Multi-Characteristic Product Price Research Based on GSADF-BP Model

Abstract

According to hedonic price theory, people’s demand for product is based on the product characteristic. Pu’er tea is geographical indicator product of Yunnan province, which has edible, medicinal, drinking, investment and cultural value. It is a typical multi-characteristic product, the price bubble possibility is large, but the degree of price bubble has not been tested, and the detection effect is not clear. So, this paper applied four Pu’er tea products as study case, generalized sup ADF statistic model (GSADF) to respectively detect price bubble and Back-ProPagation Network model (BP) is used to forecast the four products price in whole sample period and non-price bubble period, so as to verify the detection effect of GSADF model. According to the GSADF result, three products have price bubble and one product have no price bubble. In the whole sample data price prediction, the best forecast effect was product 3, followed by were product 4, product 2 and finally was product 1. In the non-price bubble period sample data price prediction, the best forecast effect was product 4, followed by was product 1, and the last one was product 2. Multi-characteristic product which cover various industries and fields, Pu’er tea is not a unique case. As a pioneering research, the methods and ideas in this paper can be extended to other industries and fields, so as to conduct in-depth analysis of multi-characteristic product in the market and provide suggestion and guidance for the decision making and behavior of relevant stakeholder.

Keywords

• Multi-characteristic product
• GSADF model
• BP model