Evaluating Hedge Funds with Machine Learning-Based Benchmarks
The explanatory power of multi-factor models used to evaluate hedge fund performance is effectively zero for many funds (so-called zero-R2 funds). We find that, in general, machine learning algorithms offer significant advantages in tracking fund performance, especially for zero- R2 funds, resulting in more precise estimates of fund alphas and hence, more accurate identification of superior funds and fund failures. A key source of the improvement is the ability of machine learning methods to detect changes in the funds’ risk exposures with greater accuracy, and the ability to capture the non-linearities and interactions among risk factors that characterize hedge fund strategies.
Dr. Ashish Tiwari - Bio
Ashish Tiwari is the Henry B. Tippie Research Professor of Finance at the Tippie College of Business, University of Iowa. He is also the director of the Finance Ph.D. program at Iowa, a role he also previously served in from 2007 to 2015. Prof. Tiwari’s primary research and teaching interests are in asset pricing with a focus on the performance evaluation of mutual funds and hedge funds. He has taught courses on a variety of topics including Portfolio Management, Alternative Assets and Portfolio Strategies, Real Estate, Financial Markets and Institutions, Corporate Finance, and Advanced Empirical Finance. In addition to teaching at the University of Iowa, he has also taught in Hong Kong and Auckland, New Zealand. Prof. Tiwari’s research addresses a range of issues in financial economics including the impact of uncertainty on optimal asset allocations, the use of model pooling (combination) to address model error in various decision contexts including the performance evaluation of mutual funds and hedge funds, the design of optimal portfolio management contracts, the impact of market design on order execution quality, and the link between portfolio diversification gains and business cycles. His current research includes projects involving the application of Bayesian machine learning techniques to model the distribution of stock and hedge fund returns. Prof. Tiwari’s papers have been published in several academic journals including European Financial Management, Journal of Business, Journal of Empirical Finance, Journal of Finance, Journal of Financial Intermediation, Journal of Financial Markets, Journal of Financial and Quantitative Analysis, Journal of Financial Research, Journal of Investment Management, Journal of Portfolio Management, Management Science, Review of Finance, and the Review of Financial Studies. His research and commentary have also been featured in various business and popular press outlets including the Financial Times and the Wall Street Journal. Prof. Tiwari has served as a consultant for a number of organizations including the American Stock Exchange. Before entering academe, his corporate experience involved the areas of Corporate Finance and Information Systems Design. Prof. Tiwari received his Ph.D. in Finance from the University of Iowa and MBA (Finance/MIS) degrees from the University of Windsor, and Panjab University.