An Analysis of Forecasting Wholesale Market of Selected Offshore Fishery Products in Taiwan

• Main Authors: Sheng-Yu Jian, 簡聖育
• Other Authors: Fu-Sung Chiang
• Format: Others
• Language: zh-TW
• Published: 2009
• Online Access: http://ndltd.ncl.edu.tw/handle/51442071620454833079

碩士 === 國立臺灣海洋大學 === 應用經濟研究所 === 97 === The purpose of this thesis is to investigate whether or not the wholesale market prices of selected offshore fishery products are affected by special events in Taiwan, including traditional festivals, typhoons and other special dates. We use ARIMA time series model to forecast prices of five coastal and offshore fishery products by adding dummy variables for these special events. Previous studies on fishery product prices were based on estimation of demand functions, which help us understand the determinants of market price and quantity of output and the price elasticity. However, by using the time series model, we are able to predict the price fluctuations of fishery products without having to specify the underlying market structure. Therefore, we take an approach different from the traditional demand analysis to forecast prices by including event dummy variables to determine how prices move five days in before the festivals, five days before and after typhoons, the periods for filing income tax and the beginning of the school semesters. We focus on specific fishery products to understand the reason for consumption change. We hope that by making different specifications of price forecasting model for these products we may choose most suitable forecasting model under different circumstances in order to increase forecasting accuracy. By doing so, we may provide fishery administration relevant information for formulating transportation and sales policy to stabilize prices of fishery products. This will prevent fishery product prices during festivals from excessive fluctuation, affecting the rights and interests of the distributors suppliers and consumers. The empirical results are as follows: (1) Using ARIMA models to determine forecasting accuracy and the data on the five selected fishery products prices, we find that the absolute mean error percentage (MAPE) is between 6.594% and 10.45%. (2) By applying fishery price data to forecasting models, this research is able to find out about price movements in Taipei area during festivals typhoons and other specific periods. We also obtain the conclusions based on different characteristics of the fishery products by selecting the most suitable model. Consequently, in strengthening the market distribution mechanism, relevant authorities may use the information to manage and control fishery product prices by adjusting supply to meet changes. This will stabilize prices and increase the welfare of fishermen and consumers at the same time.