基于非线性映射分析的期货逼仓风险判定模型及其应用

选用资金、价格、持仓量等三类指标来反映期货交易中的逼仓风险,采用非线性映射分析将交易中存在逼仓风险的交易日分离成独立的点群.根据离差极小、最大欧氏距离等原理提出逼仓风险阈值计算方法及逼仓风险判定方法.其特点一是运用非线性映射分析等原理给出了逼仓风险阈值的求解方法及风险判定方法,使逼仓风险在酝酿初期就可被及时、准确地判定出来,为期货交易逼仓风险管理提供了预警功能.二是通过与系统聚类结果的对比,验证了非线性映射分析方法的科学性和有效性.三是以豆粕s0409的244个有效样本作为总样本,举例说明了逼仓阈值的求解及风险判定过程.并随机抽取了该合约中的30个样本进行检验,检验结果表明该判别方法的精度高达93.33%.

The Discriminate Model for the Corner Risk in Future Based on Nonlinear Mapping Analysis and Its Application

Select seven indexes from three sorts-capital,price and open interest to reflect the risk of the corner.By the method of nonlinear mapping,the corner trading days are successfully separated with others,and form a independent points group.According to the principle of difference particle and distance maximal,computer method is brought forward to effectively calculate the threshold.It provides theory foundation and concrete method for the judgment and monitor of the corner risk.The characteristics are as follows: Firstly,adopt the principle of nonlinear mapping method and different particle to bring the method of solving threshold and the concrete method of the corner judgment.It solves the problem that there isn't scientific method to judge the corner cases.It can accurately judge the corner risk in the beginning,provides alert function for the monitor of the corner risk in future trade and,further perfects the system of risk control in future trade.Secondly,by comparing with the results of system cluster,verify the science and effectiveness of applying nonlinear mapping method to classify the corner phenomena.This provides a basis for the applications of the method in the future and other finance fields.Thirdly,the study selects 30 samples from 244 effective samples in soul meal contract s0409 to enplane the course of threshold's solving and classifying and the course of the corner judgment.And to verify the precision of judgment by the comparison between discriminate results and the real case,the results show that the precision of the method is up to 93.33%.