The WACC Fallacy: The Real Effects of Using a Unique Discount Rate

• Published date: 01 June 2015
• DOI: http://doi.org/10.1111/jofi.12250
• Author: DAVID THESMAR,AUGUSTIN LANDIER,PHILIPP KRÜGER
• Date: 01 June 2015

THE JOURNAL OF FINANCE •VOL. LXX, NO. 3 •JUNE 2015

The WACC Fallacy: The Real Effects of Using

a Unique Discount Rate

PHILIPP KR ¨

UGER, AUGUSTIN LANDIER, and DAVID THESMAR∗

ABSTRACT

In this paper, we test whether firms properly adjust for risk in their capital budget-

ing decisions. If managers use a single discount rate within firms, we expect that

conglomerates underinvest (overinvest) in relatively safe (risky) divisions. We mea-

sure division relative risk as the difference between the division’s asset beta and a

firm-wide beta. We establish a robust and significant positive relationship between

division-level investment and division relative risk. Next, we measure the value loss

due to this behavior in the context of acquisitions. When the bidder’s beta is lower

than that of the target, announcement returns are significantly lower.

IN THIS PAPER,WE PROVIDE evidence that firms fail to properly adjust for risk in

their valuation of investment projects, and that such behavior leads to value-

destroying investment decisions. According to the standard textbook formula,

the value of an investment project depends on both its expected cash flows

and its discount rate, which is a measure of risk. In practice, however, survey

evidence shows that most firms use only a single discount rate to value all of

their projects (Bierman (1993), Graham and Harvey (2001)), a behavior that

we label the “WACC fallacy.” The weighted average cost of capital (WACC)

fallacy is a failure to account for project-specific risk, which is particularly

damaging when the firm has to decide between heterogeneous projects. The

value of riskier projects will be overestimated, while that of safer ones will be

underestimated.

We expect the WACC fallacy to have real effects: in relatively complex firms,

investment will be biased against safe projects, which should lead to the

∗Kr¨

uger is with the University of Geneva (Geneva School of Economics and Management

and Geneva Finance Research Institute); Landier is with the Toulouse School of Economics; and

Thesmar is with HEC Paris and CEPR. The authors greatly appreciate comments and suggestions

from Malcolm Baker, Andor Gy¨

orgy, Owen Lamont, Oliver Spalt, Masahiro Watanabe, Jeff Wur-

gler, and seminar participants at AFA,CEPR, EFA, EFMA, and NBER meetings, HBS, Mannheim

University,Geneva University, and HEC Lausanne. The authors also thank three referees, the Ed-

itor (Cambell Harvey), and an Associate Editor for their constructive suggestions. Kr¨

uger thanks

Gen`

eve Place Financi`

ere for financial support. Landier acknowledges financial support from a

Scor Chair at the JJ Laffont foundation and from the European Research Council under the Euro-

pean Community’s Seventh Framework Programme (FP7/2007-2013) Grant Agreement no. 312503

SolSys. Thesmar thanks the HEC Foundation for financial support.

DOI: 10.1111/jofi.12250

1253

1254 The Journal of Finance R

destruction of value as capital is not optimally used. The economic magni-

tude of this bias is potentially large. For example, suppose that a firm invests

in a project that pays a dollar in perpetuity. If it takes a discount rate of 10%,

the present value of the project is $10. By contrast, a rate of 8% would imply a

present value of $12.5. Hence, underestimating the discount rate by only two

percentage points leads to overestimating the project’s present value by 25%.

The present paper is a first attempt to document and measure the distortions

induced by the WACC fallacy by relying entirely on field data. To implement

our empirical tests, we focus on two types of projects: investment within con-

glomerates, and mergers and acquisitions.

First, we use business segment data to investigate the extent to which di-

versified firms rely on a firm-wide discount rate. To do so, we examine whether

diversified companies are inclined to invest less in their low-beta divisions than

in their high-beta divisions, controlling for standard determinants of invest-

ment, such as growth opportunities. The intuition is as follows. A company

using a single firm-wide discount rate will tend to overestimate the value of a

project whenever the project is riskier than the typical project of the company.

If companies apply the net present value (NPV) principle to allocate capital

across divisions,1then they will tend to overestimate the value of projects that

are riskier than the firm’s typical project and vice versa. This, in turn, should

lead to overinvestment (underinvestment) in divisions that have a beta above

(below) that of the firm’s representative project.

Using a large sample of divisions in diversified firms, we show in the first

part of the paper that investment in noncore divisions is robustly positively

related to the difference between the cost of capital of the division and that of

the most important division in the conglomerate (core division). We interpret

this finding as evidence that some firms discount investment projects from

noncore divisions using a discount rate close to the core division’s cost of capital.

We next discuss the cross-sectional determinants of this relationship and find

evidence consistent with models of bounded rationality: whenever making a

WACC mistake is more costly (e.g., the noncore division is large, the CEO has

sizable ownership, or within-conglomerate diversity in the cost of capital is

high), the measured behavior is less prevalent.

In the second part of the paper, we examine the present value loss induced

by evaluating projects using a single company-wide discount rate. To do so, we

focus on diversifying acquisitions, a particular class of investment projects that

are large, can be observed accurately, and whose value impact can be assessed

via event study methodology. We look at the market reaction to the acquisition

announcement of a bidder whose cost of capital is lower than that of the target.

If this bidder uses its own WACC to value the target, it tends to overvalue the

target, and thus announcement returns should be relatively poorer, reflecting

relatively lower shareholder value creation. We find that such behavior leads to

a relative loss of about 0.8% of the bidder’s market capitalization. On average,

1Survey evidence of CEOs and CFOs presented in Graham, Harvey, and Puri (2014) suggests

that the NPV ranking is the predominant principle governing capital budgeting decisions.

The WACC Fallacy 1255

this corresponds to about 8% of the deal value, or $16 million per deal. This

finding is robust to the inclusion of different control variables and to the use of

different specifications.

Following Stein (1996), our approach is connected with the idea that betas re-

sulting from the capital asset pricing model (CAPM) capture some dimension of

fundamental risk. On the positive side, our investment regressions show that,

irrespective of whether the CAPM holds, managers do use CAPM betas but fail

to adjust them across projects. Most corporate finance textbooks recommend

the use of CAPM betas to compute discount rates, but require that managers

use project-specific betas. Our investment regressions show that investment

in noncore divisions depends strongly on core betas. Hence, managers do use

a CAPM beta to make capital budgeting decisions, even if they use the wrong

one (core instead of noncore). On the normative side, our M&A results suggest

that using the wrong beta to value the NPV of an acquisition is value destroy-

ing. This may come as a surprise given that the CAPM fails to predict stock

returns. As shown in Stein (1996), however, this empirical failure is not in-

consistent with the normative prescriptions of textbooks. Indeed, CAPM-based

capital budgeting is value-creating if CAPM betas contain at least some infor-

mation on fundamental risk relevant for long-term investors. Our mergers and

acquisitions (M&A) results suggest that they do.

Our paper is related to several streams of research in corporate finance.

First, it contributes to the literature on the theory and practice of capital

budgeting and mergers and acquisitions. Graham and Harvey (2001) provide

survey evidence regarding firms’ capital budgeting, capital structure, and cost

of capital choices. Most relevant to our study, they show that firms tend to use

a firm-wide risk premium instead of a project-specific one when evaluating new

investment projects. Relying entirely on observed firm-level investment behav-

ior, our study is the first to test the real consequences of the finding in Graham

and Harvey (2001) that few firms use project-specific discount rates. More pre-

cisely, we provide evidence that the use of a single firm-wide discount rate

(the WACC fallacy) does in fact have statistically and economically significant

effects on capital allocation and firm value. Since we assume that managers

rely on the NPV criterion, the present paper is also related to Graham, Harvey,

and Puri (2014), who show that the NPV rule is the dominant way to allocate

capital across divisions. The same reasoning, however, also applies to firms

using an internal rate of return (IRR) criterion: if the minimum IRR required

for projects is similar across all the firm’s projects, there will be overinvest-

ment in risky projects. Thus, regardless of whether firms base decisions on

NPV or IRR, the use of a single rate leads to a similarly biased investment

policy.

Second, our paper contributes to the growing behavioral corporate finance

literature. Baker, Ruback, and Wurgler (2007) propose a taxonomy organizing

this literature around “irrational investors” versus “irrational managers.” The

more developed irrational-investors stream assumes that arbitrage is imper-

fect and that rational managers, in their corporate finance decisions, exploit