**1. Introduction**

Peer-to-peer (P2P) lending has passed the shakeout period and entered a steady growth period. Its development experience can provide valuable inspiration for current market players. The fast development of disintermediated online lending in the past decade, while providing convenience and efficiency, also generates significant concealed credit risk for the financial system (Huang 2018). For example, due to the fragile auditing process and high default rate, in August 2018, the Chinese P2P market ushered in its consolidation period and experienced a reduction of 42% in P2P platforms when 168 platforms ended operation. Even after the Interim Administrative Measures for the Business Activities of P2P Lending Information Intermediaries was established, the default rate in the P2P industry was still at a high level (You 2018). According to (Gao et al. 2021), Chinese P2P lending platforms have an astonishing default rate of 87.2%, based on data available in 2019. Thus, questions are generated. Does disintermediation really provide more efficiency to the lending market, or does it actually add unforeseen credit risk to the system? Does peer screening work efficiently? This paper used a large dataset with 251,887 loan listings from the pioneer P2P lending platform RenrenDai to investigate the efficiency of the credit-screening mechanism under a disintermediated environment by comparing the performance of loan funding signals and repayment determinants.

A group of scholars (Dorfleitner et al. 2016; Santoso et al. 2020; Liao et al. 2015; Lin et al. 2013; Pötzsch and Böhme 2010; Khan and Xuan 2021) attempted to investigate the determinants of credit rationing in the field. However, the findings in the literature regarding the determinants of loan application success and repayment behavior were inconsistent. Moreover, due to data limitation, the analyses of the default determinants were insufficient. The purpose of our paper was, therefore, to contribute to the literature that explores the determinants of the loan application's performance and the default

**Citation:** Wang, Yao, and Zdenek Drabek. 2021. Adverse Selection in P2P Lending: Does Peer Screening Work Efficiently?—Empirical Evidence from a P2P Platform. *International Journal of Financial Studies* 9: 73. https://doi.org/ 10.3390/ijfs9040073

Academic Editors: Sabri Boubaker and Thanh Ngo

Received: 19 October 2021 Accepted: 6 December 2021 Published: 20 December 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

behavior of the online P2P lending platform. More importantly, the comparison of the results can provide evidence for our research question: Does the peer screening mechanism in the P2P platform efficiently diagnose the signals provided by the borrowers in their loan applications? Due to limitations in the repayment history data, no similar study has been performed using an emerging-market dataset. The only reference is (Iyer et al. 2016), who explored the question by using a Prosper dataset and US credit bureau data. However, this paper did not go deeper and explore the specific determinants which resulted in the misspecification. Our paper will fill this gap and also enrich the literature for emerging markets. We used the dataset from P2P pioneer RenrenDai to test our hypothesis. We divided the information provided by the borrowers into two categories: hard (financial) information and soft (social) information. Our findings showed that the hard (financial) indicators were given great importance when lenders were deciding whether to lend money. However, hard information was either unimportant or even acted in the opposite direction when it came to predicting the repayment behavior of a borrower. Meanwhile, soft information had much less inconsistency in the two models. This proved the existence of a TYPE II error in the investors' decision-making process, which indicated that the investors were predisposed to making inaccurate diagnoses of signals, and gravitated to borrowers with low creditworthiness while inadvertently screening out their counterparts with high creditworthiness. Due to the growing size of the fintech industry, this may pose a systematic risk to the financial system, necessitating regulators' close attention. Since, in contrast to hard financial-based signals, investors can better diagnose the soft signals, this implies enlarging socially related soft signals, and the building up of a comprehensive credit bureau could mitigate the adverse selection in a disintermediation environment.

The paper is divided into five sections. In the literature review, we provide an overview of the existing literature concerning the determinants of loan application success and loan defaults in the P2P market. We compare inconsistencies to find the gaps, then we define our scope. In Section 3, general information about the dataset will be introduced, and our model and the descriptive summary of the chosen variables will be presented. In Section 4, the results of the model are analyzed in detail. Finally, we conclude and discuss the policy implications in the discussion and conclusions section.
