中文摘要：

随着交易数据和信用数据的积累和完善，网络借贷的利率定价机制逐步演化到平台直接标价的形式。国内一家代表性网络借贷平台经历了这一变化，本文用倾向性分数匹配法对这一变化进行估计。研究发现，平台识别风险的效率提高，改善了贷款质量，表现为违约率的降低。标价机制上调了原本狭窄的期限利率区分度，改善了平台呈现的信息质量，从而影响了投资人的投标行为，获得了更快的投标速度、更高的投标比例以及更弱的羊群行为。相对于平台原先“借款人选择利率-平台审核借款人”的利率形成机制，平台标价具有信息优势，能够根据大数据对借款人的风险进行准确定价，利率区分度的提高改善了平台的信息质量，并影响了投资人的投标行为。加入利率区分度的模型很好地解释了上述变化。这说明，标价机制实现了更高的定价效率，达到了改善平台经营的目的。

英文摘要：

As transaction data and credit data accumulate, interest price mechanism in peer-to-peer lending has moved to posted price mechanism. A leading peer-to-peer lending company in China has accomplished the promotion of loan quality and decreased default rate after changing to posted price mechanism. This step also improved investors bidding behaviors and bidding percentage, and decreased herding behavior. Using Propensity Score Match method, the author finds that compared to former price mechanism which can be concluded as ＂bor- rower chooses price-platform chooses borrower＂, platform under posted price has the information advantage to recognize and price loan interest. Posted price mechanism also increases interest distinction degree between good loan and bad loan to decrease information asymmetry, Interest rate discrimination perfectly explains all the changes in trading behavior. This attempt to post price mechanism improves platform＇ s operation performance.