Research On Efficiency Evaluation And Numerical Fitting Of Securities Companies Based On Statistical Method

With the construction of China's multiple-level capital market and the rapid growth of national economy,it has higher requirements and more severe challenges to improve the competitive strength of securities companies.In particular,how to evaluate the efficiency of securities companies is the fundamental problem that they face in expanding the market.In this context,data envelopment analysis(DEA)and other statistical methods are used to evaluate the efficiency of securities companies,through the examples of a better verification,and achieved good results.First,15 securities companies in China were selected as the research samples,and the size of the output and input data of the sample companies in 2016 were taken as the research objects.The traditional data envelopment analysis,super-efficiency data envelopment analysis and generalized data envelopment analysis were applied to evaluate the efficiency of securities companies.Second,when the DEA model and the G-DEA model for evaluation of data are failed,comprehensive evaluation of cross efficiency DEA model can be established through two strategies of a Self-evaluation and Peer-evaluation.This kind of DEA model can distinguish decision units that evaluate securities companies of the same rank well.Third,according to the grey relational degree method in grey system theory,the correlation degree between each factor and performance of securities companies is obtained.According to the rank of correlation degree,we find that the efficiency of the three security companies at the bottom is close.So the cross efficiency DEA model is used to evaluate its efficiency again,the efficiency ranking of the three companies changed.So it's easier for decision makers to make new decisions.Fourth,in the grey model,in order to overcome the disadvantage that the algorithms cannot change parameters,the Simpson GM(1,N)model optimization algorithm based on disturbance factors is proposed.By analyzing relevant data of securities companies,through the influence of the numerical variation of the perturbation factors onthe parameters,the optimal predicted value of the changing characteristic factors can be obtained.According to the average relative error index,the error analysis and comparison of the predicted results show that the fitting accuracy of this algorithm is obviously improved compared with that of the original algorithm.Therefore,the reasonable application of the model of Simpson GM(1,N)based on perturbation factors is of practical significance.In conclusion,the traditional data envelopment analysis,super-efficiency data envelopment analysis,generalized data envelopment analysis,cross-efficiency data envelopment analysis,clustering data envelopment analysis based on the correlation degree and the model of perturbation factor Simpson GM(1,N)are mainly used to evaluate the efficiency of securities companies and used for numerical fitting,and good results have been achieved.