Bad Advice

History Is Bunk

Fund A grew 15 percent a year. Fund B grew 3 percent. Don’t be a sucker and buy Fund A.

Browse the entire Bad Advice series here.

In recent months, T. Rowe Price has been bombarding the media with ads that brag: “For each 1-, 5-, and 10-year period ended 12/21/06, 70 percent of our funds beat their Lipper average.”

T. Rowe Price isn’t unusual in boasting of its funds’ past performance: Many fund-firm ads say something similar. This is not surprising given that the most persuasive attribute of a potential investment—the one that, above all else, makes people reach for their wallets—is great past performance. For many seemingly sensible reasons, investors avoid investments that have done poorly and flock to those that have done well. 

Unfortunately, as the fine print in the T. Rowe Price copy notes, past performance “cannot guarantee future results.” In fact, to put it in plainer language, past performance is nearly worthless as a predictor of future results. Any firm that argues, therefore, even indirectly, that a mutual fund will do well because it has done well is taking advantage of your natural tendency to be too impressed by the past.

Most investors have heard the “past performance” warnings before, but like other common mantras, do not heed them. Why not? Because they defy common sense. Above-average fund managers should have beaten the market, while below-average ones should have lagged it. So, all we need to do, the logic goes, is to look at some past performance—and pick a few managers who have put the market to shame.

The first of many reasons why this logic is flawed is that excellent past performance is often the result of something other than skill—namely, chance. In any given period, a random selection of stocks will beat the market about half the time. Similarly, a random selection of fund managers will beat the market about half the time (before costs). As a result, the difference between a supertalented fund manager and an average one is often as hard to discern as the difference between a .350 hitter and a .280 hitter in baseball. Over many seasons, with the help of detailed statistics, the difference is obvious. Over a few dozen at-bats, however, the hitters often look about the same.

Second, strong past performance is often the result of the temporary dominance of a particular investment style: growth stocks in the late 1990s, for example, or value stocks and small stocks from 2000 to 2006, etc. When a particular fund manager’s style is in vogue, the fund can post extraordinary returns. These returns can disappear quickly when the market environment changes, however. (If you could predict the future, you could theoretically switch from style to style, but the whole problem with stock-picking, market-timing, etc., is that most people aren’t Nostradamus.)

Third, even if you do manage to find a fund whose excellent past performance is the result of skill, something critical to the performance might soon change—leaving you with a frightfully ordinary fund (and facing a big transaction and tax hit if you decide to ditch it). For example:

  • The genius fund manager responsible for those fantastic returns might leave the firm. Why? Because he might be bored. Or annoyed that his bosses are gliding along on his coattails and eager to start his own firm. (If the S&P-beating Bill Miller ever leaves Legg Mason, everyone who bought his fund because of him will face a choice: settle for another manager or potentially take a big tax hit to follow Miller to another firm.)
  • The genius fund manager might begin to rest on his laurels, devoting the evenings and weekend time he used to spend researching stocks to researching medieval art or private-jet cabin design.
  • The fund firm, having posted fantastic returns, might be swamped with new capital, which will lead to diseconomies of scale and lower future returns. (The more money a fund manages, the harder it is to deliver superior returns—except in the case of low-cost index funds).
  • The fund firm might decide that it deserves more of the profit than it has been getting—and raise fees (which will act as a drag on future returns).

Still have doubts? Then take a look at these charts, which show the persistence in five-year performance of 800 mutual funds from 1991 to 2004. In the top half of each chart, the funds are ranked according to their five-year performance relative to the mean: good ones on the left, bad ones on the right. In the lower half, the funds are in the same order, but the bars show their performance in the second five-year period.  As you can see, good performance in the first five years was of almost no use in predicting performance in the second five years. (In fact, in the second chart, good performance presaged crappy performance—a common pattern.)

Skeptics will note that these charts were produced by Index Funds Advisors, a firm committed to indexing. This is true, but IFA developed its strategy after studying this kind of evidence—and discovering that studying the past to pick funds that will do well in the future was for the birds. 

The bottom line: Although common sense suggests that scrutinizing past performance should enable you to select funds that will perform well, it usually won’t. There are some instances in which past performance is a good predictor of future performance: with index funds, for example, or in cases in which a fund’s good results are due to low costs. In most cases, however, past performance is irrelevant.