Why CAPM is useless
Beta is meaningless as a risk measure, much like standard deviation. Beta measures the sensitivity of an asset to an index, that is all it measures. If you care about relative performance than you might be content with using Beta as a risk measure, but I personally don’t care about relative performance, I only care about absolute performance. Beating the market by 2% is irrelevant to me if the market is down 20%. And believe it or not, there are many investors that hunt for absolute returns, and they handily beat the indices.
The problems with Beta come from the model it uses, the Capital Asset Pricing Model derived from Modern Portfolio theory. So what is the problem with CAPM? CAPM is designed as a theoretical model, this fact is highlighted by the restrictive assumptions that the model makes. There are other problems with the model but I will focus on two of the main assumptions that appear frivolous to me: investors are rational and risk averse, and all information is available at the same time to investors.
1- Investor rationality: This is an easy one, investors and specifically retail investors do not tend to act rationally all the time, human beings are emotional and therefore irrational, especially when it comes to money.
2- All information is available at the same time to investors: obviously this is not true and even if all information was available, many investors would not have the know-how to utilise the information.
To me these are major assumptions that are purely theoretical, and therefore in the real, practical world are not that useful. And definitely not useful when trying to make money.
Even if Beta actually measured some type of risk, from the statistical evidence it does not explain much of the return in stocks. We can see from the chart below that the CAPM’s Beta (systematic risk) does not imply a strong relationship between 'risk' and return.
Source: Fama & French (2004)
From the sample, we can see that the lowest Beta decile of 0.6 has an average return of approximately 11 percent, while the Beta decile of 1.8 has an average return of 13 percent over the period. Beta is pretty insignificant and massively overpredicts the effect of Beta.
So why do we still care about CAPM?
Investors care about CAPM because many fund managers and institutional investors still use CAPM as a determinant of risk and therefore performance. Imagine being a fund manager and your mandate is to carry a portfolio with a Beta of 0.6. Well, on average you should easily be able to beat your ‘CAPM expected’ return. For retail investors, Beta or the failings of Beta are extremely important as volatility of a stock relative to its index does not explain much of its returns. This is important because retail investors tend to be greatly emotionally affected by volatility. An emotional, quasi-rational investor trying to achieve a high rate of return would be greatly impacted by a high beta portfolio and would likely make the wrong decision at the wrong time.
The reason we still care about Beta, standard deviation, VaR, alpha and multiple others, is because those measures are numbers, and human beings find a lot of comfort in numbers. But risk cannot be entirely quantified, there is a need for business sense and skeptical thinking when investing.
Do not find too much security in numbers, it is just your mind fooling you.