Behavioral Experimental Study Of Information Acquisition And Bidding Risk In Auctions

Information is one of the most fundamental and core concepts in auction theory. Classical auction theory assumes that bidders own the accurately information of the object being auctioned, and these infomation being exogenously given. However, in real life, bidder's private value for the good is often uncertain, and bidders need to spend resources to obtain more information about their valuation. In recent years, an increasing number of studies have attempted to bring endogenous information acquisition into the auction framework, but such studies most remain at the theoretical level, rarely researchers experimentally investigate the information acquisition behavior of bidders in actual auctions.In this dissertation, we describe an information asymmetry problem in two heterogeneity items sequential auction. Based on dominant strategy analysis given by previous study, we theoretically discuss the information.valuation issue in this problem. Further more, by placing an information auction before the sequential auction, we design and carry out a sequential auctions experiment which containing an information acquisition process. Experimental results show that the majority of bidders tend to underestimate the value of information, and also not sensitive to the information value difference when probability in the auction changes. In addition, significant tendency of subjects' risk aversion in the auction has been observed, and women are significantly more risk aversion than men, bidding more conservative in the auctions.The conducting process of behavioral experiments is very time-consuming and resource-consuming. In order to increase the size of subjects and reduce the costs and operational complexity of conducting auction experiments, on the basis comparison and classification of the existing decision-making experiment platforms, we design and implement a web-based, research-oriented and scalable auction experimental system. This experimental system uses asynchronous, event-driven programming model and modular, component-based thinking to gurrentee scalability and real-time interaction. The successful completion of sequential auction described above indicates that the system basically meet the predetermined requirements.On the one hand, we make a theoretical supplement to the information valuation theory in sequentialauction, and reveals both information acquisition detail and bidding behavior pattern of bidders in real auctions. On the other hand, we provide a complete software solution for creating and conducting new risk decision-making experiments and auction behavior experiments in the future.