رفتار گله ای در اعطای وام P2P آنلاین : تحقیقات تجربی
|کد مقاله||سال انتشار||مقاله انگلیسی||ترجمه فارسی||تعداد کلمات|
|14449||2012||9 صفحه PDF||سفارش دهید||8231 کلمه|
Publisher : Elsevier - Science Direct (الزویر - ساینس دایرکت)
Journal : Electronic Commerce Research and Applications, Volume 11, Issue 5, September–October 2012, Pages 495–503
We study lender behavior in the peer-to-peer (P2P) lending market, where individuals bid on unsecured microloans requested by other individual borrowers. Online P2P exchanges are growing, but lenders in this market are not professional investors. In addition, lenders have to take big risks because loans in P2P lending are granted without collateral. While the P2P lending market shares some characteristics of online markets with respect to herding behavior, it also has characteristics that may discourage it. This study empirically investigates herding behavior in the P2P lending market where seemingly conflicting conditions and features of herding are present. Using a large sample of daily data from one of the largest P2P lending platforms in Korea, we find strong evidence of herding and its diminishing marginal effect as bidding advances. We employ a multinomial logit market-share model in which relevant variables from prior studies on P2P lending are assessed.
Peer-to-peer (P2P) lending is a breed of financial transactions that occur directly between individuals without the intermediation of a traditional financial institution (en.wikipedia.org/wiki/P2p_lending). It has a short history, but has rapidly grown in recent years. The first online P2P lending company was Zopa (www.zopa.com), launched in 2005 in the United Kingdom. In the United States, Prosper (www.prosper.com) was the first P2P lending firm, and opened to the public in February 2006. It is now the largest P2P lending platform, with over a million members and over US$219 million in personal loans funded as of February 2011. P2P online exchanges are growing in the United States and United Kingdom as an alternative platform to traditional saving and investment (Slavin 2007). Harvard Business Review reports that every major bank will have its own P2P lending network within five years, and that P2P lending will be among the most important financial service innovations in the coming decade ( Sviokla 2009). This new phenomenon has garnered significant attention from researchers. Many of them focus on social networks in P2P lending (Freedman and Jin, 2008, Herrero-Lopez, 2009 and Lin et al., 2011). In the P2P lending market, transaction costs are reduced by eliminating expensive intermediaries, but information asymmetry problems become more severe than in traditional markets. This is because most individual lenders in online P2P lending lack financial expertise, and the lending experience takes place in a pseudonymous online environment (Klafft 2008). In this situation, social networks between individuals mitigate adverse selection and lead to better outcomes in all aspects of the lending process (Lin et al. 2011). Social networks on Prosper reveal some soft information about borrower risk, and therefore have the potential to compensate for the lack of hard information (Freedman and Jin 2008). Besides social networks, borrowers’ characteristics, including demographic characteristics, financial strength, and effort prior to making a request, are regarded as determinants of funding success in P2P lending (Herzenstein et al. 2008). Despite those new experimental mechanism designs and system features, the risk of information asymmetry lenders face may not be erased easily. It has been studied that players exhibit herding behaviors in online commerce when they face risk of uncertainty such as information asymmetry. Herding behavior describes many social and economic situations in which an individual’s decision-making is highly influenced by the decisions of others (Duan et al. 2009). Therefore, it has been theoretically linked to many economic areas such as investment recommendations (Scharfstein and Stein 1990), price behavior of initial public offerings (IPOs) (Welch 1992), fads and customs (Bikhchandani et al. 1992), and delegated portfolio management (Maug and Naik 1996). Duan et al. (2009) suggest that herding behavior could be especially prominent on the Internet for two reasons. The first is information overload. There is an excessive amount of information on the Web, so online users have difficulty understanding and using all the information (Brynjolfsson and Smith 2000). Doing what others do could be an efficient and rational way to make decisions in this circumstance. The second reason is that people can easily observe others’ choices on the Internet. Most online e-commerce websites provide a way to sort their products in the order of previous sales performance. When a customer clicks on a book in one of the largest online bookstores, Amazon, she will not only obtain information about that book, but also see other items that previous customers bought with the particular book. According to Herzenstein et al. (2008), there is a considerable difference between the number of lenders bidding on funded loan listings and the number of lenders bidding on unfunded loan listings. The average of the former is 62.6, while the average of the latter is only 1.6. What makes such significant difference? Is it the outcome of rational judgment of investors or inflated by herding behaviors? Investigating herding behavior in P2P lending market is the main objective of this study. Since P2P lending platforms are online, it is obvious that they satisfy the aforementioned conditions for herd behaviors that Duan et al. (2009) identify. When the lenders decide whether to invest their money in a loan request, they can verify the number of lenders who have already participated. If investors are influenced by the decisions of other investors (Devenow and Welch 1996), this number is a kind of signal for lenders. In other words, an auction that already has many bidders may be more attractive to lenders considering investment. We speculate that herding behavior is more common in this market due to the possibility of adverse selection and the limited institutional knowledge mentioned above when lenders face unknown borrowers over the Internet. We empirically examine lenders’ herding behaviors in the P2P lending market. Two things make this study interesting. First, we question whether herding behaviors exist in the P2P lending market because some characteristics of this market are distinctly different from those online markets where herding behaviors are observed. Herd behavior refers to people who do what others are doing instead of using their own information ( Banerjee 1992). In other words, players take herding strategy because they believe that others are better informed than they are. For example, herding behavior in the stock market is led by expert analysts. Many other cases of herding behavior show that buyers rely on information gathered by other buyers of experience goods. Prior consumers already have experienced various goods and services, and therefore, potential buyers will believe that they may have better information. Thus, they will flock to popular goods or music bands. However, online P2P lending does not seem to have such obvious sources of the better information. Most peers in P2P lending are not professional investors. Also, it will take a much longer time until “true” information is revealed by loan default or payments on time. As a result, the existence of herding in the P2P lending market is questionable, since no clearly superior information source can be identified. In other words, is herding behavior triggered by blind trust of the collective intelligence in the online market? Second, there has been no micro-data from a single P2P lending company that so nobody has been able yet to explore the details of this setting. We have an opportunity to remedy this, based on the unique data set that we have been able to construct. The rest of the paper is organized as follows. Section 2 presents the related theoretical and empirical literature on P2P lending and herding behavior. Section 3 shows our research hypotheses with the reasoning behind them, and Section 4 describes our data. In Section 5, we develop and analyze the empirical model and discuss the results. We conclude the paper by mentioning limitations and future research in Section 6.