The Power of Mean Reversion in Morningstar’s Five-Star Ratings

Stockholm (HedgeNordic) – Mutual funds with a Morningstar five-star rating seldom sustain their above-average performance over time, writes a Wall Street Journal cover story on investment research firm Morningstar’s mutual fund ratings.

The WSJ article discusses the “regression to the mean” phenomenon for mutual funds, particularly for the highest-rated ones. Morningstar rates mutual funds on a one-to-five scale depending on how well they have performed relative to their peer group (after adjusting for risk and sales charges). Within each Morningstar category, the top 10% of funds receive five stars, whereas the bottom 10% receive one star.

Many individuals and institutions, including pension funds, endowments, and financial advisers, use Morningstar’s star ratings system to decide where to allocate capital. However, WSJ’s detailed analysis points out that the rating system does not serve as a good indicator of future performance. Interestingly enough, only 12% of the funds awarded a five-star rating remained in the top category over the next five-year period. An alarming 10% of those funds performed so poorly that they received a one-star rating.

The analysis shows that ratings of one-star and five-stars rated funds converge after ten years. The average score for a five-star fund tends to decline to three stars, and for a one-star fund, it increases to 1.9 stars.

Fund companies “heavily advertise their star ratings”, says the WSJ, with investment flows often pouring into the funds freshly awarded with five stars. Soon after the article’s publication, Morningstar officials claimed the rating system should only be a first-stage screen of potential investments rather than a predictor of future performance. In defence of Morningstar, informed investors should be well aware of the “regression to the mean” phenomenon.

The regression to the mean takes place typically because most phenomena include a deterministic and a random component. Indeed, the investment skills (the deterministic factor) of fund managers matter quite a lot, but so does luck (the random factor). When a fund is registering exceptional results in a given year, most probably a talented person is having a lucky year. Of course, talent is reasonably stable over time, but luck is not. And after observing a string of exceptional performance, the fund managers are likely to see their returns regress to the mean, as they run out of luck.

Picture (c) Blotty – Dreamstime.com