A New Method To Backtest ES

The main issue of this paper is the backtesting of the risk measurement.In the previous study,the researchers proposed a method of using statistics to backtest the ES,but the result is not ideal and the power is very low.So in this paper,I research another backtesting method.The method is to divide the tails into equal parts and study the property of each tail.It is found that they have multiple distributions.So we can use Pearson’s square test to indirectly backtest the tail.In other word,we have find a new method to backtest ES.Then we use a combination of Monte Carlo simulation and empirical analysis to compare the result and power of the two methods.Finally,we found that the multi-distribution method is more applicable to extreme risk models.And in the empirical study,the multi-distribution method also has greater power.Therefore,we have found a risk measurement backtesting method that is more suitable for China’s securities market,and provides a good reference for financial institutions and regulatory authorities.