Several readers responded to my post on estimating equilibrium returns by claiming that the concept is seriously and hopelessly flawed. But the notion that analytical techniques must be all or nothing is dangerous and more than a little impractical.

It’s widely understood that modern finance as classically outlined is flawed. The capital asset pricing model (CAPM), to cite the obvious blemish, has a tough time explaining the excess returns of small stocks and value stocks. This limitation inspires some to throw out the entire CAPM/modern finance edifice. But that’s a drastic response and arguably excessive. CAPM and modern finance are flawed, and we need to be aware of those flaws, but that’s a long way from saying there’s no value here.
Indeed, it’s clear that we live in a multi-factor world. But that’s no basis for abandoning the market factor as outlined by CAPM. Granted, we shouldn’t be slaves to it. Instead, we should consider a broader range of factors in our analysis, including CAPM’s market beta. In fact, empirical studies show that analysis that combines the market, small cap and value factors does a better job of evaluating risk and return compared with any one factor on a stand-alone basis.
What’s more, beta shows its durability when used in broad-minded portfolio concepts. It’s still not perfect, but it’s far from worthless. As one example, a recent study finds that beta does a reasonably good job of predicting risk and return over the last 40 years in real-world portfolios. In particular, high beta portfolios suffer bigger losses than low beta portfolios during so-called negative black swan events. Meanwhile, the opposite is true: high beta portfolios earn higher returns when markets rally compared with low beta portfolios.
That’s hardly an argument to use beta in isolation. In fact, no single predictor is flawless. But that’s no excuse to emphasize how any one predictor fails. Rather, the solution is to combine predictors and develop robust methodologies that can withstand the inevitable uncertainties that harass every attempt to forecast risk premiums. As a recent study reminds, this is a practical and productive way to approach the problem of managing risk when the future is forever cloudy.
How does estimating equilibrium risk premiums fit in? It’s valuable baseline for analyzing expected returns. Note that I didn’t say it’s the first and last step. Nor should this process be used without doing additional analysis using alternative approaches. In addition, estimating equilibrium returns should be a process as opposed to a one-time calculation.
Let’s say that you estimate equilibrium returns once a quarter. In time, the information will be far more valuable than any one batch of numbers. Imagine that your equilibrium return estimate generally corresponds with alternative methodologies for predicting, say, the U.S. equity risk premium. But suddenly there’s a wide divergence in the estimates. That may tell us something of immense value.
Estimated equilibrium returns (assuming they’re calculated in a reasonable manner) can be useful benchmarks of what’s available in the long run future because in the long run the average investor holds the market-value weighted portfolio. By reverse engineering the implications of this reality, we can develop strategic and tactical insights for managing money. Should we stop there? Of course not. But the process is relatively painless and serves as a foundation for additional analysis when it comes to evaluating asset allocation possibilities.
Embracing the process of routinely estimating equilibrium returns doesn’t mean we have to give up anything. But adding this tool to our analysis does open the door for enhancing risk premium predictions.