

With the end of Bill Miller's highly publicized streak of beating the S&P 500 coming to an end at the end of 2006 (as well as a bad year for Bill Gross and about 80% of active managers), this is a good time to think about when investors should consider selling a mutual fund.

Litman / Gregory recently did a study of mutual fund manager under performance during long term periods of superior performance, a study referenced in Nick Murray's most recent column in Financial Advisor. Their results are compelling if not totally surprising. Litman finds that at least 93% of large cap managers who beat the index by 1% or more over a 10 year period will out under perform the index by 2% or more over a three year period during their 10 year run of superior performance. Furthermore, 64% will have a three year period where they trail 5% on an average annualized basis, very bad numbers to say the least.

Let me make those figures easier to read: even the best managers will do poorly over significant periods of time on their way to great performance records.

The lesson is simple; investing is like all human endeavors and results can be lumpy and irregular more often than they are neat and consistent. According to the study's authors the three year periods of under performance are often created by one bad year that is not immediately corrected by great performance in the following year.

This study rebuts the idea that investors need to (or pay someone to) closely "watch" their funds for index under performance and sell funds that lag over short time periods. This is a common marketing point for advisors who believe that they can earn better performance by selling funds after short periods of under performance.

The reality is that, like most things in life, good manager selection probably has more to do with understanding the people and philosophy of a firm than on what they have produced in the last day, month, or quarter. If investors spent more time getting to know the people who manage their money they would probably own fewer funds and have less anxiety when they trail some random benchmark, longer holding periods would be the result which would mean less in taxes and better overall performance.