Nonlinearity in Portfolio Performance Evaluation

I present a methodology for evaluating the performance of fixed-income investment managers, over the last ten years. A cross-section of such managers reveals that alpha does not reflect most non-market performance, unless a regime-switching model is used. The quest is for arriving at qualitative statements that may be remotely-even related to probability of outperformance, given complications in the industry- standard use of information ratio. Managers tend to react more abruptly to markets, the lower the quartile in which their performance has fallen. It also appears that managers, who do less, tend to outperform, supporting passive management. However, specific exposures to principal components of indices are germane to active outperformers. At the same time, these exposures cannot be gleaned from linear performance evaluation