Chapter 2:Investors and Risk Mgmt

[[email protected]]

Hi
Rene Stulz says "small firms earn more on average than predicted by CAPM".
The reason is investors require a risk premium to bear other risk other than CAPM's systematic risk ...
can you kindly explain the reason behind it....
why this is only for small firms and not large firms....

 

[JalilaZ]

Hi,

Small firms are riskier than larger firms that's why investors ask for more compensation than those of larger firms. CAPM doesn't account for this premium

 

[[email protected]]

To re-iterate what Jay_z said, small firms are not diversified - geographically as well as operationally.... Thus, a local economic crisis or a decreasing customer group can lead to an easy default... On the other hand, large firms are diversified, thus losses in one of the business lines in one geography is offset by the gain on another business line in another geography, escaping a possible default.... I think the extra premium is due to the default risk which is not captured by CAPM.. Someone can correct me if I'm wrong...

Alan

 

[David Harper CFA FRM]

Stulz almost certainly refers to the Fama-French study, here is a short 2-page summary @ http://db.tt/ccuDHC9 To my. knowledge, the findings were strictly statistical/empirical; i.e., it was first an mere observation in the data without an underlying fundamental narrative, then narratives sought after (I think).

The new FRM chapters on CAPM (Elton, Gruber et al) are from their book with much (FRM) unassigned detail. I copied snippets here on this so-called size effect, emphasis mine (note there is a big theory difference between 1. a correct CAPM model but where the beta is mis-measured for small firms; and 2. a CAPM that incorrectly omits a small firm risk factor ...). Note, the possible explanations are consistent with both Jay's and Alan's answers.

"The size effect was the first of the firm variables that was shown to be related to excess return; there has been extensive research into possible explanations. One research avenue has been to hypothesize that the CAPM was inappropriately measured causing apparent excess returns. The argument is that the Betas estimated for small firms were too low. If Beta is too low, then the estimate of expected return using the CAPM is too low and the difference between actual return and expected return would be positive even if it was zero when expected return was correctly estimated. Two reasons have been offered for why estimated Betas are too low for small firms. Roll (1970) and Reinganum (1981) have shown that the Beta for small firms will be biased downward because they trade less often than large firms and nonsynchronous trading leads to an underestimate of Beta. Christie and Hertzel (1981) present a second reason why Beta might be downward biased. Beta is measured using historical returns. Firms that become small have changed their economic characteristics; these changes mean they are riskier and Beta measured over a prior period doesn’t capture this increased risk. These factors could partially explain the relationship of excess return to size.

Chan and Chen (1991) argue that the reason small firms are riskier is that they have low production efficiency and high leverage, and are in their terms “marginal firms” with lower probability of surviving economic hard times. They point out that size is serving as a proxy for this more fundamental risk.

Another reason why the CAPM may misestimate expected return was studied by Amihud and Mendleson (1991). They reason that investors should demand a higher expected return for less-liquid stocks since trading them involves higher transaction costs. Empirically small stocks have higher bid-ask spreads and the price impact of larger purchases would be considerable for small stocks. Thus they show the small stock effect is in part compensation for illiquidity.

Finally, a number of researchers have argued that transaction costs are very high in small stocks, so that markets are still efficient with substantial excess returns on small stocks. First, Roll (1983) and Blume and Stambaugh (1983) have estimated that the magnitude of the small firm effect is cut in half if small stock portfolios are reformed annually rather than rebalanced daily as assumed by a number of authors. If the reader wonders why not simply buy small firms and rebalance daily, the answer is that large transaction costs would be incurred. Second, a number of authors have estimated transaction costs for small stocks and then argued that the excess return is eliminated or at least reduced if realistic transaction costs are taken into account."

Source: Modern Portfolio Theory and Investment Analysis, 8th Edition. John Wiley & Sons
11/16/2009. p. 417

Hope that helps, David