Manager Change Data Show Why People Are Crucial to Ratings

Earlier this year, Scott Kolar left  Goldman Sachs Growth Opportunities (GGOAX). His comanagers, Steven Barry and Jeffrey Rabinowitz, remain at the fund. Will they be there five years from now? Maybe, but I wouldn't bet on it.

Kolar was named comanager in 2011 when Warren Fisher and David Shell left. Fisher had been named comanager in 2009 and Shell in 1999. Before they left, Herb Ehlers, Ernest Segundo Jr., and Ken Berents left in 2005, Mark Shattan left in 2003, and George Adler and Robert Collins left in 2001.

Occasionally a spate of manager instability is followed by a long period of stability or vice versa, but more often it isn't. The strongest cultures are places where analysts and managers want to stay for their whole careers. On the other hand, there are firms where everyone has his eye on the exit. Janus, for example, has been forever taking three steps forward and two steps back. In May we learned of three key managers leaving, including two of its very best--Chad Meade and Brian Schaub. Janus has had a few CEOs in the past 15 years, and it has had a hard time holding on to its best managers. We lowered our Morningstar Analyst Ratings of  Janus Triton (JATTX) and  Janus Venture  (JAVTX) to Neutral from Silver when the news came out.

I set out to see just how predictable manager changes are. I grouped funds by the number of manager departures in the five years prior to 2002 and 2007 and then looked at how many departures the funds suffered in the ensuing five years. I created five groups: no departures, one departure, two departures, three departures, and more than three departures. I looked at it by category grouping and overall.

Why departures instead of manager changes? To me, the addition of a manager to an existing team doesn't sound like a negative. Instead, I wanted to focus on subtractions.

In our Analyst Ratings process, the people pillar is a key factor. Janus Triton's profile changed dramatically when its managers left. Our focus is not on the outer shell of name, ticker, and returns, but on the key drivers behind a fund. The test below of manager departures illustrates why you want to pay attention to managers even more than returns.

Change Begets Change
The data show that, sure enough, stability trends continue. For example, among U.S. equity funds grouped by their departures from 2002 to 2007, funds that had no departures in the trailing five years were much less likely to lose a manager than funds that had lost three managers in the prior five years.

Going from no past departures to one led to a modest increase of a future 1.05 departures to 1.37. At two past departures, it surged to 2.58 departures on aver-age. The rate dipped a bit to 2.06 departures for funds with three past departures and then rose again to 3.46 departures for those that shed more than three managers in the earlier period.

The results were pretty similar in other groups. Balanced funds were particularly dramatic as there was a swing from 1.2 future departures at the low end to 5.4 departures for balanced funds that had more than three departures the prior period.

The trend worked for taxable-bond funds, too, though the disparity was smaller--1.1 to 2.6 from top to bottom. In municipal-bond funds, the trend was less visible. There, I saw a shift from 0.81 to 1.9 for the first time period, but it actually dipped from 0.90 to 0.82 in the second series. I'm not sure why munis wouldn't follow the pattern. At some firms, muni managers have a more team-oriented process and the named manager may switch from one fund to another without much upset. However, I don't know whether the practice is widespread enough to be an explanation. Certainly there are fewer dramatic ups and downs in performance in munis. Maybe that means less turbulence in the manager's suite.

Let's look at the results for all funds. In the 2002 test, funds with no change in the prior five years experienced 1.1 departures on average in the next five years. That was followed by 1.53 for one departure, 1.83 for two departures, 2.34 for three departures, and 3.00 for more than three departures. For the 2007 test, we saw 1.04 departures on average for the no-departures group, then 1.56 departures, 2.03 departures, and 3.34 departures for the last group.

In sum, the more recent departures a fund has endured, the more likely it is to have many in the coming years. If you want stability, then check the record of past manager departures. On the one-page fund reports you see by clicking on "PDF Report" at the top of an analysis on morningstar.com (Premium Members), past manager changes are marked in the returns graph with a triangle.

What About Past Returns?
That's a question I asked as I was running the above data. I wondered whether funds with bad performance were even more likely to make changes than funds with past manager departures. So, I set up a similar test, only this time I used past performance over the trailing five years.

The answer--yes, there is a slight link between performance and departures. It is not nearly as strong as past manager departures, though. For 2007, funds in the top-performing quartile saw 1.4 manager departures on average in the ensuing five years. Funds in the second quartile saw 1.73 departures on average, while third-quartile funds saw 1.94 departures on average. Bottom-quartile funds saw 2.08 manager departures on average.

That gives us a difference of 0.68 manager departures on average compared with 1.81 manager departures from the top to bottom groups based on past manager departures. So, manager departures were nearly 3 times more predictive of future manager changes.

How Can I Apply This?
Glad you asked. This suggests to me that I want to steer clear of most funds with more than a couple of departures in recent years. If a fund has had a number of manager changes, be very skeptical about claims that everything is ducky now. It just doesn't happen often.

A stable manager's suite is no guarantee of great performance, but it's pretty hard to have a good outcome if managers continue to leave. You'll have to endure strategy changes and portfolio turnover, and you won't have as much to go on in figuring out a fund's risk/return profile. If a manager has a 15-year track record, you can at least build realistic expectations about the upside and downside from calendar-year performance.

The only clear exception would be index funds where management is less important and turnover rarely signals a change in the quality of management.

Funds With Low Manager Turnover
To give you some ideas, let's look at some firms with low levels of manager departures as well as fund companies with very high manager-retention rates.

The standouts at a firm level are FPA, American, Dodge & Cox, Primecap, and T. Rowe Price. Many of the funds from these firms are good bets to have stable manager and analyst support. True, T. Rowe Price has lost a couple of good managers lately, but the data show that this is generally not the norm.

On the downside, Putnam, Columbia, American Century, and Dreyfus are near the bottom with retention rates below 90%. Getting near the top of the list is very hard to do as good cultures don't spring up overnight.

For a list of the open-end funds we cover, click here.
For a list of the closed-end funds we cover, click here.
For a list of the exchange-traded funds we cover, click here.
For information on the Morningstar Analyst Ratings, click here.

A version of this article appeared in the June 2013 issue of Morningstar FundInvestor.