’Tis the New Year, and the forecasters have been busy. I talk with advisors often, and it’s very common for me to be asked what I think the stock market will do over the year. Jokingly, but with all sincerity, I have become well-quoted: My forecast for the year’s stock market is that the standard deviation will be about 20%. I go on to predict that I am 95% confident with a forecast return of somewhere between –30% and +50%. This probably won’t shock you, but the ratio of nodding heads to blank stares is pretty low and the sound of “crickets” prevails. But why?

I am always surprised by the number of people willing to go on record with their best guesses for future stock returns. Then I’m reminded of a comment by Daniel Kahneman, who said that “most successful pundits are selected for being opinionated, because it’s interesting, and the penalties for incorrect predictions are negligible. You can make predictions, and a year later people won’t remember them.”¹ Few, though, bother to grade their own work.

In a way, this approach to forecasting is similar to New Year’s resolutions: Too often they are rarely achieved and quickly forgotten, mainly because they are unrealistic or impractical. Some pundits are so confident about their return forecasts that they include decimals, as if this extra degree of “precision” increases the chances of accuracy. This seems to be moving forecasts in the wrong direction, with more emphasis on making return estimates “interesting,” as Kahneman mentioned, rather than useful. Losing 50 pounds, quitting smoking cold turkey, or expecting the stock market to return 7.92%, all of these prospects seem overly optimistic and not likely to be attained.

Take a simple example, based on history. From 1926 to 2013, U.S. stocks returned 10.2% per year, on average. It is not uncommon for these long-run returns to influence pundits and result in a pretty narrow range of forecasts, which may be their way of trying to tilt the odds in their favor, hoping that the market delivers an “average” return. As a result, any forecast of a 20% gain or a 10% loss over this year probably seems pretty extreme, while a forecast of between a 20% gain and a 10% loss might sound simply useless. Up to a point, forecasters might even be right on this last view. But while +20% to –10% is a very large range by common forecasting standards, the stock market’s annual return fell outside this range in just more than half (45) of the last 88 years. If a 30-percentage-point range was only correct about half the time, I’m not sure I see the benefit in trying to narrow the return estimate to a single-point estimate. While it might make the forecast more interesting, it is highly unlikely to make it more useful.


I would like to thank Don Bennyhoff for his contributions to this blog.


1 Daniel Kahneman, 2011. Thinking, fast and slow. New York: Farrar, Straus and Giroux.


All investing is subject to risk, including the possible loss of the money you invest.

Past performance is no guarantee of future results.