Unformatted text preview: 9 2.53
1.35 R(r) - RF(r) = u + ~[RM(I) - RF(r)] + ($0
0.21 - O.YO
- 0.4Y — The market is ineﬃcient —investors are truly not
that smart— and we need a new model for market
— A more general model of risk and return is needed,
true. But we do not need to go as far as calling investors stupid. Multi-factor models are very
— Tests with historical data are really useless because covariances are supposed to be forward
looking. I think this would make almost 50% of statistics
useless so it’s a little bit strong for me. — (Roll’s critique) The market portfolio is unobservable, so we are not really testing the CAPM, but
the quality of a proxy such as a stock index. Failure in these tests simply tells us that the proxy is
bad, but nothing about the CAPM itself. 4 Is the CAPM used in practice?
• Does the empirical failure invalidate the CAPM as a
benchmark for capital budgeting? • Not necessarily. Why?
— Good intuition.
— It gives "reasonable" estimates of your opportunity cost....
View Full Document
- Fall '08