Price Forecasting And Hedging Based On Artificial Neural Network

With the development of economics,lots of enterprises learn the idea of avoiding risk.The question of using hedging reasonably for risk diversification has caused the attention of academia,industry and government.When the market is complete,future market can offer sufficient tools for a company to hedge.When the market is incomplete,the varieties of future contracts are limited,causing the difficulty of company’s hedging.Thus,the hedging strategy in incomplete market is significant.Research shows that the basis of hedging in incomplete market is the price transmission along a supply chain.Establishing an effective model to describe the price transmission and obtain a good price forecasting is the key to improve the hedging performance while current price forecasting models can’t get precise predictions.This paper aims to establish a model that can achieve better price forecasting to take the challenge which hedging in incomplete market faces.This paper uses different artificial neural networks(ANN)and linear method to simulate the price transmission of outputs and inputs on the supply chain so as to realize price forecasting,building a minimum variance hedging model in an incomplete market using the data obtained from prediction.The empirical analysis of two supply chain cases shows that the relationship between the input and output prices trends to nonlinearity,and the performance of hedging by use of the non-linear technique based on Elman neural network is superior to that of the linear model between the output and input prices and the one based on BP neural network.