Volume 20, Number 10
Saurin Patel’s research shows how teams are replacing individuals in the management of mutual funds.
In the world of investment, where people love to tell stories of star traders, the cult of the individual is strong. In recent years, though, the mutual fund industry has decidedly moved from an individual to a team-based approach. From 1992 to 2010, the number of mutual funds managed by teams has increased from 30 percent to 70 percent.
Academics in finance, on the other hand, have come to the opposite conclusion. Existing empirical research tends to find that individuals generate better returns than teams.
This was puzzling to Ivey Professor Saurin Patel. “The industry is saying teams are good and academics are saying teams are bad. So I was intrigued to know who’s right.”
The mutual fund industry, which is very familiar to Patel, is a good context to test this question. “The primary objective of mutual fund managers is to generate higher returns for their clients. This work is both creative and intellectually challenging,” says Patel. “It’s something an individual can do well, but so can a team. The answer isn’t obvious.”
The leading study in the field was published in the prestigious American Economic Review. Its conclusion, that individuals perform better than teams, was based on empirical data obtained from the Centre for Research in Security Prices (CRSP), a well-accepted database.
When Patel started digging into the data, he made a startling discovery: the CRSP data was seriously flawed. Many of the funds reported by CRSP to be managed by individuals had in fact been managed by teams. For his study, Patel used a relatively new database, Morningstar Direct (MD). Through cross-referencing with SEC reports, he found that the CRSP database was only 76 percent accurate, whereas the MD database was 95 percent accurate. “Almost a quarter of the data was bad,” he says, “and that was driving an opposite result.”
Patel replicated the earlier study, using the MD data set of U.S. equity funds from 1992 to 2010. He found that teams actually performed better than individuals by a significant margin - 70 basis points or .7 percent a year, after adjustment for risk and fees.
The move to teams in the mutual fund industry coincides with a substantial growth in information. To process this information, teams divide responsibilities between the group and its members. For example, a team as a whole will decide on how much to allocate to a particular sector, such as financial, but the individual with expertise in the sector chooses which stocks to buy.
In the same study Patel also looked at the relationship between team size and performance. He found that three members was the optimal size, a finding that’s consistent with existing experimental research. “Three member teams have an easier time making decisions,” he says. “In a two member team each person has equal power, which can result in a conflict situation. Four or more members can result in split votes or subgroup dynamics that hamper decision making.”
One might expect that team-managed funds would be more expensive than individually managed ones, but Patel actually found the opposite. The practice in the industry is to spread each manager over a number of portfolios, resulting in a reduction of costs. Patel also found that team-based management resulted in lower portfolio turnover. Research shows that lower portfolio turnover – fewer stocks bought and sold in a given year –lowers fees and is good for fund performance.
In another recent study, Patel found that teams were less likely to engage in “cheating.” In this study he focused on two practices that are considered illegal or quasi-illegal: “portfolio pumping” and “window dressing.” Portfolio pumping is the placing of buy orders on a stock immediately before a public report, artificially raising the price. Window dressing is the practice of getting rid of all your poorly performing stocks just before a public report, for the sole purpose of making the managers of the portfolio look smarter.
Patel found that teams cheat significantly less than individuals - as much as 50 percent. He also found a linear relationship between team size and cheating: as the size of the team goes up the magnitude of cheating goes down.
These findings have important implications for an industry known for unethical behaviour, but Patel believes his research has broader applications. There is a clear trend to group decision making and team management in many areas of business, and even in other professions such as medicine. “There are many benefits to working in a group,” says Patel. “You still have individual accountability, and at the same time you benefit from a diversity of opinion and a broader perspective.”