At last week’s Society of Actuaries annual meeting I moderated several sessions to which I have alluded in this space before.
One of the best received presentations was by Lubos Pastor from the University of Chicago Booth School of Business.
Pastor, a graduate of Wharton and student of Jeremy Siegel, presented a wake-up call for investors who have faith in Siegel’s “stocks for the long run.” Pastor, working with Wharton’s Robert Stambaugh, has clarified the difference between long-run volatility for stocks when you know the expected return and when you don’t.
In essence, the stocks for the long run idea depends on a known distribution of return. When evaluating sample data, as Siegel has done, the inference is indisputable: the dispersion of returns shrinks as the time period lengthens. Pastor refers to this as the conditional variance, and acknowledges that, due to mean reversion, it tends to shrink over longer time horizons.
However, as investors, we’re concerned about the future, not the past. So we’re concerned not only with the conditional variance but with the predictive variance of return. The predictive variance of return consists of the conditional variance PLUS the variance of the conditional mean return. Because of the uncertainty of this second component, it expands over longer time horizons, more than offsetting the shrinkage of the conditional variance.
Implications
Conventional Equity Allocation Recommendations are Probably Too High
The implication is that the uncertainty of future equity returns should lower the allocation in any risk-averse investor’s portfolio. For investors who are sophisticated enough to grasp this, target-date funds’ typical glide paths probably allocate too much to equity. For unsophisticated investors who might otherwise not allocate to equities at all, Pastor concedes that most target-date funds are probably decent investment choices for them.
The Same Could be True for Other Risky Assets
A terrific question from the audience was whether this would apply to bonds. Pastor said he and Stambaugh have only examined stock data for this paper, however it makes sense that the long-run volatility expanding due to uncertainty should apply to all risky assets, not just equities.
Here’s a link to Professor Pastor’s presentation.
Here’s a link to his paper, co-authored with Robert Stambaugh, Are Stocks Really Less Volatile in the Long Run?
Comments
2 responses to “An Inconvenient Truth about Stocks for the Long Run”
The paper is somewhat technical but well worth a read.
[…] Jacquier, Kane, and Marcus’ work modeling expected returns over future time horizons, and Pastor and Stambaugh’s work demonstrating how to think about predictive variance. These topics all relate to the fact that entering a set of risk and return assumptions will only […]