How Survivorship Bias Can Skew Your Views on Mutual Fund Performance updated thru 2016
It’s important to avoid treating the market like a popularity contest by chasing outperformers or running away from the underdogs. But neither do most investors want to go into the market entirely blind. For that, there are database services that track and report on how various fund managers and their offerings have performed.
Besides ample evidence that past performance does not predict future returns, there is another reason we advise investors to proceed with caution when considering past performance: Many returns databases are weakened by survivorship bias.
With respect to mutual funds and similar investment vehicles, survivorship bias creeps in when only the returns from surviving funds are included in the historical returns data you are viewing.
Here is what happens: As you might expect, there is a tendency for outperforming funds to survive, and for underperformers to disappear. When a fund is liquidated or merged out of existence, if its poor returns data disappears as well, overall historic returns tend to tick upward.
As such, you may end up depending on past performance data that is optimistically inaccurate.
Here is an article that further explains how survivorship bias works. In addition, consider the following illustration from Dimensional Fund Advisors’ report, The US Mutual Fund Landscape 2017. It illustrates how survivorship bias can skew your view on fund performance.
In the beginning – in this case January 1, 2002 – there were 2,587 US equity mutual funds. Now fast-forward 15 years to December 31, 2016. By then, only 48% of those funds (roughly 1,242 funds) had survived the period. Out of the survivors, only 17% (about 440 funds) had both survived and outperformed their benchmark over the 15-year timeframe.1
In the illustration above, you can readily see that the small blue box in the lower-right corner represents relatively low, less than 1:5 odds that any given fund in January 2002 went on to outperform its peers by the end of the 15 years.
If a database instead eliminates the “disappeared” funds from its performance data, the larger gray box disappears from view as well, as in the illustration below. Without this critical larger context, you may conclude that those 440 outperforming funds only had to compete against the 1,242 survivors, versus the actual universe of 2,587 funds. While it may seem as if nearly half of the fund universe has done well, in reality, the less than 1:5 odds have remained unchanged.
You see 100% of what has survived and just over 45% outperforming !!!
But wait, maybe you could “take a look at the past performance, pick the funds that have outperformed after the first 10 years, and pile up on those seeming winners. Dimensional’s report also shares the results from that exercise:
Do Winners Keep Winning? Past performance vs. subsequent performance—Equity Funds
The left-hand side of this diagram shows the top 25% funds (in blue) and next 75% of funds (in gray) during the first 5 years of the 15-year analysis.2 The right-hand side of the diagram shows what happened to the top 25% subset during the next ten years. Only 23% of the initial “winners” continued to stay in the top 25% of performers. This demonstrates that is it is extremely hard to predict “winning” mutual funds based on past performance. Your odds are even worse than what you can expect from a basic coin toss!
So let’s take a moment to reinforce our ongoing advice: Invest for the long-term. Instead of fixating on past performance, focus on capturing future available returns within your risk tolerances and according to the best available evidence. Aggressively manage the factors you can expect to control (such as managing expenses) and disregard the ones that you cannot (such as picking future winners based on recent past performance).
These principles guide the actions we’ve advised all along. We will continue to embrace them unless compelling evidence were ever to inform us otherwise. They are the ones that serve your highest financial interests, which is our highest priority as your advisor.
1 The sample includes funds at the beginning of the five-, 10-, and 15-year periods ending December 31, 2016. Survivors are funds that had returns for every month in the sample period. Winners are funds that survived and outperformed their respective Morningstar category benchmark over the period. US-domiciled open-end mutual fund data is from Morningstar and Center for Research in Security Prices (CRSP) from the University of Chicago. Past performance is no guarantee of future results. See Data Appendix for more information.
2 At the end of each year, funds are sorted within their category based on their five-year total return. The tables show the percentage of funds in the top quartile (25%) of five-year performance that ranked in the top quartile of one-year performance in the following year. Example: For 2007, only 30% of equity funds in the top quartile of previous five-year returns through the end of 2006 maintained a top-quartile ranking for one-year returns in 2007. US-domiciled open-end mutual fund data is from Morningstar and Center for Research in Security Prices (CRSP) from the University of Chicago. Past performance is no guarantee of future results. See Data Appendix for more information
Robert J. Pyle, CFP®, CFA is president of Diversified Asset Management, Inc. (DAMI). DAMI is licensed as an investment adviser with the State of Colorado Division of Securities, and its investment advisory representatives are licensed by the State of Colorado. DAMI will only transact business in other states to the extent DAMI has made the requisite notice filings or obtained the necessary licensing in such state. No follow up or individualized responses to persons in other jurisdictions that involve either rendering or attempting to render personalized investment advice for compensation will be made absent compliance with applicable legal requirements, or an applicable exemption or exclusion. It does not constitute investment or tax advice. To contact Robert, call 303-440-2906 or e-mail firstname.lastname@example.org.
The views, opinion, information and content provided here are solely those of the respective authors, and may not represent the views or opinions of Diversified Asset Management, Inc. The selection of any posts or articles should not be regarded as an explicit or implicit endorsement or recommendation of any such posts or articles, or services provided or referenced and statements made by the authors of such posts or articles. Diversified Asset Management, Inc. cannot guarantee the accuracy or currency of any such third party information or content, and does not undertake to verify or update such information or content. Any such information or other content should not be construed as investment, legal, accounting or tax advice.