Of all the useful data Morningstar produces for investors and financial professionals, its mutual fund categories are among the most frequently cited for peer group performance and investment approach comparisons. The company uses portfolio holdings to construct its categories, classifying equity funds by the types of securities they own. This data is then used to place funds into the familiar categories (based on each portfolio’s three-year average coordinates) of the Morningstar Equity Style box.

The categories are useful in helping investors to identify and compare funds that share similar traits. Morningstar derives them from a formulaic analysis; they are not based on the fund’s stated strategy or investment approach. Morningstar categories for domestic, non-sector equity funds reflect a portfolio’s primary investment focus over the past three years. Morningstar reviews category assignments semiannually, incorporating all portfolio data over the prior three years up through the most recent quarter-end. The process is partially quantitative, involving a calculation of the three-year averages for various statistics and a recommendation about the appropriate category for the portfolio. There is also a qualitative element in which Morningstar’s fund analyst team reviews suggestions from the quantitative program and recommends whether any changes should be overruled or upheld. However, while both our investment philosophy and process, which date back to 1972, have remained steady over the years, most of our Funds have experienced frequent movement in and out of Morningstar’s equity style categories. This led to us wonder how common it’s been for funds to move between categories.

To quantify the degree of flux among Morningstar categories, we first compiled historical fund category data by taking 2003, 2008, and 2010 year-end snapshots of all the funds in the Small Value, Small Growth, and Small Blend categories, including funds that had merged or been liquidated. We then compared each of those snapshots to their peer group at March 31, 2014 and computed the differences.

Morningstar Royce Funds

Small Value

We were surprised by how, starting from the end of 2003, fluid and dynamic the Morningstar categories have been. Of those funds that were in the Small Value category at the end of 2003, 32% have since shifted to another category, and 27% moved from the end of 2008. More important, from our perspective, close to 50% of the funds that made up the Small Value group at the end of 2003 were no longer in the category at the end of 2010. We suspect that many of these portfolios have disciplines similar to those we use here at Royce. Yet this category thus showed less persistence than Small Growth, which was somewhat surprising considering our own and others’ long-term buy-and-hold approach.

Small Growth

The Small Growth category also saw frequent movement, with 45% of the portfolios in the Morningstar Small Growth category 10 years ago having since merged or been liquidated while another 16% of the total migrated to another style box. Interestingly, of the three small-cap categories, Small Growth funds have actually shown the most persistence.

Small Blend

The Small Blend category showed similar trends, with only 42% of the funds that were in that group at the end of 2003 still residing in the same home at the end of last year. Our data also showed category instability over even shorter time periods. Comparing the Small Blend category from the end of 2010 to our March 31, 2014 peer group saw 74% of the funds remaining in place—nearly a quarter had moved out of the category in only 39 months.

Large-Cap Categories Show Some Similiar Patterns

To determine if category shifts were primarily a small-cap phenomenon, we ran the same analysis for the Large Cap categories, where we also found some significant differences over time. There was one important variance—we discovered less shifting between categories. Only 7% of the funds in the Large Growth group had changed categories in the 10-year span that compared the 2003 peer group to the 3/31/14 group, and the three-year comparison found that only 8% of the funds found themselves relocated.

We did see, however, a large number of fund liquidations and mergers—56% of the portfolios that comprised the Large Blend category in 2003 have since merged or been liquidated. When comparing the set of Large Value funds from three years ago to the same category group at the end of 3/31/14, we found that only 74% remained in the peer group—as we observed with the Small Blend group, approximately a quarter of the portfolios had moved out of the category in only 39 months, including these that have been merged or liquidated.

Conclusion

When considering peer group comparisons, we would all like to think that we are comparing apples to apples. However, our study found that membership in a Morningstar category can evolve considerably over time. Many funds move between categories while others merge or are liquidated out of the peer group. We think it’s important to note that funds often move even if their investment approach has remained consistent. This can create a dilemma for investors who base their decisions solely, or even mostly, on peer group percentile rankings.

While the composition of the Morningstar fund categories we examined was not constantly changing, it exhibited enough flux to raise a series of questions about how investment approaches can best be assessed and classified, as well as how to effectively evaluate any fund’s performance against its ostensible peers. Our research suggests that a fund’s category is changed far more often than seems commonly acknowledged, and this should be a consideration when screening, evaluating, and/or monitoring portfolio performance.

Portfolio managers (and their funds) should certainly be held accountable through a well-defined performance measurement process. At the same time, the analysis must be flexible. It must recognize that a fund’s peer group will change with some regularity even if the fund’s investment approach has not changed. We found that, most commonly, funds in growth or value categories move into blend categories. It must also be kept in mind that mergers and liquidations can significantly alter the category’s landscape over fairly short time periods (that is, in periods as brief as three years).

One simple but comprehensive way to address these issues would be to use broader peer groups, which is what we typically do at Royce. For example, we often start by comparing a given Royce fund to the entire universe of small-cap mutual funds. Alternatively, using custom-designed peer groups can also limit category instability. More robust peer groups can be formulated using a combination of holdings- and returns-based style analysis, which could then be made even more useful by incorporating relevant benchmark comparisons.

After all, the goal is to use an effective method with which to compare portfolios with similar approaches and cap ranges in order to make informed investment decisions. Taking note of the style box or category in which a portfolio resides is only the beginning.

Via Royce Funds