As more and more people are putting their money into the care of professionals, it's time to ask if the people investing your money are properly motivated — that is, if they are given incentives to make the right decisions when they construct portfolios composed of stocks and other risky securities. It's a good question because 44 percent of the $8 trillion in U.S. equities are now held by institutional investors, compared with just 6.1 percent in 1950. In the last 40 years, pension funds, life insurance companies, and hardworking baby boomers have poured money into the hands of stock portfolio managers instead of investing it themselves.
That huge shift in securities management from the individual to the professional prompted finance professors Paul Pfleiderer and Anat Admati to look carefully at the incentives used to compensate portfolio managers. Pfleiderer, the William F. Sharpe Professor of Financial Economics, and Admati, the Joseph McDonald Professor of Finance and Economics, were especially interested in analyzing how managers of active funds are rewarded for their performance. They found that the way active managers are compensated doesn't always generate the best results for the investor.
Active portfolio managers attempt to beat the market by identifying over- and undervalued stocks. They invest in the securities they deem to be undervalued and in some cases short-sell the ones they believe are overvalued. By contrast, passive fund managers adopt a buy-and-hold strategy in which their goal is to mimic the performance of a market index such as the Standard & Poor's 500 or the Wilshire 5000. Passive managers buy stocks in their market proportions — if the market value of Exxon is exactly twice that of Ford, then a passive manager has exactly twice as much invested in Exxon as in Ford.
Experts have long noted that the average performance of all active funds cannot exceed the performance of a passive fund that buys the market index. This is because, on average, the active managers must hold the market index. If one active manager is bullish on IBM and wants to buy more IBM for his portfolio, he must buy it from someone else who is bearish on IBM and wants to reduce his holdings. In other words, a manager can deviate from holding securities in his market proportions only if someone else deviates in the opposite way. Across all of the active funds, these deviations cancel out and so the average performance of active funds cannot be greater than the performance of the market. Active managers are engaged in a zero sum game with the gains of the winners exactly offset by the losses of the losers. In fact, since active managers incur trading costs, the game is actually a negative sum game.
On average, active managers do under perform the market and often by a wide margin. But this does not rule out the possibility that some active managers are able to beat the market with some consistency. In their study, Admati and Pfleiderer assume that there are managers who have information that allows them to earn a superior return (at the expense of other investors). They look at how these managers should be compensated with incentives to use their information in a way that benefits those who have entrusted their money to them.
An active portfolio manager's compensation package can be assembled from three components:
- A fixed fee that is paid regardless of how well or poorly the portfolio performs
- A fee that is based on the overall return earned on the portfolio
- A fee that is based on how well the portfolio does relative to the return earned on a benchmark index such as the Standard & Poor's 500.
In the past, most active managers were rewarded using schemes based only on the first two components. Recently, greater emphasis has been placed on the use of benchmarks. In this case, a manager is rewarded for beating the benchmark, not for simply posting a high return. On the face of it, this sounds like a reasonable approach and a better way to motivate the manager to create value for the investors. After all, the investors can achieve the return of the benchmark portfolio without the help of the active manager. The active manager should be compensated for what he adds over and above the benchmark return. But, says Pfleiderer, "It turns out it doesn't work that way."
In fact, benchmark compensation has some potentially serious drawbacks. The use of benchmarks distorts the way a manager uses information because the manager's and the investor's goals are no longer the same. The manager perceives risk and return in a different way than the investor, causing the manager to form a portfolio that is not optimal for the investor. The manager is getting a payoff for how well the fund does relative to the market benchmark, but the investor is interested only in how well the portfolio does overall. "You want the manager looking at the problem in the same way you are looking at it," says Admati.
Pfleiderer suggests a simple analogy to understand the ramifications of their findings. A contractor gives a bid — his benchmark — to build a house complete with wood paneling. His compensation depends on his ability to finish the house at that price. If the cost of wood rises 50 percent, the contractor will either skimp on wood finishings or lose money. Clearly, the contractor's incentives are not the same as the investor's. A scheme that tends to align the contractor's interest with the investor's would force the contractor to build two identical houses, one for the investor and one for himself. Any decision he makes for one house must be implemented in the other. Obviously, this scheme is impractical for contractors, but it is quite easy for portfolio managers — the manager simply receives a portion of the overall value of the portfolio.
Admati and Pfleiderer also discovered that benchmark distortions can significantly cost the investor. Their findings suggest that somewhere between 20 percent and 50 percent of the value of an active manager's information can be lost due to the distortions.
When Pfleiderer and Admati considered whether there might be some benefits that benchmarks offer to justify the loss, they were unable to find any compelling reason for their use. They found, for example, that the use of benchmarks did not create a greater incentive for the manager to work hard than might be created by other compensation schemes. They also found that the use of a benchmark to determine compensation did not allow the investors to "screen out" undesirable managers any better than they could through other schemes. While they won't say it is always a mistake to use benchmarks in fund manager compensation, Pfleiderer and Admati remain skeptical about their widespread use.