An Empirical Investigation of Heavy Tails in Emerging Markets and Robust Estimation of the Pareto Tail Index

In this work we analyze and compare the performances of VaR-based estimatorswith respect to three different classes of distributions, i.e., Gaussian, Stable and Pareto, and to different emerging markets, i.e., Egypt, Qatar and Mexico. This is motivated by the evidence that there are points of distinction between emerging and developed markets mainly relating to the speed and reliability of information available to investors.We propose a computational Threshold Accepting-VaR based algorithm (TAVaR) for optimally estimating the Pareto tail index. A Monte Carlo bias estimation analysis is also carried out by comparing our proposed methodology with the Hill estimator and a variant of it.