VALOR EN RIESGO CON APROXIMACIONES CUADRÁTICAS

Abstract

This paper proposes a quadratic approach for measuring Value at Risk of a portfolio with n assets and m risk factors. It is based on Cornish-Fisher expansion and the delta and gamma of each asset. This Methodology has advantages over simulation techniques based on total valuation of portfolio because it is less computationally intensive, and over lineal approximations techniques that tend to overestimate the risk exposure when the portfolio has assets with a non-lineal relationship with respect to its underlying risk factor, which is normally distributed.