The Study Of Spectral Risk Measures Of Credit Risk

As one of important tools of risk measure and management of current financialfacilities, VaR method combines potential loss and probability of known Portfolios ina simple way to describe the risk of Portfolios. This method shows its specificsuperiority in risk management and control and is broadly used. However, as a singleand general statistical measure of ordinary financial risk, VaR method reveals itstheoretical limitation on its absence of general subadditivity and description to tailinformation. Against this limitation, in recent a few years, a new risk measureframework is proposed, that is coherence risk measure. Under this framework, manyrisk measure technologies are established, one of which is spectral risk measure.The spectral risk measure, which combines the distribution of profit and losswith the subjective risk aversion from the investors, provides an available choice inrisk measuring properly and effectively.On the basis of introducing the definition and superiority and limitation of VaR,this paper introduces the formation of coherent risk measure framework and spectralrisk measure, and applies spectral risk measure into credit risk measure. By somebasic mathematical methods, some essential characteristics of credit risk and theapproximation of discretized risk spectrum are deduced under credit risk condition.Thus an available method in risk measure analysis is obtained.Based on the above method, the operational performance of spectral risk measureis analyzed by an example. The analysis results show that in measuring credit risk byspectrum, for one portfolios, different investors will form not the same spectral riskmeasure and will make different invest decisions. This provides an interpretative pathfor the diversity of investment behavior in practical operation. Also, by comparingspectral risk measure to VaR and Expected shortfall, it is found spectral risk measurecan obtain a measured value of VaR or Expected shortfall tantamount to the higherlevel of confidence by using a set of simulated value. So the method of measuring thecredit risk by spectral measure has both practical and theoretical value.