The FundQuest study
FundQuest, an investment consulting arm of the European bank, BNP-Paribas, recently issued a report on a study it performed analyzing thirty years of Morningstar mutual fund data. The study tries to determine when active portfolio management has been successful.
the results
general portfolio characteristics
According to FundQuest, there are three general characteristics of overall actively managed portfolios that are highly correlated with investment success. they are:
–manager tenure. The best portfolios are those with the longest-serving managers. Having managed a number of portfolios for fifteen years or more, I’m tempted to blindly agree with this conclusion. But even after a pause for consideration I still think it’s true. Occasionally, a weak manager may survive for an extended period, based on great marketing or office-political skills. But they’re the exception. Poor performance leads quickly to clients withdrawing their money from a manager, which, in turn, leads to his dismissal.
–low expenses. I guess this one makes sense, too. On the one hand, a ind group can hope to distinguish itself by having excellent performance, which may or may not happen. But so long as it sets reasonable risk parameters for its managers–in other words, as long as it has some reason to think its investment performance won’t be truly terrible–it has a good chance of delivering superior returns to clients if it has low expenses.
This may also be just a statement about economies of scale. Successful managers draw in more assets, which tends to lower their administrative expenses, if not management fees.
–low volatility. The study doesn’t give an adequate explanation, in my mind. This is a complex issue that I have strong view on (I don’t like volatility as a risk measure at all!). I’ll just leave it at that.
More interestingly, the study enumerates several factors that, contrary to conventional wisdom, it says have no bearing on performance:
–turnover ratio. That is, a calculation of the dollar amount of securities bought (or sold) during a year, as a percentage of average assets. Conventional wisdom in the consulting community is that the lower the turnover the better. The conventional arguments are two: trading makes the manager feel good but runs up costs and adds no value; lots of trading means the manager has no strategy.
–fund size. Conventional wisdom says that smaller funds have a wider array of stocks they can buy in significant size, and that they’re more niumble.
–number of holdings/concentration of fund in its top ten stocks. This ends up being a closet way of separating growth managers from their value compatriots. Growth managers typically run highly-concentrated, fifty-stock portfolios; value managers will typically have over a hundred (maybe two hundred–how they follow them all is a mystery to me), and as a result run a less concentrated top ten.
What I find interesting is that pension consultants thrive on collecting and analyzing such data, even though Paribas suggests it has no value.
overall manager trends
Hang onto your hat for this one. I’m really surprised by the finding.
FundQuest says that the typical manager underperforms his index during bull markets and outperforms in bear markets. Whether he’s a value manager or a growth manager makes no difference.
But–and this is the signal to grab hold of your hat–this results from the fact that managers run portfolios with a beta of about .8. In other words, managers as a group are risk-averse. In the aggregate, they take considerably less risk than the market. If we use an (unspecified in the report) correction for risk, the performance situation reverses itself. On this basis, managers outperform in bull markets and underperform in bear markets.
It makes some sense as a business decision that mutual fund managers should have as a high-level objective to minimize possible losses, even at the possible cost of performance during up markets. One of my first bosses used to frequently say the pain of underperformance lasts long after the glow of outperformance has gone.
I’m just not sure that it’s right to factor that decision out of the evaluation of manager performance, as Paribas does.
The conventional wisdom, by the way, is that growth managers outperform in up markets and underperform when stock prices are going down. Value managers are thought to do the opposite.
manager specifics
FundQuest lists five types of funds where active managers outperform in upcycles and down. They are, with the most successful first:
Emerging Markets Bonds
Industrials
Consumer Staples
World Allocation
Foreign Large Growth.
In addition to these groups, winners in up markets include:
Mid-cap Value
Small Growth
High Yield
Utilities
Commodity-related and many foreign categories.
Winners in down markets also include:
Energy
Large Value
Small Value
World Stock.
Two things strike me about these lists. The first is the prevalence of foreign-oriented or world funds. I don’t think this is particularly surprising, although others may. To my mind, US-trained managers have a considerable technological edge as securities analysts vs. their counterparts in foreign market. Yes, there may be as many foreign-generated reports. But read the latter, ask yourself how much content there is in them and you’ll see what I mean.
The second is the nearly complete absence of domestic growth funds on the outperformance charts. I have no explanation. Growth investing is generally thought of as being more difficult to do than value investing. As a result, there are supposedly only a few truly excellent growth managers, in a field made up of less-talented wannabes. Still, that’s really stretching for an explanation. (If I hadn’t been a growth investor for most of my career, I’d have no trouble finding an answer. I’d say growth investing is all smoke and mirrors.)
What to do with this information? The FundQuest recommendation, which seems good to me, is to try to use actively managed funds in areas where managers have a demonstrated ability to beat their markets, and buy index funds in the rest.