Past performance is not a reliable indicator of future returns: every SEC-registered investment firm is required to disclose some form of this message in all public advertising media. Yet past returns – most often recent past returns – do play an outsize role in influencing portfolio selection decisions among competing mutual funds and similar assets. Reliance on this one measure of performance can be value-subtracting to portfolios with long term investment goals.
The influence of past performance on selection decisions is attributable in part to human behavioral economics. Our brains are wired to detect patterns, and to supply narratives explaining those patterns. Investors may interpret a “hot streak” – a pattern of sequential years of outperformance – as indicative of consistent superior ability on the part of a fund manager.
Another contributing factor to the focus on past performance is the highly competitive nature of the fund industry itself. Mutual fund managers, most of whom seek to outperform a passive market benchmark, compete intensely for assets under management. For a fund to distinguish itself from its peer universe requires communicating some easily understandable metric to the investing public. Annual and quarterly calendar returns are an obvious choice for this metric.
In this paper we pose the question as to whether there is statistical validity to these hot streaks (and to their converse, an underperformance streak). In other words, we seek empirical evidence either in support of or against the idea that recent past performance is useful as a predictive measure of how the fund may perform in the near to intermediate future.
For our analysis we used the Morningstar universe of U.S. large cap blend funds, a diverse and liquid universe of professional money managers. We analyzed the peer group of funds with a track record extending back at least to 1990. We employed different ways of measuring recent past performance, and then evaluated the subsequent performance of funds that scored high (or low) based on those past measures.
In our analysis we found no statistically meaningful evidence to support the predictive power of past performance. A fund that ranked in the top 20% of its peer group for three years in a row, or a fund that ranked first among its peers for the past calendar year, was no more or less likely to perform well in the next one to three years. Conversely, funds ranking at or near the bottom of the universe were no more or less likely to underperform in the near to medium term.
Prudent investment selection relies on more than a cursory analysis of recent past returns. At MV Financial we obtain a composite picture from a range of quantitative and qualitative metrics to arrive at an evaluation of suitability. Moreover, fund selection itself is only one component of portfolio management. In-depth research, asset allocation, operational execution and ongoing monitoring & repositioning all play key roles in the prudent management of portfolios around the defined goals and risk considerations of each individual client.

Download Article