified many errors that can be made in undertaking DCF valuations3. Furthermore, behavioural finance which has its roots in the pioneering work of psychologists Daniel Kahneman and Amos Tversky has traced the mistakes that humans commit when making decisions to four common causes that include self-deception as characterised by such effects as over-optimism and over-confidence 4,5.

Over-optimism and over-confidence may often be encountered in DCF valuations in estimating the growth in free cash flows within the explicit forecast time horizon but, more ©Roger Mills 5 Henley Discussion Paper Series HCVI HDP No. 7

importantly in the period beyond. Conventional wisdom in estimating the value of businesses with long-term growth potential is to use a two-stage model, where the first stage represents a time period in which growth is assumed to be captured in the explicit growth in individual cash flows over a finite time period. The second stage is assumed to be a perpetuity thereby avoiding the need to forecast the growth in individual cash flows; instead a composite growth rate, ‘g’, for free cash flows can be assumed. As a consequence, value from the second stage can be very simply calculated as follows, where FCF (t+1) represents the estimated prospective perpetuity cash flow and WACC the weighted average cost of capital:

FCF (t+1)

Terminal Value =

WACC-g

Therefore, to summarise, whereas the first stage of a conventional DCF valuation is typified by the use of explicit forecast assumptions being made for individual line items of the projected free cash flow, g in the second stage often represents a number that is designed to capture in some way the sum total of expectations about individual line items. The deduction of a growth rate in the denominator of the calculation has a significant impact upon the value and the higher the growth rate the higher the value6.

As anyone who has worked with such models will know, a major concern typically arises about the impact of the g on the resulting value and, therefore, how it might be estimated in any meaningful way. The user of investment research needs to be able to establish confidence with the assumptions used and the problem is that within corporate finance there is limited evidence of any readily available framework.

©Roger Mills 6 Henley Discussion Paper Series HCVI HDP No. 7

There are many examples of the application of the perpetuity with growth approach illustrated in equation (2), of which a good one was provided in the valuation of Jordan Telecom, the first case study subject of this paper. The background was that in September 2002, the

Jordanian government sought to sell 15% of its share capital (37.5 million shares) in an initial public offering (IPO).

The IPO was assessed by a number of analysts of which the following is the example that will be used here7:

‘Using the discounted cash flow (DCF) method, we arrived at an estimated fair value for

Jordan Telecom in the range of JD618.8-671.9 million, which translates into JD (Jordanian

Dinars) 2.48-2.72/sharei.

In terms of the 2-stage model discussed earlier, a seven year explicit forecast period was used to estimate free cash flows, which were then discounted to a present value to produce JD272.4 million as the value of stage 1. The value of stage 2, the terminal value, was calculated using two assumptions for the perpetuity growth rate. Based upon a 5% perpetuity growth rate it was JD428.5 million (61.14% of the total enterprise value) and it was JD489.5 million using a

6% perpetuity growth rate (64.25% of the total enterprise value).

As with all such perpetuity with growth valuations, the all important question arises, ‘why 5% or 6%?’ The purpose of this case study is to offer a framework for challenging such growth assumptions. It is based upon the tools and techniques of corporate finance and, as with all such frameworks, there are strengths and weaknesses, but our experience has been that the former outweigh the latter.

i Assumes 5% and 6% growth in perpetuity assumptions

©Roger Mills 7 Henley Discussion Paper Series HCVI HDP No. 7

In what follows, the framework proposed for challenging the assumptions used in such analysis will be introduced with reference to the case of Jordan Telecom. The use of this company has the advantage that it is a relatively recent issue in a sector often beset by forecasting challenges and, most importantly, it draws upon information predominately available in the investment research report itself. Thereafter, the outcome is contrasted with that for Swisscom, for which the same framework was applied.

Proposed Framework for Challenging IPO Pricing

The 5 steps outlined in Figure 1 are proposed as a framework to use to interrogate the results from a DCF valuation for an IPO.

Figure 1: 5 Step Framework

1. Select comparable businesses with the target company being valued for IPO purposes,

obtain a relevant share price for them, up-to-date financials, estimates for cash flow

growth and the WACC. The objective here is to obtain estimates of all valuation

inputs other than time, so that we can use Market Implied Competitive Advantage

Period (MICAP) analysis to solve for the explicit time period implied in the share

price, assuming a simple perpetuity in the period beyond.

2. Find the time period for all comparable businesses that equates prospective cash flows

based on the assumptions, with the ‘relevant’ share price. This time period represents

the MICAP for the comparator company and will be used for defining the second time

period in a 3 stage valuation model.

©Roger Mills 8 Henley Discussion Paper Series HCVI HDP No. 7

3. Use this 3-stage model to calculate 3 values, the first being the original explicit

forecast; the second value representing the discounted cash flow value of free cash

flows from the end of the first stage up to the MICAP for the comparator business, the

free cash flows being estimated from the growth profile of comparable businesses; and

a third stage where the value is calculated from a simple perpetuityii.

4. The sum of the values of stages 2 and 3 of the 3-stage model are analysed using the

perpetuity with growth model for each of the comparable businesses to establish what

the perpetuity growth rate would have to be to justify the free cash flows in the

perpetuity period beyond stage 1 of a 2-stage model.

5. Apply the growth rates from Step 4. to the 2-stage valuation model and iterate on

share price as an input to the WACC calculation and the share price as an output.

This 5 Step framework essentially provides a methodology for the fundamental questioning of the cash flow assumptions used in a DCF analysis, particularly for the terminal value, which typically represents the largest part of the valuation. This questioning focuses upon the use of comparables analysis to form a view about what is a reasonable perpetuity with growth estimate, based upon the experiences of similar businesses. In effect, Market Implied

Competitive Advantage Period (MICAP) is used to ask the question, “Using the value implied in competitors’ share prices, for how many years does the return on capital exceed the cost of capital”8. This question is posed with the assumption that free cash flows and the cost of capital can be estimated, such that the number of years of explicit cash flows required to arrive at the current market price of the share consistent with a steady state in the terminal period can be calculated as the only unknown variable . As such, it is important to note that

ii Terminal Value= FCF(t+1)/WACC, i.e. zero growth

©Roger Mills 9 Henley Discussion Paper Series HCVI HDP No. 7

the terminal value method proposed here in the application of the approach is the simple perpetuity because all growth is assumed to be captured over the MICAP. In other words, a 2- stage model is used, where stage 1 corresponds with the assumption that the business will generate an economic return that exceeds the cost of capital, whereas in stage 2 it is assumed a steady state is reached in which the business generates an economic return that is equal to the cost of capital.

An alternative to avoid concerns with the impact of perpetuity with growth estimates in 2- stage models is to use a 3-stage model, so that the growth element can be modelled more explicitly. In such a 3-stage model, as with the 2-stage version, the first stage involves a detailed explicit forecast for all items comprising the free cash flow such that assumptions about the envisaged growth characteristics of the company being valued are captured for this time period. The second stage involves making a growth forecast for simply the free cash flow over a finite time period, possibly based upon perceptions of growth in the sector, and the third stage assumes that all growth has been captured and is based upon the valuation of a single cash flow into perpetuity.

‘Explicit’ Cash Flow Simple Cash Flow Forecast - Perpetuity Forecast - Cash Flow Based upon Based upon Detailed Company or Company Sector Growth Projections Characteristics

Stage 1 Stage 2 Stage 3

©Roger Mills 10 Henley Discussion Paper Series HCVI HDP No. 7

The problem frequently encountered with using the 3-stage model relates to the estimation of the duration of the second stage. It will be illustrated that this time horizon can be estimated using MICAP and the results of its estimation can be used to provide a cross check for growth estimates in a 2-stage valuation model. In addition to challenging the original perpetuity with growth rates, those resulting from the analysis can be used to produce a revised equity valuation for the company.

In applying the 5 Step Framework we acknowledge that it would be unrealistic to assume that there is any definitive way for estimating a correct growth rate to use – that would be tantamount to suggesting that the future can be foretold! However, in this paper it is argued using Jordan Telecom as an illustration that this framework can be used to challenge growth in perpetuity assumptions.

Applying the 5 Step Framework to Jordan Telecom9

In 1971, Telecommunications Corporation (TCC) was established as a state-owned corporation to provide and operate telecommunications services in Jordan (telephone, telegraph, and telex). In 1997 TCC was privatized and adopted the name Jordan

Telecommunications Company (JTC). In January 2000, a 40% stake was sold to the Joint

Investment Telecommunications Company (JITCO), thereby establishing Jordan Telecom. In

September 2002 the Jordanian government wanted to sell 15% of its share capital (37.5 million shares) through an IPO.

The prospective IPO generated a flurry of research reports. The one drawn upon here was representative and offered the following evaluation: ©Roger Mills 11 Henley Discussion Paper Series HCVI HDP No. 7

Using the discounted cash flow (DCF) method, we arrived at an estimated fair value

for Jordan Telecom in the range of JD618.8-671.9 million (Jordanian dinars), which

translates into JD2.48-2.72/share.

The report provided a two-stage discounted cash flow valuation model based on a seven-year explicit forecast period and two terminal value scenarios based on growth in perpetuity of 5% and 6%. The analyst estimated free cash flows over the seven-year forecast period (see Table

1) and then discounted them to a present value of JD272.4 million as the value of Stage 1.

Based on a 5% perpetuity growth rate, the terminal value was JD428.5 million (61% of total enterprise value) and it was JD489.5 million with the 6% perpetuity growth rate (64% of total enterprise value). The analyst then subtracted the market value of debt and preferred stock

(JD82 million) to arrive at an equity market value for Jordan Telecom in the range of

JD618.8–671.9 million, or JD2.48–2.72 per share based on 250 million shares outstanding.

Table 1: Analysts’ Free Cash Flow Estimates (millions of Jordanian dinars)

STAGE 1 Year 2002 2003 2004 2005 2006 2007 2008

Free Cash Flow 43.10 55.70 59.60 70.20 73.90 76.00 78.70

The critical issue is that the terminal value, whether based on 5% or 6% growth, accounted for more than 60% of the total value even with a seven-year explicit forecast period. With such a large proportion of the total value affected by these assumed growth rates, it is critical to understand their validity. Since much of the analysis was based upon comparisons with

©Roger Mills 12 Henley Discussion Paper Series HCVI HDP No. 7

selected peer group companies, the validity of the two perpetuity-with-growth estimates can be reviewed by applying the five-step model outlined earlier.

In the case of Jordan Telecom, the peer group comparables were from Central and Eastern

Europe which, according to the analyst, offered the closest match with the Jordanian market situation in terms of penetration levels, per capita GDP, credit ratings, and the standing of their mobile subsidiaries. The more geographically proximate Gulf operators were judged to be in a more advanced stage of development than Jordan Telecom.

Once the comparable businesses have been identified, the next step is to obtain up-to-date financial information, generate prospective cash flow forecasts, and estimate their cost of capital (WACC). Financial information is usually available; and WACC estimation, while challenging, is not normally a major problem. Estimating free cash flow growth rates is a little more difficult because consensus estimates for the future may not be available, although the initial use of historical trend analysis balanced against an assessment of the prospective strategy is probably a good starting point.

The projected free cash flow profile from Step 2 is extended until the current share price is reached and a simple perpetuity terminal value is obtained for the continuing period. This can be viewed as a steady state period in which the business generates an economic return that is equal to the cost of capital; there is no economic incentive to invest other than to replace assets that have worn out or become obsolete. Nonetheless, there is no incentive to exit the business, because in principle all costs including those relating to capital are covered and economic theory would suggest that a normal profit is being earned. The time period up until

©Roger Mills 13 Henley Discussion Paper Series HCVI HDP No. 7

steady state is an estimate of the MICAP, or the period of competitive advantage implied in the share price.

As regards Jordan Telecom, one of the peer group companies cited in the analyst’s report was the Polish telecom company Telekomunikacja Polska SA (TPSA), which was trading between

11.40 and 11.50 Polish zloty at the time of the analysis. Using a mid-range share price of

11.44 zloty as the reference point, projected cash flows based on a historical trend analysis resulted in a MICAP of ten years, which reflected annual free cash flow growth rates of

18.6%, 17.1%, and 15.9% (based on historical trend analysis) for the three years beyond the seven-year forecast period used in the Jordan Telecom analysis—that is, in Years 8, 9, and

10.iii In other words, the only way to obtain a value close to the mid-range share price of

11.44 zloty was to extend the explicit cash flow forecast period to ten years, after which point a simple perpetuity resulted. Table 2 summarizes these calculations.

iii The annual growth in Jordan Telecom’s free cash flow over the seven-year forecast period never reaches these levels, which further supports the five-step approach because, as we see later, even with these much higher annual growth rates in cash flows we do not get near the analyst’s 5% to 6% perpetuity-with-growth estimate.

©Roger Mills 14 Henley Discussion Paper Series HCVI HDP No. 7

Table 2: MICAP Analysis for TPSA (millions of zloty)

Year 8 9 10

PV of Future Cash Flows 14.3 16.4 16.4

Terminal Value 48.1 55.8 60.8

PV Terminal Value 14.3 14.5 15.8

Total Market Value 28.6 30.9 32.2

Less: Net Debt 16.2 16.2 16.2

Equity Value 12.4 14.7 16.0

Number of Shares 1.4m 1.4m 1.4m

Share Price (zloty) 8.89 10.52 11.44

In terms of the three-stage model described earlier, if the growth profile for TPSA is applied to Jordan Telecom, then years 1–7 represent Stage 1, years 8–10 represent the Stage 2 balance of the MICAP period of ten years, and Stage 3 is represented by all years after year 10.iv

If TPSA’s growth in free cash flow for each year from the end of the seven-year period up to the end of year 10 is representative of the potential growth in Jordan Telecom for Stage 2, then the growth rates of 18.6%, 17.1%, and 15.9% can be applied to the Jordan Telecom free

iv As indicated earlier, the difficulty with three-stage models is identifying where Stage 3 in particular starts, but the analysis of a comparable business allows this to be estimated. With regard to the start of Stage 2, the original seven-year estimate for Jordan Telecom was accepted on the grounds that only the growth rates used in perpetuity were being challenged in this instance.

©Roger Mills 15 Henley Discussion Paper Series HCVI HDP No. 7

cash flow forecast in estimating the cash flows in Stage 2 and also the simple perpetuity for

Stage 3. This is illustrated in Table 3.

Table 3: Application of TPSA Free Cash Flow Growth Rate to Jordan Telecom (millions of Jordanian dinars)

STAGE 2 STAGE 3 Year 2009 2010 2011

Growth Rate from TPSA MICAP 18.6% 17.1% 15.9%

FCF Forecast for Jordan Telecom 93.3 109.3 126.7

Discount Factor 0.882 0.778 0.686

Present Value (Free Cash Flow) 82.3 85.0 86.8

Cumulative Present Value 82.3 167.3 254.1

Terminal Value Without Growth 647.7

The value of Stage 2 of 254.1 million Jordanian dinars is the cumulative present value (as of year-end 2008) of the discounted free cash flows for years 8, 9, and 10 (2009–2011 inclusive) using a WACC of 13.4%.v The value of Stage 3 of 647.7 million Jordanian dinars is the terminal value without growth, or the simple perpetuity of 2011 cash flow (assumed to equal t+1) discounted back to a present value as of year-end 2008 (86.8/0.134).

v The 13.4% WACC is estimated from information provided in the analyst report: a risk-free rate of 9.5% (obtained using the stripped yield of a ten-year Brady bond), a beta of 0.96, an equity risk premium estimate of 6% for CEEMA markets, a cost of debt of 6%, outstanding debt of JD82 million, and a target equity-to-debt ratio of 4:1.

©Roger Mills 16 Henley Discussion Paper Series HCVI HDP No. 7

The original perpetuity growth estimates used by the analyst can now be assessed by drawing upon the information gathered from the peer comparison and applying it to the rearranged perpetuity-with-growth formula to find the growth rate associated with the sum of the values from Stages 2 and 3 of the three-stage model.

FCF (t+1)

Terminal Value (TV) =

WACC-g

FCF (t+1)

g = WACC -

With a WACC of 13.4%, a perpetuity free cash flow of JD93.3 million, and a terminal value of JD901.8 million (equal to the sum of the Stage 2 present value of JD254.1 million and the perpetuity present value of JD647.7 million from Table 2), the implied perpetuity growth rate beyond Stage 1 using the formula is 3.1%.

This perpetuity growth rate is substantially less than the 5% or 6% used in the original valuation. Of course, it was derived from an assessment of one of Jordan Telecom’s peer companies, but TPSA was one of the comparables used as the basis of the original analysis.

Still, it is quite reasonable to ask what the result would have been had the other peer companies been analyzed similarly—and in point of fact, there would have been no perpetuity growth warranted at all, because the MICAPs of the companies for which such analysis could be undertaken did not extend even to seven years.

©Roger Mills 17 Henley Discussion Paper Series HCVI HDP No. 7

The final part of the analysis involves using the growth rate calculated from the analysis of the peer group in a traditional two-stage model to gauge the impact on equity value.vi The

3.1% derived from applying the TPSA growth estimates to Jordan Telecom can now be applied to the two-stage model developed in the Jordan Telecom analyst report to “test” the implied IPO price.

In IPO pricing, of course, there is no established market equity value to use in the WACC calculation. In fact, within the WACC-based model used by the analysts, the market value of equity is both an input and an output; it is an input in terms of establishing the market-based target debt-to-equity ratio and it is also the output of the valuation exercise. One approach, therefore, is to iterate on the price, so that the IPO price and the price used in calculating the

WACC are mutually consistent. Since the amount of debt and all other variables for undertaking the valuation are known, we can treat the market equity value as the unknown variable and solve for it accordingly. In effect, an iteration is undertaken until the input market value of equity and the resulting output market value are the same. Of course, variation in the debt-to-equity ratio will affect the beta used to estimate the WACC, but we can apply the Hamada levering and unlevering formula in calculating the cost of equity to ensure consistency. Based on the information provided in the analyst report, the result is a

WACC of 13.9% and a price of JD1.89, as illustrated in Table 4.

vi The valuation can be undertaken within the three-stage model outlined; however, great user benefit has been demonstrated from relating the approach back to the original two stages.

©Roger Mills 18 Henley Discussion Paper Series HCVI HDP No. 7

Table 4: IPO Price Iteration for Jordan Telecom

The unlevered beta is estimated from the levered beta of 0.96 in the analyst’s report based on the debt-equity ratio associated with the price “guess,” assuming a tax rate of 40%; for Round

1, the unlevered beta is 0.96/[1.0 + (1 – 0.4) * 82/620]. The initial price guess is the lowest price estimate from the analyst’s report.

Round 1 Round 2 Round 3

Unlevered beta 0.89 0.87 0.87

Price (guess) (JD*) 2.48 1.85 1.89

Number of shares 2.5m 2.5m 2.5m

Equity value (JD) 620.0m 462.5m 472.5m

Debt/equity ratio 1/7.6 1/5.6 1/5.8

Levered beta 0.95 0.97 0.97

WACC 14.1% 13.9% 13.9%

DCF—Stage 1 (JD) 261.3m 263.0m 262.9m

PV of terminal value (JD) 282.9m 291.2m 290.6m

Business value (JD) 544.2m 554.2m 553.4m

Less: Debt (JD) 82.0m 82.0m 82.0m

Equity value (JD) 462.2m 472.2m 471.4m

Price (JD) 1.85 1.89 1.89

* JD: Jordanian Dinars

©Roger Mills 19 Henley Discussion Paper Series HCVI HDP No. 7

Swisscom9

Whilst the Jordan Telecom case illustrates the potential for using the approach to identify over optimism, the Swisscom case reveals quite the contrary: as will be illustrated the 5 Step

Framework can be used to identify potential IPO under-pricing.

With the liberalization of the Swiss (and European) telecommunications market on January 1,

1998, the legal monopoly of the Swiss Telecom PTT in the field of telecommunications was broken. The national telecom provider was separated from the postal activities, became a public limited-traded company on October 1, 1997 and changed its name to Swisscom AG.

In the fall of 1998 Swisscom and the Swiss Confederation offered to sell 34.5% of the shares

(to be outstanding after the offering) of Swisscom in IPO. Approximately one third of the total shares offered were newly issued shares being sold by Swisscom and two thirds were currently outstanding shares being sold by the Confederation.

After completion of the offering the Confederation still owned 65.5% of the outstanding shares of Swisscom. The Shares were approved for listing on the Swiss Exchange (SWX) and the ADSs were approved for listing on the New York Stock Exchange.

At that time, Swisscom’s revenues and net income information for 1996 and 1997 were as follows (CHF m):

©Roger Mills 20 Henley Discussion Paper Series HCVI HDP No. 7

1996 1997

Voice 6881 6668

Mobile 610 823

Data & Multimedia 762 824

Carrier Services 719 697

Total net revenues 9,532 9,842

Capitalized cost and changes in inventories 318 277

Operating Expenses 7345 9805

Operating Income 2505 314

Income before taxes 2044 -89

NET INCOME 1830 -415

A research report produced by one of the lead managers (Warburg Dillon Read, a division of

Swiss UBS) in support of the IPO provided a 2-stage valuation model for Swisscom, based upon a 7 year explicit forecast period and a terminal value based on a growth in perpetuity scenario. Using the DCF method a value for Swisscom in the range of CHF 41-54 billion was estimated, implying a share price of CHF 520-700 per share. After publication of the research report general market conditions deteriorated quickly due to turmoil in Emerging markets.

The bookbuilding price range was finally set at 330 to 410 CHF. Effective demand for stock was not overwhelming and the issuing price had to be set at the lower end of the price range,

which resulted in an issuing price of 340 CHF. The pricing of the initial research report

(only official source from the issuing consortium with an explicit DCF calculation) was adapted to the issuing price of 340 and deducted the implied assumptions about the future cash flows the issuing banks was thereby making.

Similar to the case of Jordan Telecom, how these assumptions can be analysed will be reviewed by applying the 5 Step Framework outlined earlier.

Applying the 5 Step Framework to Swisscom10

Step 1 - Select comparable businesses with the target company being valued for IPO

purposes, obtain a relevant share price for them, up-to-date financials, estimates for

cash flow growth and the WACC.

In the case of Swisscom, a number of comparable businesses were provided in the valuation document that will be drawn upon here for purposes of illustration. In order to select the best performing companies and the one, which according to the analyst, matched with the Swiss market situation most appropriately, Dutch KPN and British Telecom were chosen as relevant peers. For obvious reasons, data availability was also a point we had to consider when selecting.

As with Jordan Telecom, we used historical trend analysis to produce inputs for the development of free cash flows.

Step 2 - Find the time period for all comparable businesses that equates prospective

cash flows based on the assumptions, with the ‘relevant’ share price.

As regards Swisscom this analysis was undertaken for the peer group companies cited above.

Taking KPN as the example, the share price at the time of the analysis had been around NLG

85. Using historical trend analysis it was established that a 14 year time horizon equated with a mid range share price of NLG 85. Using this price as the reference point for projecting historical cash flows resulted in a 14 year MICAP and implied annual free cash flow growth rates of 4.39%, 3.78%, 3.32%, 2.92% 8.95%, 4.50%, 4.30% for the additional 8 years beyond that used in the Swisscom forecast, i.e. Years 8 to 14 inclusive.

Step 3 - Using the 3-stage model calculate 3 values

The average growth rates in free cash flow obtained from analysing the peer group were applied to the Swisscom free cash flow forecast. In other words, in addition to the value of years 1 to 7 (1998 to 2004) from the original free cash flow forecast illustrated in Table 5, values were calculated for the Years 8 to 14 inclusive (2005 to 2011) and for the period beyond Year 14, i.e. a simple perpetuity without growth. This is illustrated in Table 6.

Table 5 Free Cash Flows for Swisscom

Period 1

Year 1998 1999 2000 2001 2002 2003 2004

CHF m CHF m CHF m CHF m CHF m CHF m CHF m

Free Cash Flow 872 1630 1648 1770 1952 2047 2207

Table 6 Application of KPN Free Cash Flow Growth Rate to Swisscom

Period 2 Period 3

Cost of Capital Terminal Growth Rate Year 2005 2006 2007 2008 2009 2010 2011 CHF m CHF m CHF m CHF m CHF m CHF m CHF m Growh rate peer group 5.00% 14.31% 5.33% 4.88% 7.67% 5.26% 4.99% FCF Forecast Swisscom 2317 2649 2790 2926 3151 3316 3482 Discount Factor 0.945 0.893 0.844 0.798 0.754 0.713 0.674 PV FCF 2190 2366 2356 2335 2377 2365 2347 Cumulative PV 2190 4557 6913 9248 11625 13990 16336 Terminal Value without growth 40458

FCF Growth - Peer Group MICAP no growth

Step 4 – Analyse the sum of the values of stages 2 and 3 of the 3-stage model using the

perpetuity with growth model for each of the comparable businesses to establish what

the perpetuity growth rate would have to be to justify the free cash flows in the

perpetuity period beyond stage 1 of a 2-stage model.

In the case of Swisscom from the information shown in Table 6, the implied perpetuity

Table 7 Valuation of Swisscom

Year 1998 to 2004 Beyond Cost of Capital 6.45% 6.45% Terminal Growth Rate 3.18% Year CHF m CHF m

FCF Forecast Swisscom 2207 2207 Discount Factor 0.646 PV FCF Cumulative PV 9263 Terminal Value without growth 43602

Valuation CHF m Cumulative Present Value of Free Cash Flows 9263 Present Value of Terminal Value 43602 Business Value 52866 Marketable Securities 150 Corporate Value 53016 Less Market Value of Debt and Preference Shares 7114 Strategic Value 45902

Number of Ordinary Shares (millions) 73.55 Strategic Value Per Ordinary Share 624

*Step 5 - Apply the growth rates from Step 4 to the 2-stage valuation model and iterate

on share price as an input to the WACC calculation and the share price as an output

Iterating on share price resulted in the price of CHF 624 shown in Table 7, significantly different from the issuing price of 340.

Conclusions to be Drawn

A 5 Step Framework has been outlined and illustrated for use in challenging key assumptions in free cash flow models used for valuing IPOs. Using the information provided in the analysis of the potential IPO valuation of Jordan Telecom we have illustrated using one of the comparables cited that the estimates for perpetuity growth are inconsistent. This inconsistency is also reflected in the estimated original IPO price of JD2.48 to 2.72. The result of applying the growth rate estimated from an analysis of TPSA was an implied share price of JD1.89 and there would have to be some very substantial changes to justify the original IPO share price estimates. Furthermore, the application of the estimates from the other Central and Eastern

European peer companies resulted in much lower growth rate and share price estimates;

TPSA was the most optimistic.

In the case of Swisscom the situation was the contrary and the 5 Step Framework revealed potential under-pricing of the IPO. The valuation was heavily influenced by very volatile market conditions in the fall of 1998 which resulted in a series of readjustments of the initially evaluated price range for Swisscom shares. The lead manager's initial estimate (mid range share price of 610) of the market value of Swisscom seems to be most appropriate if analysed ex post. Setting the bookbuilding price range much lower than this estimate and then issuing at a price at the lower end of the price range is certainly a result of the willingness to issue the shares even in times of very negative market conditions. If the final price setting is challenged, the IPO price is inconsistent with the information implicitly contained in the peer ©Roger Mills 26 Henley Discussion Paper Series HCVI HDP No. 7

group's share price. There would have to be some substantial changes to the input assumptions for its justification.

The main conclusion is that Swisscom was sold to a much lower price compared to the valuation of similar telecom companies at that point in time. The negative premium (at least part of it) can certainly be explained by the growing uncertainty about future growth rates at that point in time, which, one can reasonably argue, affect a newly listed company more severely than a company with a long track record (although Swisscom's track record was not that short). Since Swisscom was by objective means, as shown in this analysis, clearly undervalued by investors (as shown by bookbuilding, price is a result of market demand and not necessarily controlled by the issuing banks), waiting and postponing the IPO would have been the better option. One has to note, that this was also ex ante the case since all available information at that time indicated a valuation of Swisscom that was too low compared to its fair value derived from comparable companies at exactly the same point in time (thus, also having suffered a price decline, see textbox 2). Even if it can be conceded that no one could know that markets would quickly recover, there is no sense in selling the company when the valuation is objectively too low. At least, there seems no obvious reason why Swisscom would have to be priced at such a discount compared to its peer group. Whereas difficult market conditions resulted in a weak retail response, a strong response from institutional investors was reported in the news after the IPO, thus indicating that more sophisticated investors most probably recognized the low valuation.

It is recognised that the selection of comparables is a key part of the proposed framework and is often a significant challenge insofar as there will typically be a good deal of debate about criteria for comparability and possibly questions about whether there really are any

comparable businesses. Acceptance of there being no real comparables leaves potentially little scope for the analysis described here, but we believe there is always some form of peer group comparison that can be undertaken. If the fundamental concern is that the growth rates selected for 2-stage models are too aggressive, we have found that selecting one of the best performing companies in the marketplace can be a useful proxy in the absence of real comparables. This is because in our experience at least this will force the right questions to be asked, i.e. do the growth assumptions make sense if compared against other high growth businesses’ recent performance? Setting an IPO price is a major challenge because of the numerous uncertainties, getting it right is nirvana and maybe not getting it too wrong is a better objective; that is exactly what this approach proposes by forcing the right questions to be asked. To summarise, the contention is that in IPO pricing the 5 Step Framework proposed provides a potential line of investigation and this is the real value of the approach – it provides a systematic framework for challenging assumptions!

Postscript

“The Jordan Telecom Company (JTC, Amman Stock Exchange code JTEL)) began trading on the Amman Stock Exchange (ASE) on 4 November 2002. During the first six months since it’s listing on the market, JTC’s share price has slipped into a definite downward trend losing a total of 14.28 percent from JOD 2.45 (closing value on 4th November 2002) to JOD 2.17 (closing value on 7th August 2003), trading habitually in shallow volumes. In the period of January - August 2003 JTC’s price reached the highest value of JOD 2.37.”

Sources: Basel Khraisheh, Amer Mouasher, Talal Touqan, Samer Sunnuqrot, Market Performance Report (3-7)

August 2003 Jordan National Bank plc., www.arabfinance.com

Textbox 1

“Swisscom began trading on the Swiss Stock Exchange (SWX) on 5 October 1998.

During the first three months since it’s listing on the market Swisscom’s share price showed a steadily upward trend reaching a temporary high of CHF 625 at the end of

January 1999. In the following month, after dipping down to CHF 450 again,

Swisscom climbed to an all time high of CHF 735 in March 2000.

Textbox 2

1998 1999 Date Sep 16 Oct 2 Oct 5 Jan 25

Publication of research report by WDR Aug 28

%change to Aug 28 Swiss Market Index -1% -23% -24% 4% British Telecom 5% -5% -7% 17% KPN -15% -24% -25% 5%

Swisscom -39% -44% -38% 2% Average peer group %changes -5% -15% -16% 11% Swisscom (adjusted prices) 579 521 513 676

Projected mid range share price 521

It can be fairly argued that the results from the research report published at the end of August

1998 had to be adjusted to the price decline in telecom stocks since that date. Table 2 shows that if the initial mid range price of CHF 610 would have been adjusted by the average price decline of the peer group a mid range share price of 512 would have resulted as bookbuilding mid range. Applying the 10.8% difference from the midpoint to the lower end of the pricing range, the minimum price would have been 465 – a still 41% higher issuing price than the observed one.

About the author

Professor Roger Mills

Roger Mills works with executives in delivering practical and effective solutions to cost management, performance measurement and value related problems. He has worked with major banks and financial institutions, as well as companies and not - for - profit organisations, both in the UK and internationally. He has analysed and applied his state of the art ideas about internal controls, decision-making, corporate governance and managerial accountability to a wide range of organisations including major banks and financial institutions. He is Professor of Accounting and Finance at Henley Management College and has been involved with major research projects, including the application and implementation of Activity Based Costing and he has published widely in books and journals.

1 Mills, R. W. (2003), “Raising Equity Finance – IPOs and the Alternatives”, Journal of General Management, Manager Update

2 Rock, K. (1986), “Why New Issues are Underpriced”, Journal of Financial Economics 15, pp 186-212.

3 Fernandez, P. (2001),“Company valuation methods. The most common errors in company valuation.”, IESE Business School

4 Kahneman, D. (2003), ‘Maps of Bounded Rationality: Psychology for Behavioral Economics’, The American Economic Review, pp.1449-1475

5 Montier, J. (2003/4), ‘Part man, part monkey’, Professional Investor, pp. 12 to 17

6 Mills, R. W. (1998), The Dynamics of Shareholder Value: The Principles and Practice of Strategic Value Analysis, Mars Business Associates Ltd

7 Arab Jordanian Investment Bank Equity Research Report (2002), Saturday September 28th, page 22.

8 Mauboussin, M.J. and Johnson, P. (1997), “Competitive Advantage Period ‘CAP’, The Neglected Value Driver”, Credit Suisse First Boston, Equity Research—Americas, January 14

9 Mills, R.W., 2005 “Assessing Growth Estimates in IPO Valuations – a Case Study”, Journal of Applied Corporate Finance, Vol. 17, No. 1, Issue Winter, Morgan Stanley Publication

10 ADR Research: Telecom Italia (1998), ABN Amro, August 14, London. Credit Suisse First Boston, Equity research: British Telecom (1998), October 13, London. Credit Suisse First Boston, Equity research: KPN (1998), July 16, London. JP Morgan Securities, Like Clockwork, Part 1: Investment Review (1998), November 2, London. JP Morgan Securities, Like Clockwork, Part 2: Business Overview and Analysis (1998), November 2, London. JP Morgan Securities, The Data Source of Swisscom (1998), November 2, London. JP Morgan Securities, The Financial Model for Swisscom (1998), November 2, London. Lehmann Brothers, Equity Research: Swisscom (1998), October 7, London. Warburg Dillon Read, Swisscom - Delivering the turnaround (1998), August 28, London.