Novel Network DEA Models And The Application In The Performance Evaluation Of Securities Investment Funds

Data Envelopment Analysis(DEA)is a method to evaluate the efficiency of homogeneous decision making units(DMUs).However,the traditional DEA model does not consider the production process within each DMU,we can not know the reason for its ineffectiveness.The essence of network DEA is to open the internal production process of DMU and find out the reason for its inefficiency.Based on the existing research,this paper conducts a series of researches on the network DEA model.The specific research contents are as follows:Fist,Based on the traditional single-stage DDF model under the CRS,a DEA model with serial network system structure is built.The overall efficiency of network DEA model and the efficiency decomposition of each sub-stage are studied.Numerical example shows that the network DEA model has a great improvement in validity and rationality.Second,Based on the two-stage RDM model under VRS,two novel types of network DEA models which can handle negative data are built:the efficiency weight has lower bound and the index weight has lower bound.Numerical example shows that compared with the model with lower bound of index weight,the results of the model with lower bound of efficiency weight are more reasonable.Finally,Empirical analysis.A novel network DEA model with lower efficiency weight is applied to evaluate the performance of 70 mixed funds from 2016 to 2018.The results show that compared with the traditional DEA model,the novel model can differentiate the fund performance evaluation better.