add value, investors with significant passive exposures are effectively creating drag on their overall portfolio performance. Finally, because the spectrum strategy diversifies the active risk budget, we believe that investors can achieve a higher return per unit of active risk by including structured equity products in their portfolios. These themes will be explored in detail. We'll first examine the historical track records of structured and traditional active equity managers. We'll then explore the methodological differences that drive these performance differences. Later, we'll show how investors can apply these findings, together with active risk budgeting techniques, to their large-cap U.S. equity portfolios and reach some more general conclusions. COMPARING STRUCTURED AND TRADITIONAL MANAGERS Many investors implement their long-term asset allocations to large-cap U.S. equities by combining passive and traditional active management. Because we believe that investors should also include structured equity in the mix, let's review the historical risk and performance characteristics of traditional and structured managers. Viewing these historical results will motivate further discussion of the methodological differences that distinguish these two management styles. For our analysis, we will use historical tracking errors to segregate managers, classifying lower tracking error managers as structured, and higher tracking error managers as traditional. Market conventions place structured equity managers in a target tracking error range of 100 to 250 basis points. Given that realized (or historical) tracking errors could exceed targets, we identify structured managers as those with realized tracking error levels between 100 and 300 basis points. Market convention also suggests that traditional (or concentrated) managers have tracking error targets-to the extent they are benchmark sensitive and have tracking error targets-in excess of 600 basis points. Of course, realized tracking error levels can also undershoot targets. Hence, we define traditional managers as those with realized tracking errors in excess of 500 basis points, but below 1,500 basis points. (The upper bound is meant to exclude managers who may have significant holdings in other asset classes, such as small-cap equities, international equities, or bonds.) We judged it too difficult to classify managers with realized tracking errors between 300 and 500 basis points; such managers were thus omitted from further analysis. However, our results are not sensitive to omitting these managers. Table 14.1 summarizes our results. Using the Plan Sponsor Network (PSN) database,2 we constructed a set of quarterly time-series returns for 1,052 large-cap U.S. equity managers. The returns, which are gross of fees, cover the period 1989 to 2The Plan Sponsor Network is a database of institutional manager returns. These returns are gross of fees and contain both self-selection and survivor bias. That is, only managers who choose to submit are included (presumably those with better returns), and managers who fail or merge are dropped. Thus, our median results may actually be closer to the 55th percentile results. Nonetheless, despite these biases (which affect both manager styles), we believe the comparisons between structured and traditional managers are valid.