考虑信息不对称的电子市场动态契约设计

近年来，伴随着互联网技术的快速发展以及电子商务在广大消费者中的日益普及，以淘宝网、eBay等为代表的电子市场的销售规模和交易金额都实现了跳跃式的增长。但是，在买卖双方交易数量快速增加的同时，提供交易平台的电子市场所有者始终缺乏有效的收费模式来实现自身的利润最大化，这一窘境已经得到产业界和学术界越来越多的关注。当前，各类电子市场大多采用的收费模式是以每个卖家前一期的销售历史来预测该卖家下一期的销售数量，并以此作为基准来动态更新电子市场对于该卖家下一期的收费标准。这种收费方法暗含的假设是相邻周期间产品销量为强正相关，但忽略了产品销量因季节性、流行性、替代产品威胁等各类因素影响而导致的更加复杂的中长期变化趋势。针对这一问题，本文为电子市场所有者提供了一类新的个性化动态契约机制。这种契约机制假设相邻周期的产品销量服从联合正态分布，在每个周期开始前，卖家能够充分考虑到各类因素的影响，从而对于下一期销量做出比电子市场所有者更准确的条件预期。但是，卖家对于相邻周期销量间的相关系数的观测属于私有信息，为了诱导卖家披露这一私有信息，电子市场所有者在每周期开始前需要根据该卖家上一周期的实际销售数量向卖家提供一系列契约（a menu of contracts）供卖家选择。本文构建了这种个性化动态契约的参数优化问题。求解结果揭示了一个简单的决策准则，即最优的契约参数可以表示为关于上述相关系数的failure rate的函数。通过一组数值试验，本文将这种契约与其他三类契约进行了对比，验证了这种契约能够充分地利用卖家的私有信息来提升电子市场所有者的利润。本文所述的契约机制在互联网环境下具有较强的可操作性，因此可以为电子市场所有者提供直接的管理参考和决策依据。

Dynamic Contract Design in the Electronic Markets under Information Asymmetry

The recent decade has witnessed the development of internet technology and widespread use of electronic commerce. While the emerging electronic markets, such as Taobao and eBay, have achieved huge number of market transactions at a daily basis, efficient contract designs are still missing for the market owners to extract greater monetary benefits from the third-party sellers' sales in the market. Currently, the most popular form of contract is the commission contract based on the sales history of each seller, which implicitly assumes that sales history is positively correlated with future sales but overlooks the influences of other factors such as seasonal effects, fashion pattern, and potential substitutes, etc. In this paper, an innovative contract design for the owners of electronic market is provided by taking advantage of third-party sellers' private expectation on future sales. In the beginning of each period, upon observing the sales history, the owner offers a customized menu of contracts to the third-party seller. The seller reveals her private information by selection the contract designed for her type. The optimal contract design problem is solved and a simple decision rule is discovered for the contract design. The optimal contract parameter can be given by a function of the failure rate of the estimating distribution of the correlation coefficient. The superiority of this contract is verified by a set of computer simulation study. The numerical results verify that the market owner's profit can be improved by benefiting from the seller's private information. Given the convenience of implementation in the online channels, this type of contract is expected to help owners of electronic market quickly refine their contracting with third-party sellers.