Funds routinely leave one fund category for another, making it difficult for investors to accurately compare offerings.
This article first appeared in the August/September 2011 issue of Morningstar Advisor magazine. Get your free subscription today!
We'd all like to think that we make apples-to-apples comparisons in evaluating funds. Often, we do so using peer groups such as the Morningstar mutual fund categories. Indeed, it's one element of the grading process that Morningstar Investment Services uses in choosing funds for the portfolios that we manage for clients.
Anecdotal evidence suggests, however, that categories evolve over time. Funds change peer groups, either through birth, death, or switches into different categories. This can raise comparability questions for fund researchers and investors who might base their decisions at least partly on relative performance.
In this piece, we attempt to more-systematically quantify the degree of change that fund categories experience, as well as potential implications on portfolio construction and decision-making.
The Morningstar mutual fund categories aim to help investors identify and compare funds that share similar traits. The system is holdings-based, meaning that it classifies a fund based on the types of securities it owns. For example, it categorizes diversified-stock mutual funds based on their three-year average coordinates in the Morningstar Style Box. An equity fund's style-box classification depends on the types of stocks it owns, such as value or growth, small cap or large cap. Morningstar uses various buffer zones in an effort to limit migration between categories.
To quantify the extent of change within the Morningstar categories, we compiled historical fund category data on the oldest share class of every U.S. domestic-stock fund that existed at some point during the 15-year period ended in May 2011. There were more than 4,300 funds in this cohort, including more than 1,600 that had been merged or liquidated away. We then took monthly snapshots of each fund's Morningstar category. Thus, for a fund that existed for the complete 15-year span, we took as many as 180 different snapshots. We compared those snapshots over rolling five-year windows, at one-month steps. For example, in the case of a still-existing fund that launched in June 2001, we took 120 snapshots and made the first of 60 five-year comparisons in May 2006 (when we compared the fund's category classification with its classification in June 2001).
After making these comparisons, we tallied up the results by category into groupings such as "Left" (for funds that finished a rolling period in a category different than the one they started in), "Stayed" (funds that finished a rolling period in the same category they began in), and "Died" (funds that existed at the start of the rolling period but merged away or liquidated during the period).
The Morningstar fund categories we examined were dynamic, with funds regularly switching categories. Exhibit 1 shows the median percentage of funds in various categories that either stayed put for the entire five-year rolling period, left the category by the end of the five-year period, or died through merger or liquidation at some point during the period.