UK CORPORATE VALUATION METHODS: A SURVEY

184 APPENDIX This appendix presents the results of a survey of ex perienced corporate fi nance practitioners in the UK and Ireland condu cted in May and Ju ne 1998. Qu estionnaires w ere sent to 452 corporate fi nance directors at 71 banks, stock brokers, venture capitalists and accountancy firm s listed in C rawford s Directory. A total of 113 u sable responses w ere retu rned from 41 different fi rm s by the 14 Ju ly 1998 cu t-off date. This represents a response rate of 25.0% of indiv idu als and 57.7% of firm s surveyed. Although the survey w as conducted a num ber of years ago, its resu lts continu e to be broadly v alid, w ith the possible exceptions of the current perceived equity risk prem iu m and the u sage of the Internet to research com panies. Scoring and tables Most qu estions requ ired a response regarding the frequ ency of u se of a m ethod or sou rce of inform ation. The scale used is as follow s: 1 2 3 4 5 Alm ost Seldom Som e- Usu ally Alm ost never tim es alw ays 0 5% 6 35% 36 65% 66 95% 96 100% The tables in this report contain the m ean score, standard dev iation and the percentage of each respondent selecting 1, 2, 3, 4 or 5. Percentages m ay not add to

APPENDIX 185 100% because of rounding and, in som e cases, m ultiple responses. A.1 VALUATION METHODOLOGY There are tw o m ain approaches to determ ining the value of a com pany: relative valuation techniques; and determ ination of intrinsic or true value. UK corporate fi nance professionals u se relativ e v alu a- tion techniqu es m ost often. This m ay refl ect the ease w ith w hich, in m ost cases, com parable com panies can be found. How ever, num erous lim itations affect the reliability of sole reliance on com parable m ethods. These inclu de differences in accou nting policies; identifi cation of su itable com panies; hidden assets and liabilities; and non-ov erlapping ranges; bu t ignores sy nergies in the case of an acqu isition. Possibly the m ost im portant concern relating to relative v alu ation m ethods is w hen m ost or all com panies are trading at levels above their intrinsic value (a speculativ e bu bble). Certain acqu isitions or fl otations m ay be attractive on a relative basis, but cannot be justified on the basis of intrinsic value. Determ ination of intrinsic v alu e is generally accom - plished throu gh the application of Discou nted Cash Flow (DC F) techniques. This m ethod w as ranked sixth

186 APPENDIX in term s of frequency of use, but as one of the tw o m ost im portant m ethods as detailed in Table A.2. DCF approaches also su ffer from lim itations: identifi cation of forecast cash fl ow s can be diffi cu lt if the corporate fi nancier does not hav e access to m anagem ent forecasts and the calcu lation of an appropriate discou nt rate is su bject to debate. The ability to achiev e arithm etically precise v alu ations u sing com pu ter spreadsheets can also seduce the inexperienced into believing that DCF v alu ations are m ore accu rate. As m any financiers noted: valuation is as m uch an art as it is a science. A.1.1 Frequency of use The full results are presented in Table A.1. One lim itation of this table is that it does not recognise that certain v alu ation m ethodologies are m ore appropriate to com panies in certain sectors. For ex am ple, sev eral respondents pointed ou t that property com panies are v alu ed prim arily on the basis of appraised net asset value. The m ost frequently cited other valuation m ethod w as an enterprise v alu e/earnings Before Interest, Tax, Depreciation and Am ortisation (EBITDA) or enterprise value/ Earnings Before Interest and Tax (EBIT ) ratio. Over 10% of respondents m entioned it.

APPENDIX 187 Table A.1 Valu ation m ethodology. Rank Method Mean Standard 1 2 3 4 5 score deviation (% ) (% ) (% ) (% ) (% ) 1 Trading m ultiples 4.59 0.69 0 2 6 23 69 of com panies in the industry 2 Capitalisation of 4.34 1.02 4 4 8 25 60 forecast earnings 3 Trading m ultiples 4.21 1.04 2 8 10 28 53 of com panies taken over (exit m ultiples) 4 Price/EBIT 4.12 1.12 4 5 18 21 52 5 Capitalisation of 4.03 1.16 4 10 14 26 47 historic earnings 6 DCF 3.85 1.08 2 11 24 28 35 7 Acquisition prem ia 3.52 1.24 7 15 23 28 26 8 Industry rule of 3.07 1.12 11 14 43 20 12 thum b 9 Internal Rate of 3.01 1.24 14 19 33 20 14 Return (IRR) 10 Div idend y ield 2.87 1.22 14 27 29 19 12 11 Other 2.59 1.74 50 0 14 14 22 12 Historic book v alue 2.07 1.27 46 24 14 9 7 13 Liquidation v alue 1.93 0.89 36 41 17 5 1 14 Replacem ent cost 1.72 0.81 47 37 14 1 1 asset value 15 Real options 1.58 0.81 57 32 7 2 1

188 APPENDIX A.1.2 Calculation of final valuation or value range The second set of questions dealt w ith the w ay in w hich the respondents w eighted the resu lts of different m ethods. Ov er 70% of respondents u su ally or alm ost alw ay s placed the greatest w eight on one m ethod and used others as a check. A num ber of respondents indicated that their reliance on different m ethods depended on the nature of the assignm ent. For exam ple, one suggested that relative valuation m ethods w ere m ore im portant in fl otations and other capital m ark ets transactions, w hile DCFs w ere m ore im portant in the case of acquisitions. Finally, a num ber of respondents stated that gut feel, Table A.2 Im portance of m ethods. Method First Second Third Total choice choice choice top 3 ( ) ( ) ( ) picks (% ) (% ) (% ) (% ) DCF 29.1 23.5 25.0 27.2 Trading m ultiples 25.6 26.5 18.8 25.0 Price/EBIT 17.4 2.9 11.8 Capitalisation of forecast 14.0 14.7 18.8 14.7 earnings Capitalisation of historic 8.1 11.8 6.3 8.8 earnings Num ber of respondents.

APPENDIX 189 m arket feel or a sense of w hat w as reasonable w ere im portant influences on the final valuation or valuation range. We attem pted to determ ine w hich m ethods fi nanciers m ost relied u pon. Not all respondents prov ided an answ er to this question, nor did they all provide three responses. The DCF m ethod w as highly ranked despite its rank of sixth in the frequency of use table. The question asked w as: On w hich of the m ethodologies listed abov e in Section A do y ou ty pically place the m ost w eight? A.2 DISCOUNTED CASH FLOWS There are a num ber of DCF approaches w hich vary in their ease of calcu lation and applicability. The su rv ey ask ed respondents w hich of fou r possible approaches they used. A.2.1 DCF approaches The fi rst alternativ e discou nts the pre-interest after-tax cash fl ow s at the Weighted Av erage Cost of Capital (WAC C ) assu m ing a constant debt equ ity ratio (the WACC m ethod). This is the m ost frequ ently u sed approach in the UK w ith 63% of respondents u sing it usually or alm ost alw ays. The second approach discou nts the cash fl ow s av ailable to shareholders (i.e., after interest and principal

190 APPENDIX repaym ents) using the cost of equity as the discount rate the Equity C ash Flow (EC F) m ethod. This is typically used in situations w here the com pany s initial gearing is very high. The cost of equity is recalculated each year as the gearing lev el decreases. Econom ic profi t also referred to as residu al incom e, Econom ic Valu e Added (EVA) or Shareholder Valu e Added (SVA) is forecast and discounted at the WACC (Shareholder Valu e Method). There is no benefi t to be gained in using this m ethod for valuation its greatest use is in providing an annual snapshot of corporate econom ic perform ance. Finally, Adjusted Present Value (APV) separately values operating cash fl ow s and the tax shields prov ided by interest pay m ents u sing different discou nt rates. APV probably provides the m ost useful approach in com plex v alu ation situ ations as it disaggregates the sou rces of value. A.2.2 Forecast period The length of the cash fl ow forecast w as also inv estigated. A w ide range of responses w as recorded (from 2.5 to 20 years). Som e of the variance can be explained by the indu stry specialities of the respondent. Those cov ering u tilities and other relativ ely stable indu stries indicated a longer cash fl ow forecast than others. The m ost frequ ent response w as 5 y ears, althou gh the average w as 7.14 years and the m edian 7.0 years.

APPENDIX 191 Table A.3 DCF m ethod. Rank Method Mean Standard 1 2 3 4 5 score deviation (% ) (% ) (% ) (% ) (% ) 1 WACC 3.71 1.38 11 11 15 23 40 2 ECF 2.59 1.42 33 19 15 22 11 3 Shareholder v alue 2.34 1.29 37 20 20 17 6 4 Adjusted Present 2.04 1.18 44 25 18 8 4 Value (APV ) A.2.3 Term inal value An ex trem ely im portant com ponent of any cash fl ow based valuation is the calculation of the term inal value at the end of the explicit cash flow forecast. Depending on the length of the forecast period, the term inal value estim ate can often provide m ore than half of the total value ascribed to the com pany. The m ost frequ ent response m u ltiple of earnings or EBIT in the final year is unfortunately not theoretically correct. It m ixes an accounting m easure (w ith its lim itations) w ith an econom ic m easure of cash flow. The corporate fi nancier is m ak ing assu m ptions abou t the m ultiple w hich is achievable in the future. Interestingly, the m ethod cu rrently fav ou red by m any strategy consu ltants the hold and fade m ethod w as cited very infrequently. Hold and fade is a refinem ent of the perpetuity calculation. The m ethod recognises that a com pany s cash flow s are highly unlikely to increase in perpetuity or that it w ill be able to continue

192 APPENDIX Table A.4 Term inal value calculation. Rank Method Mean Standard 1 2 3 4 5 score deviation (% ) (% ) (% ) (% ) (% ) 1 Multiple of fi nal 3.45 1.27 13 8 21 37 21 year earnings or EBIT 2 Value of perpetuity 2.95 1.33 23 11 27 27 12 3 Value of a grow ing 2.78 1.48 29 16 20 18 18 perpetuity 4 Hold and fade 2.16 1.25 43 20 19 12 5 to earn a retu rn greater than its cost of capital indefinitely. Therefore, it suggests a period of tim e w hen the com pany continues to earn a return in excess of its cost of capital (typically up to 7 years), then gradually reducing (fading) its return on invested capital to equal its cost of capital. A.3 COST OF CAPITAL One of the m ost im portant contributors to value is the choice of the discou nt rate u sed in determ ining the present v alu e of cash fl ow s. Most debate is generated by the calculation of the cost of equity as the calculation of the cost of debt is relatively straightforw ard. A.3.1 Determ ining the cost of equity Several m ethods are available for calculating the cost of equ ity. The div idend discou nt m odel, Capital Asset

APPENDIX 193 Table A.5 Determ ining the cost of equity. Rank Method Mean Standard 1 2 3 4 5 score deviation (% ) (% ) (% ) (% ) (% ) 1 CAPM 3.67 1.43 14 7 15 23 40 2 Risk free rate 3.01 1.42 21 18 19 23 19 Assum ed risk prem ium for the particular stock 3 Div idend discount 1.99 1.03 39 34 19 6 3 m odel 4 Other 1.42 1.12 86 3 3 2 7 Pricing Model (C APM), Arbitrage Pricing Model (APM) and option approaches are the m ost com m only taught in business schools. The CAPM continues to be the m ost used in practice. Each has lim itations either w ith respect to the ease of calculation (APM) or the inability of em - pirical testing to v alidate the theory (CAPM). A.3.2 Capital asset pricing m odel Several questions w ere asked regarding the com ponents of the form u la: the risk-free rate of interest, the com - pany s beta, and the equ ity risk prem iu m (the retu rn on the equ ity m ark et in ex cess of the risk-free retu rn that is required by equity investors) K e r f r m r f The CAPM has been su bject to ex tensiv e theoretical and em pirical exam ination. These studies indicate that

194 APPENDIX CAPM (and specifically the beta factor) m ay not be the best m easure of calculating the cost of equity and that other factors such as book to m arket values and firm size m ay be better indicators of ex pected equ ity retu rns. How ev er, CAPM rem ains in w idespread u se, partly because of the w ide availability of published betas (see Table A.8) and its continu ed acceptance by practitioners as noted above. In addition, none of the alternatives has proven to be w ithout lim itations. A.3.2.1 Risk-free rate of return Responses to the question of the source of the risk-free rate of return overw helm ingly favoured m edium -/longterm governm ent bonds. Table A.6 Risk -free rate. Rank Rate used Respondents (% ) 1 Medium -/long-term gov ernm ent bonds 80.5 2 Short-term gov ernm ent bills 13.9 3 Other 5.5 A.3.2.2 Beta One of the m ain reasons for the continued w idespread u se of the CAPM is the w ide range of sou rces that pu blish betas for indiv idu al com panies. This is illu s- trated by the alm ost equ al reliance on the top three beta sou rces: London Bu siness School, Datastream and Bloom berg.

APPENDIX 195 Table A.7 Sou rces of beta. Rank Source Respondents using (% ) 1 London Business School Risk 26.8 Measurem ent Serv ice 2 Datastream 25.3 3 Bloom berg 22.1 4 Barra 10.5 5 Internal calculation 9.5 6 Other 3.7 7 Use beta of 1.0 for all 2.1 A.3.2.3 Eq uity Risk Prem ium The Equity Risk Prem ium (ERP) has been the subject of signifi cant academ ic and practitioner debate in recent y ears. Traditionally, bank ers and analy sts had relied upon historic excess returns as a proxy for future returns. It w as believed that accurate forecasts of equity m arket retu rns w ere im possible to obtain, therefore a long-ru n av erage of ex cess retu rns w as an appropriate prox y for future expected returns. Studies in both the UK and the US found that, over the long ru n (1920 1994), equ ities prov ided a retu rn of approx im ately 8.3% abov e the risk-free rate (as calculated u sing an arithm etic m ean). When the retu rns w ere calculated on the basis of a geom etric m ean (i.e., to take account of com pounding) the historic equity risk prem ium dropped to betw een 5.0% and 5.50% leading to

196 APPENDIX a debate over w hether the returns should be calculated on an arithm etic or geom etric basis. Corporate fi nance tex ts line u p on both sides of the issue. Brealey & Myers Principles of C orporate Finance advocates the use of the arithm etic m ean, w hile others (including Copeland et al., 2000) recom m end the geom etric m ean. Table A.8 indicates the m ethods used by practitioners to calcu late the ERP. The fi nal colu m n indicates the UK risk prem iu m as calcu lated by each m ethod. Table A.8 Calculation of Equity Risk Prem ium. Rank Method Respondents UK Equity (% ) Risk Prem ium 1 Forw ard look ing 52.0 5.05 2 Historic av erage (arithm etic m ean) 26.5 5.46 3 Historic av erage (geom etric m ean) 18.6 4.59 4 Other 14.7 NA Note: The total sum s to m ore than 100 because 12 respondents gave m ultiple answ ers. Table A.9 sum m arises the results of the current ERP in use in the UK. Many respondents entered a range rather than a single num ber. When this occurred, w e used the m idpoint of the range in calculating the m ean and standard dev iation.

APPENDIX 197 Table A.9 UK Equity Risk Prem ium. Country Mean Mode Standard Minim um Max im um deviation (% ) (% ) UK 52 4.87 4.0 1.68 2 10 Num ber of responses; Mode Most frequent response. SOURCES OF INFORMATION The su rv ey also requ ested respondents to indicate the pu blicly av ailable sou rces of inform ation u sed w hile undertak ing valuations. Clearly, m anagem ent forecasts and interview s w ould be very highly ranked if they had been included in the question. Published financial statem ents and brok ers reports w ere v iew ed as the m ost valuable sources of inform ation (see Table A.10). Table A.10 Sou rces of inform ation. Rank Source Median score Standard deviation 1 Annual report and accounts 4.77 0.65 2 Interim reports 4.40 0.96 3 Brok ers reports (other fi rm s ) 4.21 1.01 4 Consensus earnings estim ates 4.17 1.04 5 Brokers reports (ow n firm ) 3.97 1.47 6 Other 3.71 1.72 7 Financial press 3.40 1.28 8 Trade journals 2.96 1.27 9 Consultants reports 2.83 1.12 10 Gov ernm ent statistics 2.77 1.11 11 Internet (World Wide Web) 2.75 1.33

198 APPENDIX Tw o categories of brok ers reports w ere inclu ded: ow n fi rm and other fi rm s. The reason that other fi rm s reports ranks higher than ow n firm is a consequence of the num ber of respondents w orking for m erchant banks, corporate finance boutiques or accounting firm s w ith no research or brok ing capacity. Within the financial statem ents, the num eric presentation of the resu lts w as deem ed to be m ore im portant than the descriptiv e, narrativ e sections. Table A.11 Content of annu al report and accou nts. Rank Source Median score Standard deviation 1 Incom e statem ent/profi t and 4.84 0.46 loss 2 Balance sheet 4.76 0.62 3 Cash fl ow statem ent 4.62 0.79 4 Operating and fi nancial 3.91 1.10 rev iew 5 Chairm an s statem ent 3.41 1.38 6 Directors report 3.30 1.42