Top Bottom   Average Median Quartile Quartile Structured Manners (64 Managers)         Active return (bps) 43 52 92 -4 Tracking error (bps) 209 221 266 147 Information ratio 0.26 0.28 0.44 -0.02 Pairwise correlation 0.08 0.08 0.27 -0.1 Traditional Managers (561 Managers)         Active return (bps) 53 53 201 -120 Tracking error (bps) 821 769 971 619 Information ratio 0.05 0.07 0.27 -0.16 Pairwise correlation 0.13 0.14 0.36 -0.1 2001, inclusive. We included all managers with at least 24 quarters of performance history. As discussed earlier, we reduced the number of managers in our database further by restricting our attention to low and high tracking error managers. Of course, our methodology might misclassify some managers. For example, a manager could intentionally switch between low and high tracking error regimes as part of the active decision-making process. If the tracking error levels in each regime are sufficiently different, and the manager spends an insufficient amount of time in the high tracking error regime, then we could mistakenly classify the manager as "structured." Unfortunately, we do not have sufficient data to easily discern such regime-switching behavior. This caveat notwithstanding, we do feel that our database is rich enough both to classify managers and to produce historical differences that are sufficiently interesting for further discussion. Table 14.1 shows the summary performance and risk characteristics for each group of managers. The table shows the historical average, median, top-quartile, and bottom-quartile figures for four performance and risk characteristics: active return, tracking error, information ratio, and pairwise correlation. We independently calculated these quartile cutoff points for each risk or performance characteristic. For example, the structured manager with the median active return may not be the same as the manager with the median tracking error. The performance and risk figures in Table 14.1 are quite revealing, and indicate why selection among different types of managers is such a challenge for institutional investors. Let's look at the performance record first, and then consider the differences in risk. Historically, the average active return was quite similar for structured and traditional managers. On average, traditional managers had an active return of 53 basis points, while the active return for structured managers was slightly smaller at 43 basis points. The median active returns were even closer at 52 basis points for structured managers and 53 basis points for traditional managers-despite significantly lower risk of the structured managers. Given that traditional managers usually charge higher fees, it would be hard to argue that, on average, traditional managers have provided higher risk-adjusted excess returns net of fees.