Finding the Next Great Mutual Fund Manager
Updated: September 1, 2015
- Identifying excellent asset managers requires us to distinguish the “good” from the “lucky.”
- We can identify “good” managers as those with
- Long streaks of very good results; and / or
- Longer periods of higher cumulative advantage.
- Most assessments of manager performance use time periods too short and / or advantages too small to distinguish luck from skill.
- Once we have identified a skillful manager
- We cannot be certain that the manager’s skill will persist;
- It takes a long time and very poor results to change our minds.
Many people are looking for the next Peter Lynch. This stock-picking superhero beat the S&P 500 by 13% per year for 13 years. Who wouldn’t want to own a fund that could deliver performance like the Magellan Fund of 1977-1990?
People pick mutual funds hoping for excellent performance. They buy funds because they believe the manager is “good” for any of a number of reasons – their broker says so, the fund company has a good reputation, they’ve done a screen on Morningstar…
What is a “good” manager? Most people mean a manager they can count on to produce better results than other managers. We’ll say a manager is “good” if his or her fund is likely to perform better than the market.
How can you tell if a manager is likely to beat the market? Most people would want to see evidence that the manager had beaten the market before. The natural question is, “by how much, and for how long?1”
If we recognize that good managers can be unlucky sometimes, we also must allow that mediocre or even bad managers can be lucky at other times. Sometimes those mediocre managers will even beat the market! How can we distinguish a lucky mediocre manager from a truly good one?
Unfortunately, in short periods, we can’t. Over longer times, say years, it gets easier. It is progressively more difficult just to be lucky both for longer times and by larger amounts.
For example, as I write this, my beloved Boston Red Sox are in last place. But it’s early in the season and I just know that by playoff time their skill will tell, and they‘ll be in the thick of the pennant race. In contrast, I’m sure the currently first-place Baltimore Orioles have been lucky to win so many games. Of course, if there is no change in the standings by the Fourth of July, I’ll have to modify my assessment. The Orioles will have proven themselves.
Similarly, great artists demonstrate their quality by the lengths of their careers, and their numbers of hit releases, hit performances, book sales, etc. The “flash in the pan” isn’t skilled, just lucky.
Statistics provide quantitative measures of luck and skill. We can compare the number of outperforming funds with how many lucky mutual funds should have outperformed the market by how much, and for how long.
If more funds beat the market than we expect based on sheer luck, we can be confident that “good” managers are responsible for at least some. If a fund has performance better enough for long enough than any “just lucky” fund could plausibly have, we can be confident that the manager is good.
We can look at sustained outstanding performance in at least two ways. We’ll say that
- A fund that delivers a better return than the market for several years in a row has a winning streak.
- A fund that has a higher average return than the market for a long time (but may have had some off years) is a career achiever.
The table shows how long a winning streak has to be before we can be confident that the manager with the streak isn’t “just lucky.” Each cell in the table indicates how many lucky fund managers we’d expect to see. Given the number of mutual funds, and the variability we see in market returns, we wouldn’t be surprised to see as many as 1550 funds outperform the market in any one year. Nor would we be surprised to see as many as 6 managers outperform the market for 10 years in a row.
Streaks that fall in the green shaded area would be very unlikely for “lucky” managers. So a manager outperforming the market by 24% for one year is unlikely to be lucky. A manager with a 20 year streak of outperformance by any amount is also pretty likely to be skilled. And Lynch? Jeremy Siegel estimates that the odds are 1 in 500,000 that he was lucky in his best stint2.
Intuitively, if a manager doesn’t win every year, we need to see more winning years before we will be confident of the manager’s quality3. The statistics confirm that intuition. The next table shows how many managerial careers could outperform the market without surprising us, by size of advantage and career length. We wouldn’t be surprised to see anyone with a slight 30 year career performance advantage over the market – whereas any manager with a 30 year winning streak is almost certain to be “good.”
Morningstar and other rating services typically assess managers and funds over 3, 5, and 10 years. Over 3 years, we’d expect more than 100 managers to outperform by as much as 6% – if the market returned 10% a year over that period, we’d expect numerous “just lucky” managers to deliver 16%. Over 5 years, we’d expect about 100 managers to outperform by as much as 5%. Even at 10 years, outperformance of 3% per year wouldn’t surprise us. It would take 8% per year over those 10 years to convince us that we are dealing with someone special, someone whose performance is very unlikely to be “just lucky.”
Finally, suppose we were to find a manager whose performance is convincing – with results way beyond what we’d be able to dismiss as “lucky.” How can we be sure that the manager’s skill will continue to be effective? It may be that others will be able to copy the successful technique, or that the manager will tire of the business. And, just as it takes a lot of time and much success to persuade us of the manager’s skill, once we are persuaded, it takes a good deal of time and failure to cause us to change our minds in the other direction.