Note: The featured writer is solely responsible for the content of this article. The opinions expressed herein are not necessarily those of MFI or BES, Inc.

Beware Of Falling Stars

By Werner Renberg

If you pay a financial planner to help you to choose stock, bond, or hybrid mutual funds and to oversee your fund portfolio, you're probably not surprised if he or she recommends that you buy funds that are given 5 or 4 stars for risk-adjusted performance and that you sell funds that are given 1 or 2 stars, as calculated by Morningstar.

Like many other mutual fund marketers -- as well as many individual investors -- who may, or may not, fully understand the methodology developed by Morningstar for assigning stars, many financial planners find the symbols of the stars to be one of the more easily understood reflections of a fund's suitability (or the lack thereof).

Which, of course, helps to explain why shares of 5- and 4-star funds tend to outsell shares of other funds, even if stars do not constitute the only explanation.

While I would hope that your planner will not have claimed that stars have predictive values -- something that Morningstar itself stopped doing many years ago -- he or she may tell you with a bit of confidence that:

o The odds are close to even -- but not significantly higher -- that funds which rate 5 or 4 Morningstar stars are likely to remain rated 5 or 4 for a year.
o The odds that funds will be -- and will remain - 5- or 4-star funds are higher for older funds than for newer ones. (Morningstar does not rate funds until they are 3 years old.)

Your planner is likely to make these points -- cautiously, I trust -- because of a new study, conducted by Mark Warshawsky, director of research at the TIAA-CREF Institute, and summarized in an article, "The Persistence of Morningstar Ratings," in the September issue of the Financial Planning Association's Journal of Financial Planning.

To understand the significance of Morningstar's ratings, which some sponsors of highly-rated funds like to advertise, or of a study of ratings such as Warshawsky's, it is important to understand how the ratings are calculated, which fund ads and sales literature do not always spell out clearly and prominently.

The ratings reflect historic risk-adjusted returns -- not raw, unadjusted total returns, as you find in funds' shareholder reports or in this newspaper -- in a way that permits comparisons with peer funds.

You don't need to know all the esoteric details of the methodology developed by Morningstar since its establishment in 1984, which differs from other ways of calculating risk-adjusted returns, but you should know the essentials, if you are going to base investment decisions on symbols.

Having divided the mutual fund universe into four groups -- domestic stock, international stock, taxable fixed-income, and tax-free municipal bond funds -- the firm calculates overall ratings every month for 3-, 5-, and 10-year periods for funds with at least 3 years of performance history.

How? By determining the differences between certain measures of return and risk.

The risk measure is based on how much a fund underperforms a riskless asset, 90-day U.S. Treasury bills, and how that compares with its group's underperformance.

The return measure is based on a comparison of how much a fund's load-adjusted return exceeds that of a T-bill with the excess return of its group. (By adjusting returns for sales loads, Morningstar puts load and no-load funds on an equivalent basis.)

Funds are then ranked within their groups and given stars for risk-adjusted performance over both their lifetimes and certain periods: 5 stars for the top 10 percent, 4 stars for the next 22.5 percent, and so on. For funds 3 to 5 years old, the 3-year and overall star ratings are the same. Overall ratings for funds 5 to 10 years old are weighted 60 percent for their 5-year ratings and 40 percent for their 3-year ratings. For funds 10 or more years old, weights of 50, 30, and 20 percent are assigned, respectively, to their 10-, 5-, and 3-year ratings.

(Morningstar also calculates more relevant 3-year ratings for each category within the four groups, such as large value, large blend, and large growth within domestic stock.)

Warshawsky and associates Mary DiCarlantonio and Lisa Mullan studied the overall ratings -- for three 1-year periods, 1997, 1998, and 1999, and one 2-year period, 1998-99 -- for stock and taxable bond funds. ("…extending the time period would not overturn the basic results," they wrote.)

What did they find?

That from 49.2 percent (1998) to 56.1 percent (1997) of all 5- and 4-star funds persisted in maintaining their ratings over 1-year periods, and only 30 percent did so for the 2-year period (1998-99).
These rates lagged the 61 percent of all funds with 5 and 4 stars at the end of 1992 retaining their ratings in mid-1995, as determined in a 1998 study, which, they said, was the only other one to examine the persistence of Morningstar ratings among mutual funds.

Warshawsky and his co-authors declared that the previous study, unlike theirs, overstated persistence because it reflected survivor bias by eliminating funds that had been merged, liquidated, or renamed during the study period.

Lest the focus on star ratings -- by studies, fund sponsors, or fund marketers -- be taken to imply that persistence in earning 5 or 4 stars for risk-adjusted returns is all that matters, I should point out, at the risk of stating the obvious, that unadjusted returns matter more. They measure your accumulation of wealth.

While you would not expect a direct correlation between adjusted and unadjusted performance, there does seem to be a greater likelihood for older funds to earn 5 or 4 stars and for such funds to remain highly rated -- persistence that Warshawsky and associates find "surprising."

Look at Morningstar Mutual Funds, and you'll find examples, such as Janus Fund, Fidelity Contrafund, and Vanguard Short-Term Corporate Fund, which have averaged over 4 stars every month for nearly 15 years and remain 5-star funds. As you might expect, their average annual total returns for the last 10 years topped those of relevant stock and bond indexes.

Copyright © 2000 Werner Renberg. Reprinted with permission.

More articles by Werner Renberg can be found here.