Quantifying optimal capital allocation principles based on risk measures

Abstract

In this thesis we address the issue of covering risks by allocating capital and solving the so-called allocation problem. For this purpose, we provide functional closed-forms representations for each allocation principle built under the general framework developed by Dhaene et. al. 2012. Furthermore, we assess the "correlation effect" which is considered to be the effect of changes in the allocated capital when changing the correlation between the losses, this effect arises when the sources of risk have different variances, otherwise correlations does not play any role in capital allocation results. We develop an R package called OCA which computes optimal Capital Allocations based on some standard principles such as Haircut, Overbeck type II and the Covariance Allocation Principle. Also it provides some functionalities for estimating two of the most popular risk measures: Value at Risk and Expectation Shortfall.. Es tracta de aplicar els procediments de assignació de capital més avançats i descrits a l'article de Dhaene et al. (2011) Journal of Risk and Insurance , amb una base de dades de risc operacional. Es té per objectiu construir un package de R per a aquest finalitat