Doomed to Repeat the Past?

To some, choosing the most talented money managers is easy. All an advisor has to do is examine their track records over, say, three to five years, and pick those that have performed the best. After all, that's more or less how we go about identifying talent from baseball players to salespeople. Just one problem: Such a strategy is useless in the world of money management. The boilerplate on fund

To some, choosing the most talented money managers is easy. All an advisor has to do is examine their track records over, say, three to five years, and pick those that have performed the best. After all, that's more or less how we go about identifying talent — from baseball players to salespeople.

Just one problem: Such a strategy is useless in the world of money management. The “past-performance-does-not-guarantee-future-results” boilerplate on fund advertisements is all too true. And that's putting it mildly: In extensive research into investment managers over more than 30 years at the Russell Investment Group, we have found that, if anything, past performance is more of a perverse indicator than it is a helpful tool. I say “perverse” because we have found that a manager with outstanding performance over the last several years will most likely be among the lackluster performers over the subsequent years. Further, for most investment styles and asset classes, strong past performance is actually inversely related to strong future results — a perverse relationship indeed and a big reason why so many retail investors' portfolios perform poorly.

The late 1990s gave investors many opportunities to succumb to the seductive allure of past performance (see diagram on p. 57). One particular “diversified” mutual fund we know had outperformed its equity benchmark by nearly 100 percentage points per year for the two years ending Dec. 31, 1999. Not only did investors think the fund manager could do no wrong, the fund manager also did. As expected, the story ends badly. Since that time, the fund has trailed its benchmark by more than 27 percent per year. If an advisor had purchased the fund at the start of 1998 and captured all its great early performance and held it through February 2004, he would have trailed the benchmark by more than 6 percent per year over the entire period.

This example of how the perverse indicator often plays out, albeit extreme, underscores the risk of drawing incorrect inferences based on past performance.

Yet, typically, people screen for performance when selecting an investment manager. The result is that many retail investors buy a fund that has a good track record over the past several years. In short, they buy at the top and then, since these funds tend to do relatively poorly over the next several years, sell at the bottom. Then they invest in the latest hot dot, only to experience the same results all over again. Consider this: The average equity mutual fund produced a 12.2 percent annualized return from 1982 to 2002, according to a 2003 study by market research firm Dalbar; yet, the average investor achieved a 2.6 percent annualized return. The likely reason for the poor performance: switching to the hottest investment managers at the wrong times.

The bottom line is we cannot select the best managers solely by looking at their past performance. Indeed, it would probably be more helpful to have no knowledge of past performance than to rely on it as a primary selection tool.

But all hope isn't lost. At Russell, we believe it is possible to select outperforming money managers. In fact, that's what we do as an investment consultancy. But — and here's the crux of the matter — we aim to pick the better managers before they put in their best performances rather than after they do so. The trick is to look forward rather than backward.

Many people who evaluate investment managers rely heavily on a variety of specific ratios that are meant to shed light on an investment manager's risk and return proposition. These tools essentially compare performance to risk in both absolute and index relative terms to identify which managers produce higher return for a given level of risk. Commonly used ratios include Sharpe ratios, Treynor ratios and information ratios. While these analytical metrics can produce some interesting insights, the perverse relationship between past and future performance means that these ratios are no better at predicting future manager results than raw performance.

How do we do it? Well, some of our methods are proprietary, and we believe relying on one or two quantitative tools is an oversimplified way of handling this complex challenge. Instead, we have the benefit of being able to visit with money managers frequently and carefully investigating them in order to find key determinants of future success, something retail investors and financial advisors don't have the ability to do. This process can't be boiled down into one cookie-cutter template that works in the same way for each manager.

But let me share a few of our key guidelines.

First, we resist the temptation to select money managers solely on the basis of past performance. We take past performance into account, of course, but we do not regard it as a key indicator. Rather, we try to separate the signal from the noise. By “signal,” we mean the “true insight” reflected in a manager's performance. The noise is the portion of performance that comes from such factors that are not expected to have a positive return associated with them. For example, many managers prefer to invest in specific sectors of the market, such as technology or health care. Over time, these biases are generally not rewarded, but they do explain a significant part of shorter-term performance.

Because the noise portion of performance usually swamps the signal, investors who select managers based on past performance are essentially making decisions based on random noise and not on the signal (or skill).

Even if a manager outperforms the index by 10 percent, no skill may be involved, or perhaps 1 percent is skill performance. After all, someone may toss 100 coins and during that time five heads come up in a row. That does not mean that person is more talented than others in tossing coins.

As a result, the factors that introduce noise, such as sector biases, preference for small or large companies, and even luck, tend to revert to the mean. The most powerful force driving this mean reversion is valuation. In other words, when one segment of the market does well for a long period, it tends to become relatively expensive. That overvaluation acts as a gravitational force that ultimately brings prices back down.

How, then, do we separate noise from signal? We use a variety of qualitative and quantitative tools to try to disentangle the managers with skill from those who are less talented. In doing so, we have found that the most challenging aspect of manager selection is less related to investment expertise or knowledge than it is to psychology. The key lies in the attributes of investment managers. The winners have certain traits that help them succeed.

So we look at their experience, integrity, humility and desire to learn. We look, too, at their investment processes, some of which are better than others. We try to discern the better ones by identifying those that make intuitive sense, have good premises underlying them and include a strong dose of self-discipline.

We also believe we can gain insight into investment managers' thought processes by watching how the portfolios they build change over time. But, above all, we look for people who love investing with a passion. Because investment management is as intensely competitive as anything gets, we like people who are exceptionally motivated to the point of being almost obsessed with winning. They are passionate about what they do and they love the game.

We have found their passion is almost tangible; it is reflected in time allocation, incentive structure, employee interaction and company culture (the strength of which is reduced staff turnover).

An advantage is that people with those traits tend to surround themselves with people who share similar traits, creating a firm of passionate investment managers and analysts.

We find, too, that successful investment managers are people who have an inner need to be in the business. It is much more than just a job. It is a calling.

They also are aware of their strengths and weaknesses. They have a realistic assumption of what they can achieve and so they are objective.

It is important to note how they respond to good and bad performance, and how they learn from their mistakes. If they say they know it all, that is a warning sign. Humility is good. Successful investment managers are always worried about what they don't know or how they can become better. If they stay the same, they fall behind because the industry moves fast and is so competitive. They have realistic assumptions about what they can achieve and they practice strong sell disciplines.

They are comfortable with change and have good feedback so they can always relate to the world around them.

So when next faced with having to hire a new investment manager, and when tempted simply to look over a manager's track record to determine whether the manager is likely to perform well in the future, stop and think.

A realistic assessment may lie in quite a different direction.

Paul Greenwood is Director of U.S. Equity Investment Policy & Research for Russell Investment Group.

Hide comments

Comments

  • Allowed HTML tags: <em> <strong> <blockquote> <br> <p>

Plain text

  • No HTML tags allowed.
  • Web page addresses and e-mail addresses turn into links automatically.
  • Lines and paragraphs break automatically.
Publish