CHAPTER 11
Return Dispersion
For an allocator within the CTA space, high daily correlations among strategies often lead investors to believe that all trend followers are all roughly the same. In reality, the strategies vary substantially in their approach, style, positioning, and amount of nontrend strategies that are mixed in. Despite the misconception that there is a high degree of similarity, realized performance exhibits a significant amount of return dispersion. This chapter focuses on discussing return dispersion both empirically and conceptually. First, return dispersion is discussed across different strategy classifications in both the short term and long term. Second, two core drivers of return dispersion, lookback windows size and capital allocation approaches, are examined more specifically. Third, return dispersion is examined from the investor perspective. Finally, return dispersion is examined both theoretically and empirically using CTA returns. This chapter demonstrates how the idiosyncratic effects of parameter selection coupled with the importance of high correlation between programs results in return dispersion over time.
Before discussing return dispersion and its subsequent relationship with correlation, a simple thought-provoking example provides some perspective on the complexities of performance and correlation. Figure 11.1 plots realizations of three artificially generated assets (Asset 1, Asset 2, and Asset 3). At first glance, it appears that Asset ...
