Performance Analytics Standard Errors

The current finance industry practice in reporting risk and performance measure estimates of assets and portfolios does not typically include reporting the standard error of these estimates: consumers have no clue as to how accurate those estimates are. With the recent work of Chen and Martin (2018), a new approach based on influence functions has been developed to provide an accurate estimate of standard errors of risk and performance of assets and portfolios for returns with both serially uncorrelated and seri- ally correlated returns. This project involves (1) developing a new R package named InfluenceFunctions and (2) integrating the R package EstimatorStandardError in conjunction with InfluenceFunctions into the existing R package PerformanceAnalytics, with the goal of giving PerformanceAnalytics pack- age users more functionality and the option for the first time to report the standard errors of a very wide range of risk and performance measure estimates of assets and portfolios when returns are serially correlated as well as when they are uncorrelated.