Font Size: a A A

The Applicability Research Of Random Time Series Model In International Market Price Forecasting Of Merchandise

Posted on:2008-09-10Degree:MasterType:Thesis
Country:ChinaCandidate:J ZhouFull Text:PDF
GTID:2189360215455278Subject:Finance
Abstract/Summary:PDF Full Text Request
After the entry of WTO, with the further opening up of domestic market, the fluctuation of international market price will affect importantly the stabilization of domestic price. It will take scientific forecast of main commodity price in the international market to carry out the successful docking of domestic and international market price. This can be a referenced guide for making price police in future. In our country, the current price forecasting method of these commodity in the international market is the Seven-Moving Average Method which is not accurate enough. This method is not adaptable to forecasting the international market price which has strong randomicity any more and what is needed is a mathematic model with higher precision and acceptable cost for forecasting. So what we focus on in this essay is to find the most applicable method for the international market price forecasting according to the applicability research of three kinds of statistical forecasting methods and the empirical comparative analysis of several Time Series Forecasting Method which are most widely used.Firstly, based on the analysis of the research about the statistical forecasting methods at home and abroad, we have learned about these research results in the theory field as a good theoretical basis for research.Secondly, we discuss and do the applicability research of three kinds of statistical forecasting methods, including Qualitative Forecast Method, Return Forecast Method and Time Series Forecast Method. Here we mainly focus on the Random Time Series Forecast Method which is already prevalent in the international forecasting field, including ARMA, ARIMA forecast model with or without structural break, etc.Lastly, we do the empirical comparative analysis of three Time Series Forecasting Method we choose for price forecast, including Seven-Moving Average Method, Exponential smoothing method and Random Time Series Forecast Method with structural break. And the result proves that the Random Time Series Forecast Method is the very method which is the most applicable method for the international market price forecasting.
Keywords/Search Tags:Price forecast, Time Series, ARMA Model, Structural break
PDF Full Text Request
Related items