An Advanced Approach for Improving the Prediction Accuracy of Natural Gas Price

As one of the most important commodity futures, price forecasting natural gas futures is great significance for hedging and risk aversion. This paper mainly focuses on pricing which considers seasonality fluctuations. In order to study this issue, we propose a modified approach called six-factor model, in influence seasonal fluctuations are eliminated every random factor. Using Monte Carlo method, first assess comparative analyze fitting ability three-factor model out sample data. It found that has better performance than prices strongly influenced by winter effect. We then apply proposed predict year 2019. have weak upward trend coming relatively volatile winter.