A Non-parametric Estimator used the Support Vector Machine for Expected Shortfall
Risk management is one of the top priorities in the financial industry today. The research of quantifying risk is one of risk managements centers. Quantifying the risk of financial time series amounts to measuring their expected shortfall. Asymmetric Power Distribution (APD) is a new family of densities for expected shortfall. The main feature of the APD is that it combines the flexible tail decay property with the asymmetry, which makes it particularly suited for modeling the behavior of financial returns. In this paper, a non-parametric estimator, as a improvement to the traditional estimators, used the Support Vector Machine (SVM) for expected shortfall based on APD is proposed. The simulated studies show that this method can depict the distribution characters of Asymmetric Power Distribution with good results.
support vector machines expected shortfall asymmetric power distribution
Qian xuan
Department of Business Administration Hangzhou Wanxiang Polytechnic Hangzhou P.R.China, 310012
国际会议
太原
英文
352-355
2010-10-22（万方平台首次上网日期，不代表论文的发表时间）