structured equities, with the allocation coming primarily from the passive portfolio. Investors should allocate significant amounts to passive products only when their tracking error targets are quite low. In our examples, a large allocation to passive management is appropriate only when the tracking error target for the entire U.S. equity portfolio is less than 100 basis points.13 CONCLUSIONS A basic issue that most institutional investors face is how to allocate assets between active and passive strategies. Many investors adopt a barbell approach in which they achieve their active risk targets by blending traditional, high-tracking-error active managers with passive index funds. However, by including passive management, investors are forgoing excess returns on what may be a significant portion of their portfolios. Most investors would benefit from putting this capital to work in structured equity programs; that is, most investors can achieve potentially significant improvements in excess returns and information ratios by reducing their passive allocations and replacing them with allocations to structured equity. By allocating risk across the active risk spectrum, investors can significantly enhance the expected active performance of their U.S. equity portfolios. The actual optimal risk allocations will depend on investor assumptions about the ability of active managers to outperform their benchmarks.14 Using historical separate account data, we have shown that the median and top-quartile information ratios for structured managers have exceeded those of traditional managers. This result is not surprising: Given their lower tracking error objectives and relative freedom from the no-short constraint, we expect realized information ratios to be higher for structured managers. (This result is consistent with the emerging literature that explores performance differences in mutual funds.) Thus, investors should not be alarmed by the relative differences in historical information ratios. If these differences persist, then the practical implication is that investors will continue to need traditional managers within their active manager rosters-although possibly with somewhat smaller allocations. Our analysis also shows that manager selection is extremely important among traditional managers. Thus, when developing a portfolio of traditional managers, investors should balance the benefits of diversification against the higher fees and monitoring costs that come with manager proliferation. Our main conclusion, however, is that investors should allocate risks across the entire active risk spectrum. Moreover, when moving from a barbell approach to a spectrum strategy, the allocation to structured managers is more likely to come 13Note that 100 basis points of tracking error should have little impact on the risk of the overall plan, given the small amount of active risk vis-a-vis the total risk in equities. "Software has been developed that can help clients determine optimal risk allocations based on their own assumptions for risks, correlations, and expected returns across various managers and management styles.