Microsoft PowerPoint - PoF.ppt

Critiqueƒ 64

Testability of the CAPM: Roll’s critique

ƒ

If the portfolio that is used for the comput

ation of beta lies on the minimum-variance

set, then the expected retu

rns have to lie on the SML

Ä

not necessary that we use

the efficient market portfolio for the co

mputation of betas (see the proof of the

SML).

ƒ

The CAPM states that the market portfoli

o is efficient. Since

it is impossible to

observe the market portfolio this

hypothesis cannot be tested.

ƒ

Empirical relevance of the CAPM

Despite Roll’s critique

there are still attempts to empirically verify

ƒ

whether expected returns are a linear function

of their betas with a market index and

ƒ

whether market betas are sufficient to explain the variation of the expected returns.

ƒ

Fama and French (1992): beta is not even significant!

ƒ

Fama and French (1992): firm size (significant and negative), book-to-market ratio (significant and positive), leverage ratio (significant and either negative or positive), and earnings to price ratio (not significan

t if firm size and book-to-market equity are

already included).

ƒ

Fama and French (1993): returns are not only determined by one factor but by more.

Single-period random cash flows: CAPM