In this paper I develop the measurement of market risk for the TES portfolio for a colombian bank, addressing the issue of forecasting Value–at–Risk (VaR) using different multivariate volatility measures: EWMA, orthogonal GARCH, robust GARCH, as well as several models of Value at Risk (VaR) with normal distribution and the student´s t-distribution, evaluating their efficiency by the methods of backtesting proposed by Candelon et al. (2011) based on the generalized method of moments (GMM), along with the test of independence and conditional coverage raised by Christoffersen and Pelletier (2004) and Berkowitz, Christoffersen and Pelletier (2010). The results show that the best specification of VaR for measuring market risk TES portfolio for the colombian banks is built from EWMA volatilities based on the normal distribution, since it satisfies the hypothesis of unconditional coverage, independence and conditional coverage, and the requirements set forth in Basel II agreements and regulations in force in Colombia.
