Nonparametric performance hypothesis testing with the information ratio

This study proposes a nonparametric bootstrap-based test to compare performances between two portfolios in terms of their information ratio. This serves as an extension to the literature that tests performance between two portfolio investment strategies that uses Sharpe ratio. Monte Carlo experiments show that the test has appropriate sizes and is powerful to most of the scenarios. However, the test does not perform well in highly correlated portfolio returns, but is better when the mean of portfolio return is modeled using an autocorrelated process.