Applying Statistical Models To International Crude Oil Price Forecasting And Natural Gas Pricing

This thesis briefly analyzes the factors affecting the international crude oil prices aswell as existing price forecasting model, and outlines the current natural gas pricingmechanism, and the goals of China’s natural gas pricing reform.Select WTI crude oil spot monthly price time series to establish the predictionmodel. Choose data from June1986to April2012as the training set to establish themodel and data from May2012to March2013to test model predicted results. From thepoint of view of time series analysis to establish the model: season-harmonic predictionmodel, EGARCH model, exponential smoothing model. Explain the principle of thethree models compare the predictions in the empirical analysis section, conclude thatseason-harmonic prediction model to perform better.Use season-harmonic prediction model to forecast the monthly price of WTIcrude oil spot from April2013to December2015. Establish the natural gas price andalternative energy fuel oil price-linked pricing mechanism. Use polynomial lag model tofit the WTI spot price of crude oil and fuel oil prices and forecast the Shanghai fuel oilprices from April2013to December2015, and in accordance with equivalent calorificvalue conversion method, the Shanghai fuel oil prices translated into equivalent calorificvalue gas prices. Then multiply some value to get final natural gas price.Another pricing mechanism of natural gas pricing mechanism is linked to the ex-factory price of gasoline. We calculate the price of natural gas under such a pricingmechanism. And compare the price of natural gas pricing mechanism. Including naturalgas pricing mechanism linked to alternative energy fuel oil prices, linked to the ex-factory price of gasoline, current "cost-plus" and "market back of the net value method"contrast mechanism. Explore the direction and goals of natural gas pricing mechanismreform.