A big issue with building an alternative risk premia portfolio is whether you believe that it should be actively managed or whether it should just be a passive diversified portfolio. This is a variation of the old issue of whether there is investment skill with predicting returns. Investment skill is not just isolated to security selection but also can be applied to style rotation just like asset allocation decisions.
Passive, however, may not mean there is no investment action. Passive portfolio management may include rebalancing and resetting allocations based on a criteria such as risk parity or equal risk contribution. This passive investment is based on the assumption that investors do not have skill at determining expected return, but there is value or skill in managing volatility and covariance.
Active management would represent changing the choice of ARP styles or weights based on some set of predictive factors of return. The active question revolves around the issue of whether risk premia are time varying in a systematic fashion. To answer this question requires a classification of risk premia based on their characteristics and whether those characteristics are predictive.
Risk premia have sometimes been classified in two forms: those associated with compensation associated with the aversion to a specific risk and those that are associated with investor behavior. In the case of risk aversion, the alternative risk premia will include value, carry, and volatility. Investors are paid to hold a specific risk that may be tied to the business cycle or market risk. Those premia associated with behavior will include momentum or trend, and defensive styles. Investors are paid as compensation for what may be the “bad” behavior of investors.
There are some premia that are structural and independent of any cyclical factors and others that are time varying and have a relationship with some cyclical factors. We would expect that premia that are driven by risk aversion will vary with changes in perceived risk that needs to be compensated. Premia associated with behavior will change with the environmental uncertainty that causes biases away from rationality. Overall, some premia styles should be more predictable than others.
Some will argue that even if these factors vary over time, it is not possible to make active decisions on when to hold or exit these premia. These investors would argue for passive portfolios that are rebalanced by risk. Others will argue that there is enough information to make these allocation judgments. They would focus on rebalancing methodologies with the potential for risk premia tilts.
We believe that risk premia prediction is difficult and should be done carefully but there are periods when the odds for performance will be stacked against a specific risk premia style. Measuring and tracking this performance behavior should be rewarded.