基于最小二乘支持向量机的VaR计算方法研究

Study of VaR Calculation Methods Based on Least Squares Support Vector Machine

VaR是使投资风险数量化的工具,旨在估计给定金融资产或组合在正常的资产价格波动下未来可能的或潜在的最大损失。支持向量机是一种基于传统统计学习理论的机器学习算法。波动率作为金融风险的度量,是风险管理中的重要指标。在对Va R的计算中,本文将最小二乘支持向量机与传统的蒙特卡罗模拟法结合,对波动率进行估计。实证分析表明,该方法可行有效。

VaR is a tool making the investment risk quantification, which aims at estimating the possible or potential maximum loss in the future of a given financial asset or portfolio under the normal asset price fluctuations. Support vector machine is a kind of machine learning algorithm based on traditional statistical learning theory. As a measure of financial risk, volatility is the important indicator of risk management. In this paper, when calculating Va R, the least squares support vector machine is combined with the traditional monte carlo simulation method to estimate volatility. The empirical analysis shows that the method is feasible and effective.