The real lesson from failed forecasts

Posted by on January 15, 2014 @ 9:00 am in Investing

Every year about this time, I bask in the warm glow of schadenfreude as I read stories such as the Wall Street Journal’s “Why market forecasts are so bad.” (Subscription required.)

“On average, the analysts thought the S&P 500 would rise 8.2% in 2013,” writes reporter Joe Light. “The S&P 500 has actually risen 27.5% this year through Friday, not including dividends—a difference of 19.3 percentage points.” (The Journal article was published on December 20, 2013. By year-end, the index had produced a 12-month return of 32.4%, including dividends.)

How can highly trained experts with vast research resources at their disposal be so … wrong?

Although I chuckle at the forecasters’ foibles, it’s not really the forecasters who are wrong. It’s the belief that a one-year stock market forecast, whether from Warren Buffett or the proverbial shoeshine boy, can convey useful information.

My 2014 outlook

A good forecast includes both an expected value and a range of potential outcomes. The use of a range “treats the future with the humility it deserves,” as Vanguard Chief Economist Joe Davis likes to say. The size of this range depends on the variability of what you’re forecasting and the degree of confidence you demand in your forecast.

In the stock market, these inputs produce a range of outcomes that seems more like a cowardly dodge than a forecast, the work of an ignorant amateur. To say that the stock market could follow a number of wildly divergent paths simply sounds less authoritative than a bold and unequivocal forecast for, say, an 8% return.

Since 1926, the S&P 500 Index has produced, on average, one-year returns of 12%.* The range of those returns has been wide—a one-year low of –43% in 1931 and a one-year high of 54% in 1933. If we assume that these patterns can serve as a guide to the future, we can use the historical data to forecast a stock market return for 2014, confident that our forecast will capture 95% of the potential outcomes.

Ladies and gentlemen, here it is: In 2014, the S&P 500 Index will return between –27% and 51%.**

You’re welcome.

Taking the long view

The high variability associated with short-term returns helps explain why our Investment Strategy Group, which publishes Vanguard’s Economic and investment outlook, doesn’t make one-year forecasts.

Instead, our economists and analysts produce 10-year forecasts, assigning probabilities to the different outcomes. Over longer periods, asset class returns become less volatile, allowing us to forecast a range of outcomes narrow enough to be useful in long-term asset allocation decisions.

Vanguard’s longer-term focus reinforces another lesson. Investing, particularly in the stock market, is a long-term undertaking, and one-year forecasts are the wrong tool for the task.

* The one-year average return is different from annualized compound returns, which are lower and more relevant for long-term financial planning.

** My tongue-in-cheek forecast reflects insights I first heard from Vanguard investment strategist Fran Kinniry.


Get every new post delivered to your Inbox.

Join 355 other followers