Using Morniningstar's Star Ranking System As A Fund Screening Tool

It’s not often that you get to back-test a key part of your fund selection methodology and have someone else do all the work! Such was the case recently when I came across a paper by Matthew Morey and Aron Gottesman of Pace University . Morey and Gottesman have produced the first study of the predictive value of Morningstar’s revised star rating system for Mutual Funds. As background material for the two or three investors out there who may not be familiar with the system, Morningstar rates mutual funds on a scale of 1 star to 5 stars with one being the worst in category and five being the best. This system was revised to its present form in 2002. The Pace University study looked at subsequent three year returns following the change to the new system.

Our interest in this at PCM is that the Morningstar ranking system is the first line in our screening process for funds used in our managed accounts. Our first step is to eliminate all funds ranked 1 or 2 stars. My thinking here is that we eliminate the riff – raff by making this initial cut. The key question is: Is this effective in eliminating funds we’re really not interested in and likewise are we eliminating some funds that we should be looking at?

Morey and Gottesman expressed the effectiveness of the star rankings by measuring the probability that any funds in a particular asset class would be outperformed by the average fund in another higher or lower rank in the same class. In other words, what percentage of Large Cap Value Funds ranked 3 stars would be beaten by the average large cap value fund ranked 5 stars, etc. This was very useful information from my point of view because a quick calculation showed that on average a fund ranked 1 or 2 stars by Morningstar was 20% more likely to be out-performed by an average fund ranked 3, 4 or 5 stars. Thus we’re eminently justified in eliminating funds ranked in the lowest two categories immediately.

I have been critical of using Morningstar’s rating system exclusively for fund selection for a number or reasons, the most notable being that they will rate funds after only three years of existence. It is possible in the investment management business to be more lucky than skilled over a relatively short time period like three years. Of course one could ignore the three year ranking and just concentrate on the 5-year or better yet the 10-year rankings. However, given the emphasis placed on these rankings by the financial press, financial product sales people and by the funds themselves, it’s not surprising that the investing public could misinterpret the significance of the three year ranking.

We strongly urge our readers not to use the Morningstar Star ranking system as an exclusive means to pick funds but rather an initial screening tool to eliminate grossly unacceptable funds. Final selection should be made using a methodology such as ours that selects for consistent long term superior performance (top 25%) within peer group. Final fund selection should also encompass fees, investment philosophy and methodology, and stewardship of shareholder funds.