ABSTRACT
This document evaluates the performance of 36 different approaches for the estimation of the Value at Risk (VaR), parametric and non-parametric, of a representative portfolio for 5 Latin American countries (Argentina, Brazil, Chile, Colombia and Peru) with and without the day of the week effects (DOW). After finding the approach that better captures the risk set out for each portfolio we found that the inclusion of the day-of-the- week effect do not improve the risk metrics in Brazil and Argentina; but it does contribute in the case of Colombia, Chile and Peru.
Keywords
Value-at-Risk, GARCH, Mean average, Historical Simulation, Back testing, Latin America, Parametric Approach, Historical Filtered.