Viva la revolution: Unshackle your money managers and cast off their benchmarks.
A cadre of investment intellectuals believes pension executives have tied the hands of their managers, boxing them into narrowly defined investment categories and unduly restricting their ability to provide added value.
"Specialization has been carried to such extremes in the investment management business that it really does severely limit the manager's ability to make common-sense judgments regarding portfolio structure," Steve Leuthold, chairman of investment research firm The Leuthold Group, Minneapolis, told the Association for Investment Management and Research's annual conference in May.
Institutional investors have handcuffed their managers by linking them to benchmarks whose compositions change every year and might not adequately reflect the market, argued Peter Bernstein, president of the economic consulting firm Peter L. Bernstein Inc., New York, and well-known financial writer.
Instead, Mr. Bernstein, with deference to Jonathan Swift, put forward "a modest proposal" to shift to benchmarks based on required return and volatility levels.
Not surprisingly, pension executives and their consultants take umbrage at these revolutionary remarks. Steve Nesbitt, senior managing director at Wilshire Associates, Santa Monica, Calif., said managers have tightly defined mandates for a good reason: "Many managers have narrow benchmarks precisely because they have shown poor performance under broad mandates."
William F. Quinn, president of AMR Investment Services Inc., Fort Worth, Texas, added that Mr. Bernstein's proposal would provide no accountability for managers.
Loosening the straitjacket
But pressure to loosen the "equity manager's straitjacket," as Mr. Leuthold refers to it, is building, especially in light of the vast split in performance by style and market-capitalization during the past few years.
That disparity, driven by spectacular gains among technology stocks from 1997 through 1999, has made life very difficult for value managers, who have been constrained to picking stocks from a poorly performing universe. And for small-cap managers, it was a long time in the desert until 1999's spectacular returns.
Value managers should be free to buy traditional growth stocks at some times, Mr. Leuthold said. In addition, large-cap growth managers should be able to downsize into midcap stocks when large-cap stocks are overpriced. They also should be allowed to keep a sizable cash position, both as a reserve and as a way of enhancing performance.
"I don't think it's sinful for a large-cap growth manager to have 20% in cash," Mr. Leuthold said.
But what really riles him is pressure on managers to sell stocks because they no longer fit into their designated style box: ". . . (L)ooking at it from a common-sense perspective, happiness should be a value stock that becomes a growth stock. And frustration is selling, and selling too soon, and then seeing the stock double or triple from there."
Mr. Bernstein argues that manager mandates are tied to narrowly defined benchmarks, which in many cases are faulty themselves. "One of the many troubles with using these indexes as benchmarks is that they are floating craps games," he wrote in a recent client paper.
Mr. Bernstein noted the membership of leading stock indexes, such as the Standard & Poor's 500 and Frank Russell indexes, change constantly.
"What should we be thinking about a manager whose benchmark is one of these indexes but whose portfolio is the same as last year's?" he asked. "Asking an active investment manager to manage against such unstable benchmarks is equivalent to cruelty to animals."
A bigger problem, he added, is deciding against which benchmark to measure performance. "But how does the client know which is the best index to choose?" he asked.
Breaking the umbilical cord
Another issue is that risk differs for the plan sponsor and the money manager. For the plan sponsor, the risk is at the total fund level, not in the relative performance of each portfolio.
From the manager's perspective, risk is "essentially tracking error" -- how far from the benchmark the manager is willing to deviate. Most managers are unwilling to take big bets against the bogey. "The result is the universal use of constraints by managers to hold their optimizers in check," he wrote, which also can make the client less wealthy.
Mr. Bernstein argued managers should be given far more leeway. "The proposal that the manager be encouraged to discard the (voluntary) corset and to take the risk of being wrong and alone is the same thing as suggesting that the whole umbilical cord that ties managers to benchmarks be either broken or at least stretched out," he said.
Referring to a recently published paper by Lee Thomas, senior international portfolio manager at Pacific Investment Management Co., Newport Beach, Calif., Mr. Bernstein said managers should be given expanded scope, taking advantage of the manager's skill.
"In other words, the manager with greater breadth will generate more active return without a corresponding increase in the volatility of that active return," he said. "I find it difficult to believe that a large-cap growth manager or an international value manager that has displayed skill . . . would be incapable of exploiting opportunities in other areas."
This point is where consultants and others part company with Mr. Bernstein.
Thomas M. Richards, principal with Richards & Tierney Inc., Chicago, said that if, say, a large-cap growth manager has a robust investment process, it doesn't necessarily mean the process will work well on international stocks.
What's more, as analysts expand the number of stocks they follow, the manager's "skill becomes diluted," he said.
Don Ezra, director of strategic advice, Frank Russell Co., New York, added: "Managers tend not to be good at everything." While agreeing in principle with much of what Mr. Bernstein said, Mr. Ezra noted the aggregate exposure could create a permanent tilt toward certain asset classes, upsetting the total fund's investment exposure.
"In our view, Peter gains simplicity and conceptual clarity, but we think it would be at the cost of added value and perhaps also unacceptable risk," he added.
Problems with benchmarks
Mr. Bernstein also questioned the use of market indexes or normal portfolios for measuring investment performance.
The problem is that identifying managers that can add value "is a mushy business. All alphas are variable, and establishing statistical significance is impossible because none of us live long enough or, at least, the circumstances that produced alphas in one environment may fail to produce alphas in a different environment," he wrote.
Mr. Bernstein proposed first setting the required return for the total pension or endowment fund, and the degree of volatility with which the client could live.
The same approach can be applied to individual portfolios. "The question should be: How much is this organization contributing to a return in excess of our required return, and what level of volatility?" he explained.
Not only would managers enjoy much greater leeway, but pension executives would not shy away from risk-taking firms, Mr. Bernstein added.
The approach also would benefit large, multiproduct organizations because the client would be able to buy and evaluate performance from the combined product line. Also, plan sponsors could shift assets to different strategies as conditions warrant, without upsetting a strong-performing manager.
Mr. Leuthold also believes larger houses offer the best potential, because they, rather than the client, could move assets around.
Boutiques also can gain by occasionally investing outside of their immediate areas, Mr. Bernstein said.
The proposal does have one major drawback, he acknowledged: "How do we manage diversification" when manager mandates are not clearly defined? But he concludes the issue is not insoluble -- some of the old style distinctions still make sense.
What's more, the pension fund's chief investment officer would have to monitor changes in the asset mix and security classes much more closely, he added.
Substitute is flawed
Mr. Richards said the proposal falls short of two of his firm's four criteria for benchmarks: It is neither investible nor achievable, in which case it's an investment objective, not a benchmark.
Mr. Quinn said he's "very sympathetic" to Mr. Bernstein's views on short-term performance, but found his solution "unrealistic. Say you want to get a 9% return when the stock market is up 20% a year, like it has been for the last five years, or if we had a down market for the next five years. . . . It would be impossible for any equity manager to reach an absolute return."
H. Russell Fogler, principal with Fogler Research and Management, Gainesville, Fla., said he agreed with Mr. Bernstein's arguments, but adds the one area Mr. Bernstein did not discuss is manager fee structures.
Current fees leave managers with no risk, he said. At worst, they lose a client. "I'm saying, create fee structures that make (managers) accountable," Mr. Fogler said.
In a letter to Mr. Bernstein, James Garland, a manager of private family money, added that standard measures of volatility are too imperfect, suggesting instead using the risk of having inadequate cash when needed.
In an interview, Mr. Bernstein replied it remains difficult to "make a judgment about how a manager is doing." Ultimately, it's a subjective decision, he said.
"You have to know your managers, understand their style, understand why they are getting high returns" -- or not -- relative to the risk they take, he said.