What are your risks?
Would you be willing to trade some average return in order to make sure that your portfolio does well in particular circumstances? For example, an investor who owns a small company would not want his investment portfolio to do poorly at the same time that his industry suffers a downturn, that there is a recession, or a credit crunch, or that the industries he sells to suffer a downturn. Thus, it makes good sense for him to avoid stocks in the same industry or downstream industries, or stocks that are particularly sensitive to recessions or credit crunches, or even to short them if possible. This strategy would make sense even if these stocks give high average returns, like the value portfolios. Similarly, he should avoid high yield bonds that will all do badly in a credit crunch. If the company will do poorly in response to increases in interest rates, oil prices or similar events, and if the company does not hedge these risks, then the investor should take positions in interest rate sensitive or oil-price sensitive securities to offset those risks as well. We’re just extending the principles behind fire and casualty insurance to investment portfolios.
This logic extends beyond the kind of factors (size, book to market, and so on) that have attracted academic attention. It applies to any identifiable movement in asset portfolios. For example, industry portfolios are not badly explained by the CAPM, as they all seem to have about the same average return. Therefore, they do not show up in multifactor models. However, shorting your industry portfolio protects you against the risks of your occupation. In fact, factors that do not carry unusual risk premiums are even better opportunities than the priced factors that attract attention, since you buy insurance at zero premium. This was always true, even in the CAPM, unpredictable return view. I think that the experience with multifactor models just increases our awareness of how important this issue is.
What are not your risks?
Next, figure out what risks you do not face, but that give rise to an average return premium in the market because most other investors do face these risks. For example, an investor who has no other source of income beyond his investment portfolio does not particularly care about recessions. Therefore, he should buy extra amounts of recessionsensitive stocks, value stocks, high yield bonds, etc., if these strategies carry a credible high average return. This action works just like selling insurance, in return for a premium. This is the type of investor for whom all the portfolio advice is well worked out.
In my opinion, too many investors think they are in this class. The extra factors and time-varying returns would not be there (and will quickly disappear in the future) if lots of people were willing and able to take them. The presence of multiple factors wakes us up to the possibility that we, like the average investor, may be exposed to extra risks, possibly without realizing it.