When plan sponsors aren’t looking for new asset managers these days, they’re apt to be searching their souls. How can a few years of down markets cause such a crisis for money management when they follow a decade of unprecedented boom?
One explanation may lie in the way that so many pension funds structured their fund manager relationships: Paying scant regard to the true goal of any pension fund -- meeting its liabilities -- plan sponsors expended enormous effort on strategic asset allocation. Once they’d settled on a split between stocks and bonds, they apportioned funds to a smorgasbord of specialist managers. And numerous funds adopted core-satellite models for stocks, turning perhaps 40 percent of their equities over to passive managers and divvying up the remainder among the specialists.
But if this appeared to be bold at the time, it may actually have been too conservative. The funds and their consultants permitted such restrictive tracking error vis-à-vis benchmarks that the specialist managers had their hands tied. The paradoxical result even now is that many pension funds take too little active risk, according to Mark Carhart, co-head of quantitative strategies at Goldman Sachs Asset Management.
“Our estimate is that the average pension fund is taking about 150 basis points of active risk above their strategic asset allocation benchmark,” he says. “To be blunt, that just isn’t enough.”
In effect, what a pension fund does when it uses a core-satellite approach and restricts additional risk to only about 150 basis points through tracking controls is create a quasi-index fund.
Even worse, contends Anthony Foley, head of the Advanced Research Center at State Street Global Advisors, constraining tracking error in a rigid fashion is a naive approach to risk management. “We have been using something called a regime-switching model that allows us to calculate how a given strategic asset allocation will behave in different market environments,” he says. “When markets are trending up, the core-satellite approach does capture market return; but in fact, the risk is huge. That’s because the correlation between the assets and the managers soars.”
Foley, Carhart and a growing band of fund managers believe that the core-satellite model should be thrown in the waste bin. What particularly irks Foley is the segregation of stocks by style and by capitalization. “All of our work suggests this is meaningless,” he declares. “But you shouldn’t need a fancy quant model to tell you that. It’s plain common sense. Why should a fund manager suddenly be unable to assess the merits of a chemical firm he has followed for years because its price-to-book ratio changes and it shifts between arbitrarily defined growth and value stock universes?”
Another reason for questioning the prevailing process is that it produces an agency problem. The agent -- that is, the fund manager -- may have a different objective from the principal -- that is, the plan sponsor. Ostensibly, the fund manager’s brief from the plan sponsor is to seize opportunities to add value. But once a fund manager gets appointed, the No. 1 priority may be to keep that coveted mandate: Play it safe and don’t jeopardize a good thing by deviating too far from the benchmark.
“The manager is incentivized to lie” about how limited the current opportunities in his specialty may be, contends Lee Thomas, chief global strategist at Pimco. “This issue worries me because pension funds aren’t able to make an assessment of whether there are opportunities for their managers or not. That’s why they put them in boxes and give them small amounts of tracking error to work with,” he says. “Nobody is optimizing the alpha [or manager] bets at the plan level, so a lot of potential return is left behind.”
Fund managers deserve some of the blame. Mark Kritzman, managing partner of Windham Capital Management Boston and a leading commentator on investment issues, plans to publish a paper in the spring 2003 issue of the Journal of Portfolio Management (a publication of Institutional Investor, Inc.) that seems sure to start an academic firestorm over manager culpability. Kritzman challenges one of asset management’s eternal truths: that asset allocation makes the most important contribution to overall return -- 94 percent of the variation in returns among U.S. pension funds, according to portfolio manager Gary Brinson’s famous 1991 study.
Kritzman argues that Brinson got it wrong because he measured how pension funds’ underlying portfolios actually did; instead, he maintains, Brinson should have considered what the fund managers hypothetically could have done, through countless simulations.
Kritzman’s basic contention is that Brinson’s findings fell prey to an all too familiar agency problem: business risk -- the hazard of losing a manager mandate. Retaining the mandate is more important to a manager than seeking the best returns available by pursuing opportunities that diverge from the benchmark, he says.
For U.S. equity portfolios, Kritzman found, stock selection was four times more crucial in determining returns than asset allocation. And he reached a startling conclusion. “The money management industry has been so caught up with the issue of business risk that it has artificially compressed the rates of return available from the capital markets,” he says.
Kritzman’s findings are bound to be controversial, but the message emanating from the cutting edge of investment research is that pension funds should stop worrying so much about strategic asset allocation and learn to love active risk. Portfolio manager Peter Bernstein, author of the classic Against the Gods: The Remarkable Story of Risk, sees Kritzman’s paper as a wake-up call for pension funds. “The only way to deal with a diminishing equity risk premium,” argues Bernstein, “is to find stock pickers.”
Inspired by the bear market, some fund managers are introducing new and refurbished products. Pimco’s Thomas proposes a new method of balanced management to allow funds to make better, though not necessarily bigger, risk bets. Besides overseeing $10 billion in global bonds, Thomas has become a kind of internal investment consultant at Allianz Dresdner Asset Management, or ADAM, which absorbed his bond house in 1999. And lately, he has undertaken a sweeping performance-attribution exercise, listening to fund managers describe how they run money, then pulling apart their track records to identify the ones who can add alpha.
Pimco’s own revamped and soon-to-be-launched system of balanced management will operate roughly like this: The firm will agree to a benchmark with a client; it could be an absolute target, like U.S. Treasury bills, or a hybrid global equity and bond index. A global policy committee will determine asset allocation. Thomas will then shift active risk among asset classes and fund managers by shrinking or expanding their risk budgets, that is, permit them more or less leeway in tracking error.
Although the main contribution to return should still come from the fund managers, Thomas is to be allocated a small portion of the capital to rebalance the client’s portfolio back to the original asset allocation, reduce unintended concentrations of risk and leverage active positions, using derivatives. “The idea is to optimize the alpha bets in the overall portfolio,” he explains. “In the current structure no one does that.” Pimco’s global fixed-income product, which Thomas runs in like fashion, outperformed its benchmark by 270 basis points last year.
Scott Donald, head of product development at consulting firm and fund-of-funds manager Frank Russell Co., like- wise believes that the core-satellite model needs to be overhauled. He says that Russell’s researchers pay little attention to whether fund managers are growth- or value-oriented but instead try to find adept stock pickers. Russell’s portfolios of managers can take more active risk than typical managers, Donald says, because Russell can uncover enough managers with uncorrelated alphas to ensure diversification.
Nevertheless, Thomas maintains that ADAM still has an edge. “Russell would say that it can choose any manager in the world, whereas Thomas only has the managers in the Allianz Dresdner group to chose from,” he says. “I would say I have more information and a far more candid assessment of the opportunities.”
Donald is not convinced. “Moving around a risk budget sounds a lot like market timing to me,” he says. “If Lee thinks he can do that, then good luck to him. If he is successful, maybe we will hire him to do it for our portfolios.”
