The world of mutual fund analysis is a curious place indeed. Just about any document that purports to analyze and evaluate fund performance comes with that tried-and-true disclaimer: “Past performance is an unreliable indicator of future returns”. The rest of the document, more often than not, will deal almost exclusively with… you guessed it, past performance. What’s going on here? Past performance is unreliable and yet an entire industry seems to be built around past performance.

I think the pertinent question to ask is this: does being unreliable mean that past performance is unimportant? That is what I will focus on in this post.

Past Performance is Unreliable

Let’s be clear about one thing upfront: the disclaimer is right. Past performance is unreliable. How a fund performed over the last twelve months, or three years, or five years, says little about what it will manage to generate in returns over a future sequence of similar time intervals. That applies both to absolute measures – the actual returns earned – and also to relative performance – how well the fund performed relative to its peer group. Numerous studies have been undertaken over the years to supply compelling evidence that the most prevalent factor explaining sustained performance success is simply the luck of the draw. Outperforming a benchmark index for three years in a row, or landing in the top 25% of a peer group universe for the same period, owes more to normal probability distributions than it does to the skills and smarts of the fund management team. That is to say, while for any given fund it is unlikely that the fund will outperform over multiple successive time periods, for the fund universe as a whole it is probable that anywhere from 5-10% of the funds in the universe will exhibit just such outperformance. The point these studies make is that which funds land on the “long tails” of these probability distributions is by and large random.

Observer-Created Reality

Now back to the question at hand: does this mean that investors should not pay any attention at all to past performance? I would argue that no, it does not mean that. Investment markets are in no small part an observer-created reality, meaning that the deliberations and actions of the actors in the markets have a decisive impact on outcomes, whether those outcomes are rational or not. Consider the other side of the probability distribution – the unfortunate souls who show up in the bottom 25% of their peer group for three years in a row. If this performance is random, like the studies say, then we should not expect a higher than average probability of their repeating this dismal performance for the next three years. Yet markets – and more specifically the investment committees of large institutional investors under constant performance pressure themselves – will be likely to dump these funds from their portfolios. The funds may then find themselves hard pressed to meet their operating expenses, they may turn to desperate, ill-advised strategies to turn things around – and voila! Perception has become reality. Investors, it would seem, should pay attention to past performance for this reason – but not for this reason alone. In fact there are additional important reasons why past performance should figure into the evaluation process.

Achieving a Composite Picture

In general, there are intelligent ways of evaluating past performance along with unintelligent ways. An example of an unintelligent way to measure past results would be something like this: “ABC Fund beat the S&P 500 for the last twelve months and has been in the top 25% of its Morningstar peer group for the last two years. Let’s sell XYZ Fund and replace it with ABC Fund ASAP!” An intelligent way to evaluate past performance is to look more carefully at multiple time periods – short, medium and long term – along with the strategy the fund managers actually pursued in order to generate those results. Did they stay disciplined in accordance with their stated investment strategy, or did they wander off the reservation to chase the ephemeral flavor of the day? The analysis over multiple time periods should provide some insight into the nature of these returns beyond the headline numbers.

But there is much more to “past performance” than the single idea of total average period return. How much risk did they take on in achieving those returns – both in absolute terms like standard deviation and relative measures like beta? Were their fees in line with the norms for funds in their asset class? How did they perform in extreme bull rallies or bear contractions?  These and other similar questions drive the thinking behind Jemstep’s algorithms that employ over 70 data points across six categories to arrive at a composite picture (driven by the priorities and preferences of each investor).The point of this kind of analysis is not to arrive at the conclusion that because the fund achieved X% for the last three years it will likely achieve X% in the next three years. Rather, these data points add up to an integrated analysis about how the fund operates and whether what they say they do more or less aligns with what they actually do. That is information that investors should reasonably consider to be important.

Past performance by itself is an unreliable indicator. But when looked at in conjunction with other factors, over different time periods and market conditions, using different measures of return and risk, it can supply insights useful to investors in choosing among alternatives for inclusion in their portfolios.

Tell Us:  Do you think past performance matters?

For a free, easy and unbiased way to find the best investments for you, visit Jemstep.com. 

TAGS: