基于SV模型和KLR信号分析的期货逼仓风险预警模型

采用期货价格波动风险、持仓量波动风险和基差波动风险三大因素衡量期货交易中的逼仓风险。采用KLR信号分析法、SV随机波动模型、GARCH模型等构建了逼仓风险预警模型，为期货市场中的逼仓风险预警监控提供了理论依据和具体的方法。本研究最为关键的突破创新有以下两点：一是结合KLR信号分析法、SV模型等建立了期货交易中的逼仓风险预警模型。经1980～2006年的文献检索发现本研究是针对我国期货交易逼仓风险预警问题首次提出的定量分析模型，解决了我国期货界对逼仓风险至今没有预警方法的难题。二是在本研究中将现有KLR信号分析法与风险指标预测模型相结合，突破了以往使用KLR信号分析方法只能采用历史数据对未来情况进行预警而造成的预警误差。经实证分析，本研究的逼仓风险预警模型精度可达76．67％。

The Corner Risk Prewarning Model of The Futures Market and Application Research Based on SV Model and KLR Signal Approach

The present research begins with fluctuation risk of futures price,fluctuation risk of hold,fluctuation risk of basis to scale the corner risk in futures trade.By adopting the KLR signal approach,according to the principle of the jar-to-signal ratio,determine the index threshold of the three risk index.By using the SV stochastic fluctuate model and GARCH model to forecast the three risk index and combining the signal of corner risk and the probability of corner risk to judge the pre-warn rank of corner risk,to build the pre-warn model of corner risk,which provide theory and method for the pre-warn control of corner risk in futures market.The most important characteristics of this research are as follows: Firstly,combining the KLR signal approach,SV model and GARCH model to build the prewarning model of corner risk in futures market.We retrieved all the documentation from 1980 to 2005.And we found that this is the first ration analysis model aiming at the problem of corner risk pre-warn in our country.This research solves the difficult problem of having no pre-warn method to corner risk in the present futures of our country.Secondly,the present research combines the KLR signal approach and the forecast model of risk scale to break through the problem of the low pre-warn precision in the past KLR signal approach.In this research,the risk index is up to 96.6% and the pre-warn precision of corner risk is 76.67%.