Estimating Flexible, Fat-Tailed Conditional Asset Return Distributions

Abstract:     
We provide a new alternative to thin-tailed regression models by introducing robust numerical methods, based on the minimum relative U−entropy (MRUE) principle, to estimate heteroskedastic, fat-tailed, flexible univariate probability density functions, conditioned on a number of explanatory variables. We benchmark our method against state-of-the-art asset return models on the 30 constituents of the Dow 30 and find that our models outperform the benchmarks out-of-sample.

Number of Pages in PDF File: 44

Keywords: Minimum Relative U−Entropy, Conditional Probability Distribution, Fat-tailed, Power-Law Distribution, Heteroskedastic, Financial Data, Asset Returns

working papers series