Abstract
Discounted cash flow, method of comparables, and fundamental analysis typically yield discrepant valuation estimates. Moreover, the valuation estimates typically disagree with market price. Can one form a superior valuation estimate by averaging over the individual estimates, including market price? This article suggests a Bayesian framework for combining two or more estimates into a superior valuation estimate. The framework justifies the common practice of averaging over several estimates to arrive at a final point estimate.
Similar content being viewed by others
Notes
Yoo (2006) empirically examined, in a large-scale study of Compustat firms, whether taking a linear combination of several univariate method of comparables estimates would achieve more precise valuation estimates than any comparables estimate alone. He concluded that the forward price-earnings ratio essentially beats the trailing ratios, or combination of trailing ratios, so much that combining it with other benchmark estimates does not help. Yoo did not consider DCF, liquidation value, or market price, in his study.
The author’s equity-valuation students enrolled in an MBA elective at a leading business school like to take weighted averages of their valuation estimates even though nothing in the class materials advises them to combine estimates. When asked why they do this, students report they learned to do this at work, other classes, or that weighing makes common sense. The students do not articulate how they determine their weights, and the weights vary greatly between students.
Net asset value is the liquidation value of assets minus the fair or settlement value of liabilities. Net asset value is not the same as accounting “book” value. Book value, because it derives from historical cost accounting, arcane accounting depreciation formulas, and accounting rules that forbid the recognition of valuable intangible assets, is known to be an unreliable (typically conservatively biased) estimate of net asset value.
The Internal Revenue Code addresses the valuation of closely held securities in Section 2031(b). The standard of value is “fair market value,” the price at which willing buyers or sellers with reasonable knowledge of the facts are willing to transact. Revenue Ruling 59–60 (1959-1 C.B. 237) sets forth the IRS’s interpretation of IRC Section 2031(b).
For example, a large block trade of micro-cap shares may cause temporary price volatility. Non-synchronous trading may cause apparent excessive price stability, as seen in emerging equity markets.
Y ∼ N(m,τ) means Y is normally distributed with mean m and standard deviation τ. Note that we leave the possibility that the standard deviations may be heteroskedatic—nothing in our model requires them to be independent of V.
Accounting treatments like mark-to-market accounting or cookie-jar accounting may correlate accounting numbers to market noise. If so, then DCF estimates which rely on accounting numbers to make their cash flow projections may correlate to market noise, so that corr(e I ,e) ≠ 0. For this reason, the Appendix provides the generalization of Theorem 1 to the case when corr(e I ,e) = ρ I ≠ 0.
Bell v. Kirby Lumber Corp., Del. Supr., 413 A.2d 137 (1980).
For comparison, a random number homogeneously distributed between 0 and 1 has average value 50% and standard deviation 28.9%. Since the market value and earnings value weights have significantly smaller spread and their means differ from 50%, it does not appear these weights are homogeneously distributed random variables.
References
Armstrong J (1989) Combining forecasts: The end of the beginning or the beginning of the end? Int J Forecast 5:585–588
Arnott R, Hsu J, Moore P (2005) Fundamental indexation. Financ Analysts J 61(2):83–99
Bates J, Granger C (1969) The combination of forecasts. Oper Res Q 20:451–468
Beatty R, Riffe S, Thompson R (1999) The method of comparables and tax court valuations of private firms: an empirical investigation. Accoun Horiz 13(3):177–199
Black F (1986) Noise. J Finance 16(3):529–542
Black F, Litterman R (1991) Asset allocation: combining investor views with market equilibrium. J Fixed Income 1(2):7–18
Black F, Litterman R (1992) Global portfolio optimization. Financ Analysts J 28–43
Bhojraj S, Lee C (2003) Who is my peer? A valuation-based approach to the selection of comparable firms. J Account Res 41(5):745–774
Chan L, Lakonishok J (2004) Value and growth investing: review and update. Financ Analysts J 60(1):71–87
Chen F, Yee K, Yoo Y (2007) Did adoption of forward-looking methods improve valuation accuracy in shareholder litigation? J Account Audit Finance, forthcoming
Clemen R (1989) Combining forecasts: a review and annotated bibliography. Int J Forecast 5:559–583
Clemen R, Winkler R (1986) Combining economic forecasts. J Econ Bus Stat 4:39–46
Copeland T, Koller T, Murrin J (2000) Valuation: measuring and managing the value of companies, 3rd ed. McKinsey & Company
Coulson N, Robins R (1993) Forecast combination in a dynamic setting. J Forecast 12:63–67
Damodaran A (2002) Investment valuation: tools and techniques for determining the value of any asset, 2nd edn. John Wiley & Sons Inc, NY
DeAngelo L (1990) Equity valuation and corporate control. Account Rev 65(1):93–112
Diebold F (1988) Serial correlation and the combination of forecasts. J Bus Econ Stat 6:105–111
Diebold F, Pauly P (1987) Structural change and the combination of forecasts. J Forecast 6:21–40
Diebold F, Lopez J (1996) Forecast evaluation and combination. In: Maddala GS, Rao CR (eds) Handbook of statistics. Amsterdam, North Holland
English J (2001) Applied equity analysis. McGraw Hill, New York
Gilson S, Hotchkiss E, Ruback R (2000) Valuation of bankrupt firms. Rev Financ Stud 13(1):43–74
Granger C, Newbold P (1973) Some comments on the evaluation of economic forecasts. Appl Econ 5:35–47
Granger C, Ramanathan R (1984) Improved methods of forecasting. J Forecast 3:197–204
Grinold R (1989) The fundamental law of active management. J Portf Manage 15(3):30–37
Kaplan S, Ruback R (1995) The valuation of cash flow forecasts: An Empirical Analysis. J Finan 50(4):1059–1093
Lie E, Lie H (2002) Multiples used to estimate corporate value. Financ Analysts J 58(2):44–53
Liu D, Nissim D, Thomas J (2001) Equity valuation using multiples. J Account Res 40:135–172
Penman S (2001) Financial statement analysis and security valuation. McGraw-Hill, Boston
Perold A (2007) Fundamentally flawed indexing. Harvard Business School working paper
Seligman J (1984) Reappraising the appraisal remedy. George Washington Law Rev 52:829
Sharfman K (2003) Valuation averaging: A new procedure for resolving valuation disputes. Minn Law Rev 88(2):357–383
Siegel J (2006) The noisy market hypothesis. Wall St J A14
Treynor J, Black F (1973) How to use security analysis to improve portfolio selection. J Bus 46:66–86
Yee K (2002) Judicial valuation and the rise of DCF. Public Fund Dig 76:76–84
Yee K (2004) Perspectives: combining value estimates to increase accuracy. Financ Analysts J 60(4):23–28
Yee K (2005) Aggregation, dividend irrelevancy, and earnings-value relations. Contemp Account Res 22(2):453–480
Yee K (2007) Using accounting information for consumption planning and equity valuation. Rev Account Stud 12(2–3):227–256
Yoo Y (2006) The valuation accuracy of equity valuation using a combination of multiples. Rev Account Finance 5(2):108–123
Author information
Authors and Affiliations
Corresponding author
Additional information
I thank Francis Diebold for a useful discussion about statistical research on combining forecasts, and Jennifer Hersch for editorial assistance. The statements and opinions expressed in this article are those of the author as of the date of the article and do not necessarily represent the views of Bank of New York Mellon Corporation, Mellon Capital Management, or any of their affiliates. This article does not offer investment advice
Appendix: Proof of Theorem 1
Appendix: Proof of Theorem 1
This section establishes that the specific choice of weights κ I and κ C in Theorem 1 maximizes the precision of \(\mathop{V}\limits^{\frown}\) or, equivalently, minimizes var[\(\mathop{V}\limits^{\frown}\)]. To this end, consider an arbitrary linear weighing scheme \(\mathop{V}\limits^{\frown}=\kappa _P P+\kappa _I V_I +\kappa _C V_C,\) where κ P , κ I , and κ C are yet to be determined real numbers. Plugging in V = P + e, V I = V + e I , and V C = V + e C and requiring \(E[\mathop{V}\limits^{\frown}]=V\) immediately yields κ P = 1−κ I −κ C , which implies \(\mathop{V}\limits^{\frown}=V+(-1+\kappa _I +\kappa _C )e+\kappa _I e_I +\kappa _C e_C.\) Computing the variance of this expression yields var\([\mathop{V}\limits^{\frown}]=(1-\kappa _I -\kappa _C )^{2}\sigma ^{2}+\kappa _I^2 \sigma _I^2 +\kappa _C^2 \sigma _C^2-2(1-\kappa _I-\kappa _C )\kappa _C \rho \sigma \sigma _C.\) The first order conditions for minimizing var\([\mathop{V}\limits^{\frown}]\)with respect to κ I and κ C may be written in matrix form as
Inverting the matrix recovers the expressions for κ I and κ C given in the Theorem. These weights minimize var\([\mathop{V}\limits^{\frown}],\) which means they maximize \((\hbox{var}[\mathop{V}\limits^{\frown}])^{-1}\), the precision of \(\mathop{V}\limits^{\frown}.\) ■
1.1 Generalization of Theorem 1
Theorem 1 assumes that error e l from the intrinsic valuation estimate is uncorrelated to market noise e. If an analyst believes that corr(e I ,e) = ρ I = ρ I ≠ 0, he or she should replace the weights in Theorem 1 with the following formulas:
and
Proof
Same logic as the proof of Theorem 1. When corr(e I ,e) = ρ I , the new first order optimization condition becomes
Inverting the matrix recovers the expressions for κ I and κ C . These weights maximize the precision of the triangulation estimate \(\mathop{V}\limits^{\frown}.\)
Rights and permissions
About this article
Cite this article
Yee, K.K. A Bayesian framework for combining valuation estimates. Rev Quant Finan Acc 30, 339–354 (2008). https://doi.org/10.1007/s11156-007-0055-6
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11156-007-0055-6