Model Uncertainty Of Risk Measure

Risk measurement is a central element of financial decision-making and risk man-agement.In this paper,risk ratios is used to study the model uncertainty of risk measure-that is,the potential for risk measure based on different underlying risk forecast mod-els to provide inconsistent outcomes—increases significantly during market turmoil.So risk readings should be interpreted and evaluated with caution.In addition,VaR and ES are significantly affected by model uncertainty when in-creasing the value of p.Therefore,it is not advisable to raise economic capital by increasing the level of confidence simply.The Basel Committee proposed to replace VaR(p = 0.99)by ES(p = 0.975)for the quantification of market risk in 2013.The results show that the latter is less sensitive towards regulatory arbitrage and model uncertainty.Sensitivity analysis shows that the divergence across models is not dependent on any particular model.Finally,this paper suggests that one would be better off estimat-ing a model at a low probability,and scale it up to the probability needed—probability shifting.