"It may be significant that a less risky group of funds are emerging at this relatively high point in the market."
That sentence ended the analysis of the second-quarter returns. Since then, the third quarter witnessed the steepest decline in more than 20 years. It seems appropriate at this one-year anniversary of the Omega return to review what it did and did not do.
The reason offered a year ago for the Omega return was that risk-adjusted returns based solely on manager data understated the underlying risk because the previous five years had very few months of negative returns. I believed the incorporation of style analysis offered the hope of more reliable risk estimates.
The top three funds recommended one year ago were Putnam New Opportunity, T. Rowe Price Science and Technology, and T. Rowe Price New Horizons. Hardly conservative funds. It wasn't until last quarter that funds with income components found their way into the top 10. (See table on page 30.) That was propitious timing, but did it save investors from losses in the third quarter?
Third-quarter performance for the top three Omega funds last quarter (American Century Income & Growth, Vanguard Institutional Index, and Vanguard Growth & Income) averaged a loss of 10% for the quarter, while the Vanguard 500 Index fund lost only 8%.
Of course, many funds lost more than that. If one stayed with the initial top three funds, the loss for the quarter would have been 35%. So while it could have been much worse, the Omega return did not enable one to beat the market in the downturn, and it required a willingness to switch funds at some point in order to prevent substantial losses. Longer-term Omega performance for the 18 years ended Sept. 30, 1997, were shown in the Nov. 10, 1997, issue of Pensions & Investments.
A crucial determinant of the reliability of the Omega return is the R squared, which indicates the degree to which the fund's style can be identified. The top performing fund this quarter was Kemper Dreman High Return fund: 92% of the return, and therefore the risk, is captured by the manager's style. Yet the fund earned an excess return of 3.6% over a passive strategy using style index funds. Part of the reason for this performance is because the manager took less risk than the passive style benchmark (downside beta = 0.71). Federated Utility, on the other hand, had an R squared of only 68%. The Omega measure for this fund is not capturing enough of this manager's style to be reliable
Why didn't the Omega return switch into bond funds or money market funds? Fixed-income funds are not part of this analysis. And, performance measurement is not intended to replace asset allocation. The Omega return is conceptually an improvement in estimating which funds performed best for the past five years on a risk-adjusted basis, but that information is inadequate for making an asset allocation decision.
Asset allocation based on some period in the past repeating itself, is doomed to failure. The model we have developed at the Pension Research Institute requires an estimate of the current valuation of the market. The decision-maker should at least be able to answer this question in order to justify an active asset allocation strategy.
LCG Associates in Atlanta, and Mobius Group in Research Triangle Park, N.C. offer models using the Omega return. The fund data and the statistics in the table were generated on the LCG model.
Frank Sortino is director of the Pension Research Institute, Menlo Park, Calif
