Forecasting compositional risk allocations

Abstract

We analyse models for panel data that arise in risk allocation problems,when a given set of sources are the cause of an aggregate risk value. We focus on the modeling and forecasting of proportional contributions to risk. Compositional data methods are proposed and the regression is flexible to incorporate external information from other variables. We guarantee that projected proportional contributions add up to 100%, and we introduce a method to generate confidence regions with the same restriction. An illustration using data from the stock exchange is provided.