The Geography of Bidder Behavior in Peer-to-Peer Credit markets Garrett T Senne The Ohio State University, Department of Economics Original Draft: November 10, 2014 Current Draft: January 11, 2016 Abstract. Theoretical and empirical research on the traditional credit market finds strong evidence that investors and lenders are sensitive to their distance from the borrower, especially early on in a venture. This is due to the cost of information gathering and monitoring. However, new online platforms could overcome the geographical constraints on investing. Recent empirical work, across many types of crowdfunding, has found mixed results. In this paper, using transaction data from a peer-to-peer lending site, I find that local lenders tend to bid earlier, both chronologically and relatively to other bidders in the auction, and bid larger amounts than nonlocal lenders Additionally, local lenders are more informed in the sense that they are better able to evaluate the underlying risk of borrowers. This is demonstrated by the fact that they bid significantly higher interest rates on loans that ex-post default and lower rates on loans that ex-post pay back in full. Lastly, I develop a simple model of social learning with heterogeneous agents that provides testable predictions. My results are consistent with this model; a listing with more early local bidding activity will attract more lenders, leading to a higher probability of funding and a lower final interest rate, if funded. This work suggests that the behavioral differences between local and nonlocal lenders are driven mostly by informational frictions and not merely preferences. Local lenders are better informed because they have easier and cheaper access to information, and this asymmetry contributes to explaining why geographic-based frictions are still present and relevant in online lending markets Keywords: Geography, Peer-to-Peer Lending, Informational Frictions JEL Classification: DS2, D83, L10, L86 1. Introduction It is an established fact that most of the inefficiencies in the credit market are due to the existence of asymmetric information between lenders and borrowers( Stiglitz and Special thanks to Jason Blevins, Maryam Saeedi, and Stephen Cosslett for their assistance and guidance, Xiang Hui and Robert Munk for their comments and encouragements, and Efraim Berkovich for providing the data Electroniccopyavailableathttp://ssrn.com/abstract=2721756
Electronic copy available at: http://ssrn.com/abstract=2721756 The Geography of Bidder Behavior in Peer-to-Peer Credit Markets∗ Garrett T. Senney The Ohio State University, Department of Economics Original Draft: November 10, 2014 Current Draft: January 11, 2016 Abstract. Theoretical and empirical research on the traditional credit market finds strong evidence that investors and lenders are sensitive to their distance from the borrower, especially early on in a venture. This is due to the cost of information gathering and monitoring. However, new online platforms could overcome the geographical constraints on investing. Recent empirical work, across many types of crowdfunding, has found mixed results. In this paper, using transaction data from a peer-to-peer lending site, I find that local lenders tend to bid earlier, both chronologically and relatively to other bidders in the auction, and bid larger amounts than nonlocal lenders. Additionally, local lenders are more informed in the sense that they are better able to evaluate the underlying risk of borrowers. This is demonstrated by the fact that they bid significantly higher interest rates on loans that ex-post default and lower rates on loans that ex-post pay back in full. Lastly, I develop a simple model of social learning with heterogeneous agents that provides testable predictions. My results are consistent with this model; a listing with more early local bidding activity will attract more lenders, leading to a higher probability of funding and a lower final interest rate, if funded. This work suggests that the behavioral differences between local and nonlocal lenders are driven mostly by informational frictions and not merely preferences. Local lenders are better informed because they have easier and cheaper access to information, and this asymmetry contributes to explaining why geographic-based frictions are still present and relevant in online lending markets. Keywords: Geography, Peer-to-Peer Lending, Informational Frictions. JEL Classification: D82, D83, L10, L86. 1. Introduction It is an established fact that most of the inefficiencies in the credit market are due to the existence of asymmetric information between lenders and borrowers (Stiglitz and ∗Special thanks to Jason Blevins, Maryam Saeedi, and Stephen Cosslett for their assistance and guidance, Xiang Hui and Robert Munk for their comments and encouragements, and Efraim Berkovich for providing the data. 1
Weiss, 1981; Dell'Ariccia and Marquez, 2004). However, the modern financial ecosystem is evolving rapidly, as innovations and technological advancements are radically altering the delivery of financial services worldwide. Increasingly, through use of the internet, people have been able to interact with each other more intensively(sharing more information) and extensively(cheaper to develop larger networks). Empirical work on the economic quences of the internet consistently find two facts: the Internet can overcome g graphic isolation(Balasubramanian, 1998; Forman, Ghose, and Goldfarb, 2009; Choi and Bell, 2011), and search costs are lower online than compared to offline(Bakos, 1997; Baye, Gatii, Kattuman, and John, 2009). Utilizing these new and cheaper online connections and crowdfunding websites, individuals are starting to find novel ways to address and overcome the asymmetry in the credit market. In this paper, I explore the informational frictions, arising due to geography, that have traditionally plagued the offline credit market and evaluate the effectiveness of online peer-to-peer lending to address it. Peer-to-peer (henceforth P2P) lending is a mechanism for groups of investors to lend money directly to individual borrowers without using a bank as an intermediary. This arrangement creates the possibility that borrowers can obtain loans at lower interest rates than they would get on a credit card or a normal loan without collateral. Individual lenders get the opportunity to invest in short-duration assets with higher rates of return than would be available on certificates of deposit, bonds, or money market accounts, all due to the cost savings arising from removing the intermediary. P2P lending has been heralded as an online tool that has the potential to level the playing field in the credit market by providing access to financing to more people in a more approachable way. A potential blend of altruism and the allure of higher yield has attracted investors to lending money online, especially in the aftermath of the 2008 financial crisis, where interest rates and banks' willingness to lend have hit historic lows. Additionally, a survey performed by Zopa. com(a British P2P site) reports that, although most P2P loans are unsecured, borrowers feel a greater responsibility to repay a loan that has been created by individual people instead of a loan from a bank(Cortese, 2014). Compared to traditional banks, online lenders enjoy the advantage that online marketplaces currently face fewer regulatory constraints, albeit the Security Exchange Commission and the Treasury Department are deeply interested their activities. Moreover P2P sites are structured to facilitate faster transfers of capital to the borrowers, and are operated by internet-based companies, which have lower overhead costs than companies with physical locations. Therefore, these companies operate more profitably in a market traditionally viewed as fairly risky with low margins(Cowley, 2015). P2P lending, along with other forms of crowdfunding, has established itself as an emerging alternative source of financing for startups and people with particularly limited access to traditional means Electroniccopyavailableathttp://ssrn.com/abstract=2721756
Electronic copy available at: http://ssrn.com/abstract=2721756 Weiss, 1981; Dell’Ariccia and Marquez, 2004). However, the modern financial ecosystem is evolving rapidly, as innovations and technological advancements are radically altering the delivery of financial services worldwide. Increasingly, through use of the internet, people have been able to interact with each other more intensively (sharing more information) and extensively (cheaper to develop larger networks). Empirical work on the economic consequences of the internet consistently find two facts: the Internet can overcome geographic isolation (Balasubramanian, 1998; Forman, Ghose, and Goldfarb, 2009; Choi and Bell, 2011), and search costs are lower online than compared to offline (Bakos, 1997; Baye, Gatii, Kattuman, and John, 2009). Utilizing these new and cheaper online connections and crowdfunding websites, individuals are starting to find novel ways to address and overcome the asymmetry in the credit market. In this paper, I explore the informational frictions, arising due to geography, that have traditionally plagued the offline credit market and evaluate the effectiveness of online peer-to-peer lending to address it. Peer-to-peer (henceforth P2P) lending is a mechanism for groups of investors to lend money directly to individual borrowers without using a bank as an intermediary. This arrangement creates the possibility that borrowers can obtain loans at lower interest rates than they would get on a credit card or a normal loan without collateral. Individual lenders get the opportunity to invest in short-duration assets with higher rates of return than would be available on certificates of deposit, bonds, or money market accounts, all due to the cost savings arising from removing the intermediary. P2P lending has been heralded as an online tool that has the potential to level the playing field in the credit market by providing access to financing to more people in a more approachable way. A potential blend of altruism and the allure of higher yields has attracted investors to lending money online, especially in the aftermath of the 2008 financial crisis, where interest rates and banks’ willingness to lend have hit historic lows. Additionally, a survey performed by Zopa.com (a British P2P site) reports that, although most P2P loans are unsecured, borrowers feel a greater responsibility to repay a loan that has been created by individual people instead of a loan from a bank (Cortese, 2014). Compared to traditional banks, online lenders enjoy the advantage that online marketplaces currently face fewer regulatory constraints, albeit the Security Exchange Commission and the Treasury Department are deeply interested their activities. Moreover, P2P sites are structured to facilitate faster transfers of capital to the borrowers, and are operated by internet-based companies, which have lower overhead costs than companies with physical locations. Therefore, these companies operate more profitably in a market traditionally viewed as fairly risky with low margins (Cowley, 2015). P2P lending, along with other forms of crowdfunding, has established itself as an emerging alternative source of financing for startups and people with particularly limited access to traditional means 2
Using data from an American P2P lending site, I examine whether the internet has improved the efficiency of the credit market by loosening the geographic constraints on investing. Evidence for this potential reduction in the informational difference between local and nonlocal lenders should be observable in their bidding behavior. If the two types of lenders bid the same way when faced with similar investing options, it is most likely that the internet has removed the geography-based frictions. However, if local and nonlocal lenders behave differently, this might be explained-as pointed out in the empirical literature- by one of two possible channels: asymmetric information or preference. My analysis shows that lenders indeed behave differently based on geography and, while a local preference exists informational differences seem to be a major driver for this behavior Most of the scholarly work on P2P lending has focused on the determinants of a listing being funded and the determinants of the final interest rates. The consensus is that soft factors like demographic and network effects matter; however, they are second order in importance, after hard factors, like the verified financial information of income, debt, and credit score. Because P2P lending is still in its infancy, its full potential as an alternative or supplement to the traditional banking industry is still an open empirical question A long literature documents the importance of distance in social and economic behavior; famously the first law of geography states"all things are related but near things are more elated than far things, "(Tobler, 1970). Additionally, there is an extensive theoretical work on early investing and capital ventures that predicts that investors and lenders are sensitive to their distance from the borrower- especially true in the early stages of the venture when there is little to no observable history. This result is because the cost of gathering and processing information, as well as monitoring, are generally thought to increase as the distance between lender and borrower grows. Empirical evidence supports these finding Explicitly, recent research on angel investing and large-scale acquisition reports that most investors are located within a half day of travel of the entrepreneur that they funding The predicted informational asymmetry between potential local and nonlocal investors may derive from the informational opacity of startups and small firms. Potential investors IBelleflamme, Lambert, and Schwienbacher(2010); Schwienbacher and Larralde(2010): Mollick(2014) Herzenstein, Andrews, Dholakia, and Lyandres(2008); Berger and Gleisner(2009); Pope and Sydnor (2011); Ravina(2012); Duarte, Siegel, and Young(2012); Lin, Prabhala, and Viswanathan(2013) Klafft(2008b, a); Iyer, Khwaja, Luttmer, and Shue(2009); Weiss, Pelger, and Horsch(2010) Arthur(1986, 1988, 1990); David and Rosenbloom(1990); Krugman(1991a, b) t STribus(1970); Florida and Kennedy(1988); French and Poterba(1991); Florida and Smith(1988); Martin, Inley, and Turner(2002); Mason and Harrison(2002): Powell, Koput, Bowie, and Smith-Doerr(2002): Zook 6Sohl(1999); Sorenson and Stuart(2001); Wong, Bhatia, and Freeman(2009)
of financing.1 Using data from an American P2P lending site, I examine whether the internet has improved the efficiency of the credit market by loosening the geographic constraints on investing. Evidence for this potential reduction in the informational difference between local and nonlocal lenders should be observable in their bidding behavior. If the two types of lenders bid the same way when faced with similar investing options, it is most likely that the internet has removed the geography-based frictions. However, if local and nonlocal lenders behave differently, this might be explained– as pointed out in the empirical literature– by one of two possible channels: asymmetric information or preference. My analysis shows that lenders indeed behave differently based on geography and, while a local preference exists, informational differences seem to be a major driver for this behavior. Most of the scholarly work on P2P lending has focused on the determinants of a listing being funded and the determinants of the final interest rates. The consensus is that soft factors like demographic and network effects matter; however, they are second order in importance,2 after hard factors, like the verified financial information of income, debt, and credit score.3 Because P2P lending is still in its infancy, its full potential as an alternative or supplement to the traditional banking industry is still an open empirical question. A long literature documents the importance of distance in social and economic behavior; famously the first law of geography states "all things are related, but near things are more related than far things," (Tobler, 1970). Additionally, there is an extensive theoretical work on early investing and capital ventures that predicts that investors and lenders are sensitive to their distance from the borrower– especially true in the early stages of the venture when there is little to no observable history.4 This result is because the cost of gathering and processing information, as well as monitoring, are generally thought to increase as the distance between lender and borrower grows. Empirical evidence supports these findings.5 Explicitly, recent research on angel investing and large-scale acquisition reports that most investors are located within a half day of travel of the entrepreneur that they funding.6 The predicted informational asymmetry between potential local and nonlocal investors may derive from the informational opacity of startups and small firms. Potential investors 1Belleflamme, Lambert, and Schwienbacher (2010); Schwienbacher and Larralde (2010); Mollick (2014). 2Herzenstein, Andrews, Dholakia, and Lyandres (2008); Berger and Gleisner (2009); Pope and Sydnor (2011); Ravina (2012); Duarte, Siegel, and Young (2012); Lin, Prabhala, and Viswanathan (2013). 3Klafft (2008b,a); Iyer, Khwaja, Luttmer, and Shue (2009); Weiss, Pelger, and Horsch (2010). 4Arthur (1986, 1988, 1990); David and Rosenbloom (1990); Krugman (1991a,b). 5Tribus (1970); Florida and Kennedy (1988); French and Poterba (1991); Florida and Smith (1988); Martin, Sunley, and Turner (2002); Mason and Harrison (2002); Powell, Koput, Bowie, and Smith-Doerr (2002); Zook (2002); Mason (2007). 6Sohl (1999); Sorenson and Stuart (2001); Wong, Bhatia, and Freeman (2009). 3
llect a sizeable amount of infe ng rate. The geographic proximity of local lenders facilitates cheaper and easier access to this information. Anderson and van Wincoop(2004)find that informational frictions associated with geography, including search costs, communication barriers, and contracting cos contribute to reducing transactional efficiency when parties are physically separated from each other In addition to more standard channels of financing small to medium-sized firms alse rely on relationship-based lending to obtain funds. Relationship lending is when a financial nstitution uses a sustained relationship across multiple interactions with the potential borrower in addition to normal financial information they regularly gather in order to process a loan request. Many studies of relationship lending find that the physical distance between the banks and the borrowers has increased significantly, implying that banks from farther away are able to develop the needed relationship with the borrower that historically only local banks would have. However, other studies find no discernable change in the distance between lenders and borrowers. Elyasiani and Goldberg(2004)suggest that the observed mixed results are most likely caused by the fact that the technological developments that mitigate geographic-based frictions have only been adopted by a small share of banks engaging in relationship lending. As the adoption and use of these new tools become more widespread, it is thought that location should become a less important factor in obtaining financing The informal capital markets, alternative sources of financing outside of the traditional mainstream credit market, have not to date been the focus of much scholarly analysis However, it is generally accepted, as noted by Harrison, Mason, and Robson(2010), that they comprise a series of potentially overlapping local markets rather than one fully integrated national market. With the internet and e-commerce being omnipresent, it is important to examine whether this shift in the commercial paradigm coupled with new online tools has been strong enough to alter or even invalidate the theoretical predictions of the effects that geography has on investing. Recent empirical work has shown that the internet has the potential to allow individuals and firms to overcome many traditional barriers that have fettered offline markets by mitigatil research on online transactions finds that online platforms might reduce some, but not all of the distance-related frictions, 10 'Cyrnak and Hannan(2000); Wolken and Rohde(2000); Petersen and Rajan(2002) Degryse and Ongena(2003): Brevoort and Hannan(2003) Ratchford, Pan, and Shankar(2003); Brynjolfsson, Hu, and Rahman(2009 ); Goldfarb and Tucker(2011) Lendle, Olarreaga, Schropp, and vezina(2013). Blum and Goldfarb(2006); Hortacsu, Martinez-Jerez, and Douglas(2009); Agrawal, Catalini, and Goldfarb (2011); Lin and Viswanathan(2014)
collect a sizeable amount of information before deciding to invest and determining the rate. The geographic proximity of local lenders facilitates cheaper and easier access to this information. Anderson and van Wincoop (2004) find that informational frictions associated with geography, including search costs, communication barriers, and contracting costs, contribute to reducing transactional efficiency when parties are physically separated from each other. In addition to more standard channels of financing, small to medium-sized firms also rely on relationship-based lending to obtain funds. Relationship lending is when a financial institution uses a sustained relationship across multiple interactions with the potential borrower in addition to normal financial information they regularly gather in order to process a loan request. Many studies of relationship lending find that the physical distance between the banks and the borrowers has increased significantly, implying that banks from farther away are able to develop the needed relationship with the borrower that historically only local banks would have.7 However, other studies find no discernable change in the distance between lenders and borrowers.8 Elyasiani and Goldberg (2004) suggest that the observed mixed results are most likely caused by the fact that the technological developments that mitigate geographic-based frictions have only been adopted by a small share of banks engaging in relationship lending. As the adoption and use of these new tools become more widespread, it is thought that location should become a less important factor in obtaining financing. The informal capital markets, alternative sources of financing outside of the traditional mainstream credit market, have not to date been the focus of much scholarly analysis. However, it is generally accepted, as noted by Harrison, Mason, and Robson (2010), that they comprise a series of potentially overlapping local markets rather than one fully integrated national market. With the internet and e-commerce being omnipresent, it is important to examine whether this shift in the commercial paradigm coupled with new online tools has been strong enough to alter or even invalidate the theoretical predictions of the effects that geography has on investing. Recent empirical work has shown that the internet has the potential to allow individuals and firms to overcome many traditional barriers that have fettered offline markets by mitigating some geographic frictions.9 Other research on online transactions finds that online platforms might reduce some, but not all, of the distance-related frictions.10 7Cyrnak and Hannan (2000); Wolken and Rohde (2000); Petersen and Rajan (2002). 8Degryse and Ongena (2003); Brevoort and Hannan (2003). 9Ratchford, Pan, and Shankar (2003); Brynjolfsson, Hu, and Rahman (2009); Goldfarb and Tucker (2011); Lendle, Olarreaga, Schropp, and Vézina (2013). 10Blum and Goldfarb (2006); Hortaçsu, Martínez-Jerez, and Douglas (2009); Agrawal, Catalini, and Goldfarb (2011); Lin and Viswanathan (2014). 4
Although technology makes it easier and cheaper for individuals to gather and process information about market conditions it is unclear if informational frictions like search costs persist in the online marketplace. I examine the ongoing empirical question of whether geographic frictions exist on the online credit market, and if so what factors contribute to their existence. My analysis centers on how geography affects lender behavior in P2P lending markets by focusing on the issue of whether the new P2P lending sites can loosen the geographic constraints on investing. More explicitly, I ask relative to nonlocal lenders (1)do local lenders bid different amounts, (2)do local lenders evaluate and price the risl of the listings differently, and(3)do local lenders tend to bid at different times during he auction. Additionally, if distance-related frictions do exist, (4)does the difference in behavior arise from informational asymmetry or simply a preference on the part of local lenders, and lastly, (5)how does the presence of local lenders in the market affect other lenders'decisions to enter and their behavior after entering Using bid and listing level transaction data from lending auctions on Prosper. com, I stimate that local lenders tend to bid earlier and larger amounts than nonlocal lenders Furthermore, local lenders also seem better able to evaluate the riskiness of loan requests They tend to ex-ante bid larger interest rates when the loan ex-post defaults and less when the loan ex-post pays back in full. Reconciling theory and previous empirical work, I conclude from my results that there exists a local preference in the demand for loans, but local lenders' interest rate bids demonstrate support for the informational-based explanation of the observed behavioral differences between lenders 2. Overview of Peer-to-Peer Lending Crowdfunding, which P2P is a part of, is a process through which an individual or firm attempts to obtain financing by soliciting for usually small contributions from a large number of online investors. These new financing platforms are the result of a social movement that arose in reaction to the emergence of new technologies that are enabling new and cheaper ways of forming social networks(Adams and Ramos, 2010) The crowdfunding market has quickly grown from its creation in the early 2000s and is predicted to reach $34.4 billion globally in 2015 (Massolution. com, 2015). The movement has bifurcated into donation and financial(debt/equity) based sites both geared towards different market segments. The first American online platform to facilitate debt transactions launched in late 2005. The two current largest domestic players, Lending Club and Prosper, control about 98% of the American P2P credit market. Combined, they issued almost $5 billion in loans in 2014 and are predicted to issue about $8 billion in 2015(NSR Invest 2015). In this paper I focus on the P2P lending site Prosper. com, which has been called the
Although technology makes it easier and cheaper for individuals to gather and process information about market conditions, it is unclear if informational frictions like search costs persist in the online marketplace. I examine the ongoing empirical question of whether geographic frictions exist on the online credit market, and if so what factors contribute to their existence. My analysis centers on how geography affects lender behavior in P2P lending markets by focusing on the issue of whether the new P2P lending sites can loosen the geographic constraints on investing. More explicitly, I ask relative to nonlocal lenders: (1) do local lenders bid different amounts, (2) do local lenders evaluate and price the risk of the listings differently, and (3) do local lenders tend to bid at different times during the auction. Additionally, if distance-related frictions do exist, (4) does the difference in behavior arise from informational asymmetry or simply a preference on the part of local lenders, and lastly, (5) how does the presence of local lenders in the market affect other lenders’ decisions to enter and their behavior after entering. Using bid and listing level transaction data from lending auctions on Prosper.com, I estimate that local lenders tend to bid earlier and larger amounts than nonlocal lenders. Furthermore, local lenders also seem better able to evaluate the riskiness of loan requests. They tend to ex-ante bid larger interest rates when the loan ex-post defaults and less when the loan ex-post pays back in full. Reconciling theory and previous empirical work, I conclude from my results that there exists a local preference in the demand for loans, but local lenders’ interest rate bids demonstrate support for the informational-based explanation of the observed behavioral differences between lenders. 2. Overview of Peer-to-Peer Lending Crowdfunding, which P2P is a part of, is a process through which an individual or firm attempts to obtain financing by soliciting for usually small contributions from a large number of online investors. These new financing platforms are the result of a social movement that arose in reaction to the emergence of new technologies that are enabling new and cheaper ways of forming social networks (Adams and Ramos, 2010). The crowdfunding market has quickly grown from its creation in the early 2000’s and is predicted to reach $34.4 billion globally in 2015 (Massolution.com, 2015). The movement has bifurcated into donation and financial (debt/equity) based sites both geared towards different market segments. The first American online platform to facilitate debt transactions launched in late 2005. The two current largest domestic players, Lending Club and Prosper, control about 98% of the American P2P credit market. Combined, they issued almost $5 billion in loans in 2014 and are predicted to issue about $8 billion in 2015 (NSR Invest, 2015). In this paper I focus on the P2P lending site Prosper.com, which has been called the 5