Policing P2P Microcredit Platforms in Developing Countries

In the last decades, online P2P microcredit businesses have boomed around the world. While many have benefited, it’s important for the future of the business to ensure both lenders and borrowers are protected from frauds and scams. Credit ranking systems are commonly used but how effective are they and how can they be improved?

By Cathy Tian, PhD Candidate, Department of Management, CUHK Business School

In March 2005, Richard Duvall, James Alexander, Giles Andrews, and David Nicholson established Zopa in the UK as the first person to person (P2P) microcredit online platform in the world and realized microcredit among individuals. Since then, P2P as a new business model began to expand to other countries and became popular worldwide. In the United States, P2P platforms such as Kiva and Prosper emerged, while in Europe, Smava, Boober and myc4 were established to solve financing problems faced by individuals and small enterprises. To date, P2P microcredit platforms have become very successful in developed countries such as US, UK and Australia. By 2015, the loan volume through P2P platforms was expected to hit 77 billion dollars, and the world’s largest P2P lending marketplace LendingClub was trading at a market value of about $7 billion, according to Bloomberg.

This new business model has also expanded to developing countries such as India and China. Two well-known examples are DhanaX in India and PPDai in China. However, unlike those in the developed countries, the credit systems in developing countries still have many flaws which have restricted the development of P2P platforms.

One of the most severe problems is the lack of a credit scoring system to divide the credit ranks of different borrowers. Our working paper[1] studies how China’s PPDai solves credit problems and analyses the company’s weaknesses in solving those problems. The results can help to promote the development of the P2P business model in China.

PPDai provides a platform to connect people who have extra money with those who need more money. Through PPDai, lenders and borrowers can build multiple connections – a lender can lend money to several borrowers, while a borrower can borrow money from several lenders. Through these multiple connections, lenders can decrease their risks by diversifying investments while borrowers can increase the amount of capital by enlarging the number of investors.

The realization of the lending process is through auction. A borrower first sets an initial interest rate and then lenders begin to make bids on it. But the result of this kind of auction is contrary to that of a traditional auction. In the traditional auction, the bidder who gives the highest price wins, but for PPDai, it is the bidder who gives the lowest interest rates wins.

After a successful auction, the company then checks the result of the auction and generates an electronic lending agreement. Finally, the lender’s money is transferred to the borrower’s account through PPDai and the lending transaction is complete. Then each month, the borrower repays the principle and interest through the PPDai platform until all the money is paid off.

Compared to traditional lending, PPDai has its advantages. First, the transactions are realized through the Internet which makes transaction processes more convenient and cheaper. Second, the company encourages lenders and borrowers to establish friendship during the process which helps to increase the possibility of successful ongoing transactions in the future.

However, PPDai also has disadvantages. The biggest disadvantage is related to information asymmetry, that is, an imbalance of power in transactions. During transactions, lenders need to know more information about borrowers to reduce investment risks. But the information provided by borrowers before transactions could be fake or incomplete. Borrowers could deliberately hide some information to optimize the possibility of raising funds. Such information asymmetry will lead to adverse selection problem, which means those who are most active in searching for capital and most likely to obtain capital are most likely to break the contracts. After transaction, a borrower may simply refuse to repay the money and jeopardize the lender’s interest. Information asymmetry can therefore lead to moral hazard. Compared to traditional lending, adverse selection and moral hazard problems are more severe for microcredit platforms like PPDai.

To reduce moral hazard, PPDai adopted several methods during the development process. The company makes an online blacklist of those who didn’t repay money on time. To solve the adverse selection issue, PPDai built a credit ranking system to evaluate the credit of each borrower. If a borrower has a higher credit ranking, he or she can borrow a larger amount of money compared to those who have a lower ranking. With the help of credit rankings, lenders can make better informed decisions on whether they will lend money to a certain borrower.

Since the factors considered in the process of assigning credit ranking are important for the prediction accuracy of borrowers’ adverse selection, our study collected samples from PPDai to further study if the factors included by the company are effective. The samples included 47 borrowers who refused to repay or didn’t repay on time and 105 borrowers who always repaid on time.

PPDai chooses five factors to evaluate a borrower’s ranking: basic information, community score (i.e., the frequency of user activities in PPDai forum), the number of times of full repayments, education certification, and mobile phone number. Since a user can be both a borrower and a lender simultaneously, the study further incorporates a lending credit score (i.e., the credit score is calculated according to the user’s lending history in PPDai) to see if it can help to predict whether a user will break his/her borrowing contract.

Our results show that among the five factors, education certification and the number of times of full repayments are important for predicting if a borrower is likely to violate the contract, whereas basic information, community score and mobile phone number are not significant indicators of contract breaking behavior.

The result also shows having a lending credit score has a significant influence on a borrower’s contract breaking behavior. One of the reasons might be that if a user has a high lending credit score, the violation cost for breaking contract will be high since such behavior has a bad influence on his or her reputation. This suggests that when building a credit-ranking system, P2P firms should consider also including lending credit scores as a borrowing credit evaluation factor.

Although P2P businesses are expanding quickly in China, the country’s credit ranking system is still underdeveloped and insufficient to protect the interest of lenders. A well-developed credit-ranking system in P2P platforms can provide a reference for the country’s credit system and help its economic development in the long run.

Reference:

[1] Tian, X., Yang, D. Online microcredit and credit ranking: A case study. Working paper.