I’ve written before about behavioral biases, and how they can trick you into abandoning your investment resolve. For example, it’s easy to succumb to recency bias, giving breaking news more weight than it deserves in your financial plans. Or we regularly see investors fall for herd mentality, chasing and fleeing popular but ill-fated trends.
As if our human foibles weren’t enough, the data we turn to when considering past performance also is subject to insidious biases. Survivorship bias is one trick of the trade we must watch for when accepting or rejecting performance data analyses.
What Is Survivorship Bias?
Survivorship bias occurs when an analysis omits returns from in-sample funds that were either closed, merged into other funds, or otherwise died along the way. Examples of a “sample” might be the returns from all actively managed U.S. stock funds during the past decade, or the returns from all global bond funds from 2000–2014.
Why Does It Matter?
Funds that have “disappeared” matter to our data analysis because they’re almost always the ones that have underperformed their peers. As described in this Vanguard report, whether a fund was liquidated or merged out of existence, underperformance was the common denominator prior to closure, by “a significant majority.”
An analysis marred by survivorship bias is highly likely to report overly optimistic outcomes for the group being considered. While a degree of optimism can be admirable in many walks of life, basing your investment decisions on artificially inflated numbers is more likely to set you up for future disappointment than to position you for realistic, long-term success.
How Often Do Funds Go Under?
Does survivorship (or lack thereof) really make a dent in a fund family’s overall score? There is compelling evidence that it can indeed. In the competitive capital markets in which we operate, many fund managers launch new products and discontinue existing ones all the time – probably far more frequently than you might think.
Fund manager Dimensional Fund Advisors has been keeping an eye on the U.S. mutual fund landscape for several years now, producing an annual report that tracks the damage done. Here’s a visual of the survivorship rates they saw at year-end 2015.
Out of a universe of 3,550 funds that existed on December 31, 2012 (three years ago), about 13% of the initial funds hadn’t survived. And out of the 2,730 funds that existed on December 31, 2006 (ten years prior), the death toll is closer to 41%. That’s getting uncomfortably close to 50/50 odds that a fund you could have invested in a decade ago no longer even exists today.
Other analyses have documented similar results:
• A recent S&P Dow Jones Indices analysis found that, for the five-year period ending in December 2015, “nearly 23% of domestic equity funds, 22% of global/international equity funds, and 17% of fixed income funds have been merged or liquidated.”
• The aforementioned, January 2013 Vanguard analysis looked at a 15-year, 1997–2011 sample of funds identified by Morningstar. It found that 46 percent “were either liquidated or merged, in some cases more than once.”
• A May 2015 Pensions & Investments (P&I) article reported that Exchange-Traded Products (including ETFs) weren’t immune from the phenomenon either, having just reached the milestone of 500 products closed. According to the “ETF Deathwatch”cited source, this represented a mortality rate of just under 23 percent.
What’s the Bottom Line?
If an industry or academic study succumbs to survivorship bias, failing to include the performance data from the significant numbers of funds that disappeared during the period being analyzed, you’ll want to be highly suspicious of the results. Any alleged outperformance being reported may be more of a phantom figure than real returns.
Moreover, survivorship bias is only one of a number of faults that can weaken a seemingly solid study. One way in which we strive to add value to investors’ evidence-based investment experience is to help them separate robust data analysis from misleading data trickery. We hope you’ll be in touch if we can assist you with your own strategies and selections in a market that is too often rigged against the individual investor
1. Registration with the SEC should not be construed as an endorsement of Adviser’s investment skill or acumen. 2. Investments in securities are not insured, protected or guaranteed and may result in loss of income and/or principal. 3. This article is intended to provide general information only and should not be construed as an offer of specifically tailored individualized advice.