JEAN-PAUL PAQUIN, ANNICK LAMBERT AND ALAIN CHARBONNEAU 281risk, as measured by beta, contributed significantly less than did total risk
to the explanation of securities’ required rates of return.
From a different perspective, theoretical considerations from the CAPM
would also imply that companies in the same sector should face similar
systematic risks and, therefore, should apply similar discount rates when
assessing their investment projects. However, many surveys have found
high variation in the hurdle rates used within same companies and same
industry sectors; this would indicate that hurdle rates are not related directly
to the cost of capital as prescribed by the CAPM. For instance, Poterba and
Summers (1995) found from a sample of 228 companies (out of all Fortune
1000 companies) that only 12 percent of the hurdle rates total variance could
be explained by the industry sector. A simple linear regression of the hurdle
rates of firms on their respective beta coefficient revealed that the beta was
not statistically significant in explaining variations in the hurdle rates. Their
results also indicated that depending on the project, hurdle rates varied
substantially within a same company, strategic projects having much lower
hurdlerates.AccordingtoJagannathanandMeier(2002), firmsgenerallyuse
much higher hurdle rates than the CAPM prescribed cost of capital. What
is rationed is not financial capital, as usually assumed by a MM frictionless
and transparent economy, but managerial talent and organisational capital.
They demonstrate that when the firm has substantial real options, the project
selection decision will be optimal as long as the hurdle rate is sufficiently
high.
Finally, Roll’s 1980 article casts serious doubts on the very empirical
testability of the CAPM. Roll’s critique, which has not yet been rebutted,
is even more devastating to the CAPM for it brings us back to Popper’s
Falsifiable Principle (1934) according to which scientific theories should
always lead to propositions that are potentially falsifiable by experimental
observation.
Nevertheless, until a better theory is proposed, the CAPM will continue
to exercise a dominant role in the theory of modern finance. The same can be
said about the Modigliani–Miller perfect markets assumptions. Replacing
these by more realistic assumptions has resulted in capital investment rules
for each possible change in assumptions; each new rule not being sufficiently
compelling to justify the replacement of the perfect market assumption.
The unique decision criterion obtained under the perfect market assump-
tion thus becomes a simplifying and unifying concept for teaching capital
budgeting.
The Modigliani–Miller paradigm with its no default risk assumption
might explain why the CAPM proponents have adopted equity return vari-
ance as a measure of risk. Again, in view of the fact that under the CAPM no
default risk is assumed for any security and that only undiversifiable mean
volatility of stock returns is relevant in determining the security’s required
rate of return, then the only relevant measure of risk is provided by the