Survivorship Bias

This morning’s Wall Street Journal interviews Peter Lynch, the legendary and erstwhile manager of the Fidelity Magellan Fund, who, unsurprisingly, turns out to be an advocate of active equity management.  “People accept that active managers can’t beat the market and it’s just not true,” says Mr. Lynch, who certainly did beat the market in his day.

In support of this view, we learn that “a Fidelity spokesman says about three-quarters of its 49 equity funds managed by the same portfolio manager for at least five years were beating their benchmark over the manager’s tenure…”  A casual perusal of Fidelity’s website reveals a total of 110 actively-managed equity funds (41 U.S., 25 international, and 44 sector-specific), which means that less than half have had the same manager for five years.  If Fidelity wished to do so, they could implement a policy of changing the manager of any fund which hadn’t outperformed over the manager’s tenure (perhaps after a five-year probationary period).  Then their spokesman could announce that 100% of the funds managed by the same portfolio manager for at least five years had beaten their benchmark.  It would be a more impressive sound bite, but just as meaningless a statement about the nature of active management.

Otherwise said: suppose two new portfolio managers take over new funds this year, and after four years one is outperforming her benchmark while the other is underperforming.  Which of the two do you think has the better chance of making it to year five?  The Fidelity statistic is a classic illustration of survivorship bias.  When we control for survivorship, as in our SPIVA reports, the majority of active managers underperform most of the time.

What is true across the population of active managers does not mean that individual managers cannot be exceptions.  Indeed, Peter Lynch is famous precisely because his performance was exceptional.   If most active managers could outperform consistently, we wouldn’t celebrate the few who do.

The posts on this blog are opinions, not advice. Please read our disclaimers.

Leave a Comment

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>