Non-gaussian asset allocation in the federal thrift savings plan

Historical data suggest that returns of stocks and indices are not distributed independent and identically Normal, as is commonly assumed. Instead, returns of financial assets are often skewed and have higher kurtosis. In this study, we investigate how the optimal investment choices in the federal government´s Thrift Savings Plan (TSP) change when a non-Gaussian factor model for returns, generated with independent components analysis (ICA) and following the Variance Gamma (VG) process, is used in place of the usual Normally-distributed returns model. Using back-testing and simulation, we hope to show how this method could benefit the more than 3 million TSP participants in achieving their retirement savings objectives.