Mutual fund survivorship bias

December 01, 2017

When my son was in second grade, one of his report cards was, let's say, not perfect. But he had been wise enough to pre-check the envelope before he brought it home. After he read the contents, it magically disappeared. He arrived home claiming there was no report card this quarter, sadly it had been lost.

Unfortunately for him, one of his sisters found it, a few days later, buried in the school's clothing lost and found.

Active mutual funds love to play this game with the public. If they do well, it makes the front page of the glossy marketing material. If the fund manager is dogging it, the firm is tempted to kill off the fund or, more commonly, merge it into another fund. This obscures the poor performance statistics.

It's their version of burying bad news in the lost and found, and it's misleading. This is called "survivorship bias," because only the survivors are factored into the reporting. Vanguard Research did a study on this trend, covering the years 1997 - 2011. They found:

• Only 54% of funds managed to survive the full period.
• 87% of merged funds underperformed prior to the merge date.
• The act of merging funds did not improve performance in general.

The research firm Lipper also dove into this topic. From 1986 to 1996, 568 mutual funds were reported to return 13.39%. But when those same returns were reported a decade later, they had magically risen to 14.65%. How did the supposedly same group of funds, reported over the same time period, get a 1.26% improvement performance? In the later reporting period, there weren't 568 funds, there were only 434 surviving funds. Surprising no one, the 134 that had been liquidated or merged away were not the best performers.

Burton Malkiel, Princeton economics professor, opines: "if you look at results for dead funds as well as survivors, you show that mutual-fund returns are a lot lower than most of the published statistics." In fact, a study he did from 1982 to 1991 shows survivorship bias added 1.4% a year to published fund returns.

Perhaps Malkiel had his own experience with a lost and found long ago. He certainly knows where to look.

Dan Cunningham

WSJ: Dead-and-Buried Funds would alter averages
Vanguard: the mutual fund graveyard