Non-Parametric methods: An application for the risk measurement

Currently, the financial institutions are exposed to different types of risks, which has increased the need for new analytical instruments for the risk management, being one of most developed the Value at Risk (VaR). There are different methods of calculation; however, as it was affirmed, there exists an increasing need to be provided with analytical tools that shape the behavior of the financial markets in a more accurate way, in this sense, the present work proposes the calculation of the VaR using non-parametric methods (kernel estimator) for portfolios characterized by heavy-tailed distributions. The evidence shows that the behavior of the changes of a portfolio’s return can be estimated in a more precise way since there is not assumption about the distribution, as in case of a normal distribution.