But he goes on to explain why clients can depend on their projected income: “an investor’s portfolio will deliver a lifelong mean return consistent with the standard deviation for the given subset of equities…” Will deliver??? This is exactly the kind of deeply held belief that I fear has been created by the quarter-century Golden Age. It’s based on what statisticians call “the fallacy of reversion to the mean” which I understand to be a fancy way of saying “you’re driving by looking in rearview mirror.”
Consider my golf game. Let’s say my “mean” score over the past 20 years is 77. Then by DT’s reasoning, we can predict that while my scores over the next ten years will range from the 80s to the low 70s, they will average out to that 77 mean. What this doesn’t consider is that significant factors have changed. For one thing, I’m getting older–I don’t hit the ball as far, as straight, or putt as well as I used to–so I’m not the same golfer who averaged 77. What’s more, I had knee and shoulder injuries a couple years back, which still affect my swing. And the greens on my course have gotten bumpier, and are harder to putt. Considering all this, I’d be lucky to average 80 over the next 10 years (at which time my “mean” score will have risen to 79 or so).
Point is, financial performance numbers that are averaged into an asset class mean all come from the past, holding little information about the future. If nothing changes, then, maybe they’d have some predictive value. But when does nothing change? And things in the financial/economic world seem to have changed rather dramatically of late. Seems to me, rather than blind faith in reversion to an historic mean, advisors might want to consider whether and how much things have changed, and what effect that might have on returns–and means–in the future.