Separation anxiety

Lee Thomas III, chief global strategist at Pimco Advisors, the world's leading bond manager, can't help but feel a little anxious as he sizes up the competitive threat.

"Investors are turning to hedge funds because they want alpha, pure and simple," says Thomas, referring to the return above a market index that reflects a portfolio manager's skill. "Whether hedge funds can provide it is another matter. But the challenge to firms like Pimco is to be an alpha provider. If we fail, we risk becoming irrelevant to the needs of institutional investors."

Like Thomas, active stock and bond managers are increasingly feeling the need to justify their existence -- and their fees. A growing contingent of economists, finance professors, investment consultants and money managers argue that pension funds should identify and isolate beta, the market return, from alpha, the excess return beyond the market. In their view, investors should pay active managers to deliver alpha and use either low-fee indexers or liquid futures exchanges to gain their exposure to beta. The separation of alpha and beta, which are uncorrelated, implies a barbell approach to portfolio construction. For example, 90 percent of assets might be indexed to the betas of different asset classes, and 10 percent could be placed in funds that seek to deliver pure alpha.

Says Robert Ross, director of client services in Europe for Russell Investment Group, a Tacoma, Washington, consulting firm and manager-of-managers: "If you separate alpha and beta, it draws you to hedge funds, because to the extent that a fund takes any active risk, it should do so in the least constrained way that has the greatest opportunity of capturing alpha."

Any portfolio can be broken down into its alpha and beta components. The active risk of a portfolio is the extent to which it diverges from the market or index weights. To generate alpha, a manager must take active risk.

But many pension funds do not distinguish which component of a portfolio's return comes from the market and which portion comes from a stock picker's talent.

Goldman Sachs Asset Management estimates that the typical U.S. pension plan is exposed to 84 percent market risk (beta), 15 percent interest rate risk and only 1 percent active risk (alpha). Even though as much as 50 percent of a fund's assets may be in the hands of active managers, these stock pickers are effectively compelled to act as quasi-index managers. Why? Because tight constraints on a manager's tracking error -- the variance from the market as measured by a benchmark -- restrict his or her ability to outperform the market.

Essentially, many pension funds are paying managers for their active management skill even though their portfolios largely track the market. Says Barton Waring, head of the client advisory group at Barclays Global Investors: "It seems odd that funds should be happy to pay an active fee for what is basically an indexed product. The tracking error constraint means that managers don't have enough freedom to demonstrate their skill and just enough freedom to underperform [the index]."

Certainly, the search for alpha has never been more urgent. Despite the recent comeback in global equity markets, pension funds are still reeling from several years of falling stock prices, which depressed the value of their plan assets, and falling interest rates, which increased their liabilities.

Many plans are hoping to invest their way out of their problems. Consider: Beginning in 2003 the average pension plan of a Standard & Poor's 500 company assumed an annual rate of return on its assets of 8.8 percent. With ten-year U.S. Treasuries currently yielding 4.45 percent, that projection assumes that equities will trounce bonds.

But will they? In the U.S. the average equity risk premium -- the excess return from equities above the rate of U.S. Treasury bonds -- was 4.2 percent a year over 100 years through 2002. Without alpha, assumptions of an 8.8 percent average annual return would be unrealistic.

The £646 million ($1 billion) Shropshire County Council Pension Fund, based in Shrewsbury, U.K., hopes to find greater alpha by boosting its presence in hedge funds. This past summer it slashed its active U.K. equity exposure from 45 percent to 35 percent, with the intention of using the money to make its first investment in hedge funds. (As yet, no hedge fund managers have been chosen.)

"We are looking for ways of cutting equity risk without affecting the potential for return," says Philip Guy, treasury and pensions manager at Shropshire. "Essentially, we are trying to take risk in a different arena."

Theories about separating alpha and beta have emerged as part of wide-ranging debate in academia about asset allocation and portfolio management. Undoubtedly, the most provocative reassessment of investing strategies appeared in Peter Bernstein's March 2003 editorial in Institutional Investor's Journal of Portfolio Management. The widely admired theorist advocated a form of tactical asset allocation that looked to many readers like a form of market timing. In his article Bernstein challenged the notion that pension funds should focus exclusively on the long run. "Which run is the long run?" he asked.

Mark Kritzman, founder of currency specialist Windham Capital Management Boston, entered the debate the same month with another Journal of Portfolio Management article, "Hierarchy of Investment Choice: A Normative Interpretation." Kritzman repudiated Gary Brin- son's famous 1986 paper, which asserted that 93.6 percent of total return variation reflects asset allocation choices.

On the contrary, Kritzman concludes, security selection is by far the most important determinant of return. Using options pricing theory, he contends that swapping the skill of a median stock picker for one who places in the top quartile is four times as valuable as the same trade between a median and top-quartile asset allocator -- such is the variance of returns.

In other words, alpha is almost everything.

"Kritzman's paper is a bit like Martin Luther nailing his 95 theses to the church door, except Mark should nail his study to the doors of investment consultants," says Pimco's Thomas. "It shows that the production of alpha is the most important potential source of return to a pension fund."

Says Kritzman, "The pension fund has to identify the minority of skillful managers and be able to predict the persistence of skill into the future."

Who is helped and who is hurt by the separation of alpha and beta?

Not surprisingly, active money managers who are seen to be nothing more than overpriced closet indexers will be hurt by a strict separation of alpha and beta. Indexers stand to benefit from the trend because they deliver an inexpensive proxy for beta.

"Not everybody has the skill to find alpha," says BGI's Waring. "But if there is skill and resources, it is certainly worth trying to find managers who can deliver alpha."

Indeed, BGI and State Street Global Advisors, the world's top indexers, who together control 84 percent of the $1.7 trillion in global indexed assets, recently have been more aggressive -- and more successful -- in selling their own hedge funds as a complement to their passive strategies (Institutional Investor, July 2003). BGI now posts $5.2 billion in its hedge fund portfolios, while State Street reports $500 million.

Indexing is not the only way that pension funds can match the market, though. More and more institutional investors are turning to liquid-futures markets for their beta exposure.

They can also try to generate return beyond the market by using so-called por-table alpha strategies. As pioneered by Pimco, portable alpha grafts the excess return from active bond management onto an underlying equity index strategy implemented through futures. Global tactical asset allocation, a relative-value strategy designed to capture value across a wide range of asset classes, countries and currencies, can also be used as a portable alpha strategy on any part of an underlying portfolio. Shropshire County Council has hired Goldman Sachs Asset Management to use global tactical asset allocation to move alpha onto its U.S. equities portfolio, while Pensioenfonds Metalektro, a E13 billion ($15.2 billion) Dutch pension fund, is using the GSAM strategy to move alpha onto its European equities portfolio.

Managers-of-managers hope to profit from the separation of alpha and beta, because they sell their ability to identify true skill -- the rare ability to deliver alpha -- among the vast hordes of active money managers.

"After making a strategic asset allocation decision, we then get down to the business of choosing managers who can implement that," says Greg Stahl, head of research in Europe for SEI Investments, an Oaks, Pennsylvania, manager-of-managers. "Especially in today's mar- kets, we are spending more time searching for alpha."

Of course, it is no simpler to find true alpha in the universe of 6,000 hedge fund managers than among thousands of stock pickers. "Hedge funds are nothing more than active portfolios on steroids," says Pimco's Thomas.

Pumped up or pared down, however, money managers have rarely been able to beat the market over time.