Risk Research On Large Consumer’s Power Purchasing Portfolio Based On Generalized Utility Theory

Large consumer’s direct power purchasing is one of the most important issues of promoting electricity market construct. During direct power purchasing, large power consumers and gencos can carry out direct bilateral transactions. As we all know, electricity is an inelastic commodity with price strongly volatility. For the different price fluctuations in the different electricity market, large consumers need synthetically consider various uncertain factors and allocate reasonable purchasing ratio in the multi-market to get a given expected profit value while the risk is minimal.Risk identification is a prerequisite of risk analysis, In this paper, the financial risk, trading patterns and transaction type is described during large consumer’s direct power purchasing. Accurate price prediction is the prerequisite of power purchasing portfolio risk analysis. For the characteristic of strong volatility of spot market price, it is firstly decomposed by wavelet transform, then the transfer function model is used to forecast approximate price and detail price respectively and summed as the final result. Further, the generalized utility theory is introduced into the modern portfolio theory. A new comprehensive risk measurement index, in which both of significance level, risk attitude parameters and weight coefficient are taken into account is put forward and it is set as the short-term power purchasing portfolio model’s objective function. By use of the previous electricity price forecast result, the optimal electricity purchasing proportion of multi-market is obtained. Due to the real-time electricity price is more vulnerable to the impact of balance supply, simply mean prediction usually can’t reach satisfied results which has a significantly heteroscedasticity. This paper adopts the GJRSK-M model to fit time-varying characteristics of residual series’second, third, and fourth moments and the parameter estimated is based on the Gram-Charlier series’four moments expansion of the normal density function. The general utility theory is expanded to higher moments utility theory which is set as the spot power purchasing portfolio model’s objective function to get optimal electricity purchasing proportion of the multi-market. Calculation results show that both of the comprehensive risk measurement index and higher moments utility theory can more flexible and accurately portray different large consumers’ subjective attitudes. Meanwhile, the above models can provide decision support on power purchasing and risk measurement for the large consumers.