Price Formation And Its Dynamic In Two Types Of Multi-unit Auctions

As a pricing mechanism, auction has been extensive research and attention. Many previous papers study on a single unit auction. With economic development, Multi-item auction has become a research hotspot in recent years. In this paper, we study two types of multi-unit auctions.One is use a data set of FOU auctions to explore the bidding behavior in simultaneous auctions and obtain that it is optimal to use the cross-bidding strategy. Further, we use the ARIMA(2,0,2) model to show that the auction prices are correlated. By constructing multiple linear regression equations, we analyze the effect of some factors on the auction prices. It suggests that the number of auctions has a negative impact on the auction prices.The other is use flower auction data from KIFA to study the price formation process in Sequential auctions. Using ARIMA model, we find that the auction price is highly correlated and the prices follow an ARIMA(1,0,0) process. Based on it, we construct multiple linear regression equations and analyze the effect of some factors on the auction prices. The results show that the average auction price on the previous day and the first five auction prices today have positive impacts on the tendency of the auction price today. Meanwhile, the ratio of demand to supply quantity also has impact on the auction price significantly.