In CFAI Quant on page 260, the following regression model is presented: (Ri - Rf) = alpha + beta*(Rm - Rf) + error term Now, recall from equity that alpha is equal to expected return - required return (where required return is calculated by something like CAPM), If alpha is equal to 0, then we can conclude that the stock isn’t earning abnormal profits. If we plot the excess return of a stock over the risk-free rate (Ri - Rf) against the excess return of the S&P500 (Rm - Ri), we can find beta from the slope. ----------------- In the case of the of the Market Model in portfolio management (page 369), the equation is Ri = alpha + beta*Rm + error term In Schweser at that same section (page 228), they mention that they refer to quant LOS11.d where they "ran a regression with individual stock returns as the dependent variable and market returns as the independent variable) Isn’t this incorrect? They used _excess_ returns as I had mentioned at the top. Also, wouldn’t I get different betas using this model? Which one should be used?
Yeah, they’re definitely two different regression models and would generate different betas. For example, a beta of 1.5 in the equity risk premium model would mean you expect to get the risk free plus 1.5 times the market return over the risk free rate. Whereas a beta of 1.5 in the market model would mean you expect to earn 1.5x the absolute market return. These would def produce different estimates. The alphas would likely be different too.
janynoname, so which is used in practice? I took a Wall Streep Prep course and we regressed the stock return against the s&p500, no excess calculation. It seems the market model explained in CFAI doesn’t go to explain the difference?
I don’t know what the PMs use, but I think the absolute returns model is actually more common. That’s what we use in litigation, although the Rm is replaced with returns on comparable company indexes instead of the S&P. The absolute returns model would be more useful for picking up correlations between securities for diversification purposes as well.