when investors believe that they are unable to differentiate between the structured and traditional strategies and are neutral in their manager selection abilities, it is still optimal to follow the spectrum strategy. So far, we have developed allocations to hypothetical managers whose expected outperformance fas measured by the information ratio) resembles that of the top-quartile manager in each strategy, and whose tracking error resembles that of the median manager. Additionally, we have explored the investment implications of changing assumptions about the correlations among managers (Table 14.7) and the assumed information ratios (Tables 14.8 and 14.9). To complete the analysis, we will now develop optimal active risk budgets using results from the composite portfolio analysis shown in Tables 14.2 and 14.3. Table 14.10 shows these allocations. In Table 14.10, we continue to assume some skill in manager selection, but the bar is a bit lower. That is, we assume that the investor can develop a top-quartile portfolio of managers, rather than a portfolio consisting of only top-quartile managers. We also include growth and value managers in the analysis. We will abstract from style effects by assuming that the two style benchmarks have the same expected returns, and that the investor can select a top-quartile portfolio of managers (as measured by the information ratio) in each style group. As in our earlier analysis, we again see that it is always beneficial to include a healthy allocation to structured equity managers. For ease of comparison, let's focus on the 200 basis point tracking error target. Table 14.10 shows that an investor can hit this tracking error target with an allocation of 58 percent to structured managers and 42 percent to traditional managers. These allocations compare quite favorably with the figures in Table 14.6. Irrespective of whether our analysis develops optimal portfolios using historical results from individual managers or uses results from composite portfolios, the conclusions are the same: As long as the expected information ratios are positive, TABLE 14.10 Optimal Strategy Mix at Various Tracking Error Targets (1992-2001) Tracking Traditional Active Error Large Traditional Traditional Return Information Level Passive Structured Cap Growth Value (bps) Ratio   0.0% 100.0% 0.0% 0.0% 0.0% 0.0% 0 N/A 0.5 72.0 21.0 0.0 5.2 1.8 60 1.20 1.0 43.9 42.1 0.0 10.5 3.5 120 1.20 1.5 15.9 63.1 0.0 15.7 5.3 180 1.20 1.8 0.0 75.0 0.0 18.7 6.3 214 1.20 2.0 0.0 58.2 0.7 26.4 14.7 234 1.17 2.5 0.0 31.6 10.4 33.7 24.3 269 1.08 3.0 0.0 9.6 17.4 41.4 31.6 300 1.00 3.2 0.0 0.0 23.1 41.9 35.0 312 0.98 3.5 0.0 0.0 34.1 40.6 25.3 328 0.93 4.0 0.0 0.0 45.2 39.1 15.7 344 0.86 4.5 0.0 0.0 56.2 37.8 6.0 360 0.80