A beta return can therefore be worth an alpha-like fee, typically when circumstances discourage performing the required assembly steps. As the authors say, “few people want to move to Turkey to invest in hot dog stands or negotiate with Indian farmers for the carbon credits they’ve earned.”

William Sharpe's observation that the return on any portfolio consists of a market (beta) part and a non-market (alpha) part is “perhaps the most profound insight in modern finance.” All investments, including illiquid and non-transparent alternatives, have betas and alpha, even if they are not readily apparent to the naked eye. If all you have to do is “be there” (as in the case of Turkish hot dog stands), it is beta. But if you need to identify a manager with superior skill to justify making a superior investment, then that’s a combined beta and alpha decision.

Is timing among beta exposures alpha or beta?

Should attempts to earn a return by timing beta exposures be credited to beta or alpha? Siegel and his colleagues believe it’s understandable that investors are confused on this point, since the question contains a surprisingly large number of moving parts.

At first blush it would seem that such bets should always be credited to alpha. If an investor held the S&P 500 from January 1995 to March 2000, then Treasury bonds from April 2000 to February 2003, then the S&P 500 thereafter, they would have been a genius (or extremely lucky) at market timing and would have earned a huge alpha, regardless of the benchmark against which their performance was measured.

But let’s dig a little deeper. Many international (EAFE) managers beat their benchmark in the 1990s by underweighting Japan, which, at its peak, represented almost 60% of EAFE. Siegel, Waring, and Scanlan argue that whether the return from underweighting Japan is alpha or beta depends on the manager’s intent. If the manager underweighted Japan because they thought that Japanese stocks were overpriced, then it’s alpha. If the manager underweighted Japan for risk control reasons (say, to avoid putting 60% of a portfolio in one country), or because their peers were also underweighting Japan, then the return is arguably beta – that is, it should not be regarded as evidence of manager skill. Unfortunately, it’s essentially impossible to discern manager intent looking backward over time; managers will say they intended to make the bets that had favorable payoffs. To get the information needed to distinguish alpha from beta in the current example, one would have had to keep track of the manager’s stated intentions in “live action.”

Is rebalancing alpha or beta? In a market that is in a long uptrend or downtrend, rebalancing asset-class weights to the policy mix only after they have drifted appreciably away from the policy weights can add 5% a year compared to the policy mix (which assumes continuous rebalancing). If it is just a mechanical rule to avoid overly frequent trading, this extra return is beta. (By the way, in a trendless, mean-reverting market, the same rebalancing technique produces a negative return relative to the benchmark. This, too, is beta. As evidence that it’s beta, note that the two effects cancel each other out; there is no preferred return, over both trending and mean-reverting markets, from the strategy. Stated another way, unless you know whether the markets are trending or not, in advance, the expected return of your rebalancing strategy is simply zero.)

The authors do not provide definitive guidance as to whether a beta-timing return is really beta or alpha. When viewed after the fact, excess returns result either from intended, skillful “beta bets” or from incidental, skill-free ones just as with any other active management decision. It’s similarly hard to distinguish one kind of bet from the other. Siegel and his co-authors do not suggest there is a “bright line,” but aim to raise awareness that returns from timing among beta exposures can produce returns that are fairly credited to either alpha or beta, depending on the circumstances. Looking “under the hood” is necessary to have any chance at telling the difference. But, as the authors say, one thing is sure, “if you intentionally make beta timing bets, it’s about the search for alpha, not for beta.”

Display article as PDF for printing.

Would you like to send this article to a friend?