When big-tech lenders enter a market, they already have access to a wealth of data about potential borrowers, leaving traditional lenders at a disadvantage. Jiayin Hu, Yanting Chen and Yingwei Dong study how a car equity loan provider in China adapted when a big-tech rival entered the market. It reacted not by hiking interest rates but by raising its lending standards, a strategy that paid off.
Financial technology (fintech) has greatly changed the competitive landscape in the financial system. Notably, credit extended by large technology companies, or big-tech credit, is projected to surpass one trillion dollars worldwide in 2023. Empirical evidence has shown that fintech lending promotes financial inclusion by extending credit to unbanked and underprivileged borrowers. However, the impact of big-tech credit on traditional lenders hasn’t been fully understood, especially on small- and medium-sized banks and non-bank financial intermediaries. Will these traditional lenders ultimately disappear in the shadow of big-tech lending? Or will they coexist with big techs by differentiating their business model?
Non-bank traditional lenders more exposed
Non-bank lenders do not take deposits and thus face higher funding costs and less strict regulations than banks. As a result, they often serve borrowers with higher credit risks and find their niche in serving clients that banks may have rejected. Big-tech and fintech lenders typically begin by lending to unbanked borrowers and have a larger clientele overlap with non-bank lenders, making this clientele more exposed to big-tech competition.
Non-bank lenders face more direct competition from big tech in small loans. China’s central bank defines small and micro-enterprise loans from traditional banks as “single-account credit of less than 10 million yuan”. In contrast, non-bank traditional lenders’ loan amounts for entrepreneurs rarely exceed 300,000 yuan. Comparatively, the average loan amount granted to small businesses by digital banks MYBank and WeBank, owned, respectively, by Ant Group and Tencent (the two largest big tech platforms in China) is approximately 270,000 yuan — closer in size to the loans extended by non-bank lenders.
Impact of big tech credit on non-bank traditional lenders
To quantify the impact of big-tech credit on non-bank traditional lenders, we use loan-level data from a car equity loan company with multiple branches in China. The non-bank lender, which operates through local brick-and-mortar branches, launched its first car equity loan product in May 2015 and gradually expanded its branch network to a nationwide presence. While allowing borrowers to retain the control rights of the collateral car, the company implants a GPS device in the vehicle so that it can locate it and pull it back if the borrower defaults on the payment or if the car surpasses a specific boundary. This way, the lender ensures the collateral remains within its reach, and the borrower retains the use of the vehicle, an essential commuting tool in an entrepreneur’s daily life and productive activities.
Our paper exploits geographical differences in the penetration ratios of big-tech lending. Our big-tech credit penetration index comes from the Peking University Digital Financial Inclusion Index of China (PKU-DFIIC), a detailed data set covering penetration ratios of Ant Group’s digital finance activities. We use two instrumental variables: the great-circle distance to Hangzhou city, Ant Group’s headquarters, and the penetration ratios of their payment services, which serve as a basis for their lending businesses but do not directly compete with traditional lenders.
We find that the car equity company experiences a decline in the lending business as Ant Group expands in the area. We evaluate each branch of the company’s dynamic performance since its opening month and compare the differences between branches in cities with high penetration ratios of Ant Group credit and those in cities with low penetration ratios. Our estimates capture the differences between branches at the same development stage but with different intensities of big-tech competition.
These results show that larger big-tech credit penetration reduces the number of loans originated by the traditional lender, consistent with our hypothesis that borrowers in cities with higher penetration ratios of big-tech credit are more likely to borrow from big techs, which reduces the attractiveness of traditional loans.
While we do not find evidence that big-tech competition induces local branches to lower the collateral requirement (measured by the average price of collateral per loan), we do find a reduction in total collateral values, corroborating our argument that branches facing more intense big-tech competition originate fewer loans. Big-tech credit relaxes households’ borrowing constraints and weakens traditional lenders’ competitiveness in the lending market.
The non-bank traditional lender responds to big-tech competition by holding higher lending standards. Branches in cities highly penetrated by big-tech credit approve fewer loans per unit collateral value (defined as the loan-to-value ratios), implying a more conservative attitude towards who qualifies as a borrower. Interestingly, the interest rates charged by non-bank branches do not change and are restricted to a limited range. It seems that non-bank traditional lenders adopt a more prudent lending standard, i.e., reducing the loan-to-value (LTV) ratio but not increasing the interest rates, to contain default rates.
We argue that this cautiousness in lending reflects traditional lenders’ concern about “cream-skimming” in the loan market by big techs, which may use more advanced fintech to screen borrowers and “cherry pick” from the shared application pools. Fintech reduces the value of soft information in lending and weakens the advantage of non-bank traditional lenders, which may lose their relatively high-quality clientele to big-tech lenders.
This response of informationally disadvantaged traditional lenders to big-tech competition echoes the classical literature on asymmetric information. For instance, the quantity-based response is consistent with the credit rationing motive proposed by the seminal work of Stiglitz and Weiss (1981), where lenders find it optimal not to raise interest rates due to adverse selection and moral hazard concerns under asymmetric information. We also find that the increase in the lending standards pays off: branches facing fiercer big-tech competition do not experience higher default rates, indicating the success of risk control measures through lower LTV ratios.
While big-tech lending generally improves social welfare by reducing informational asymmetry, relaxing the collateral constraint and promoting inclusive finance, its impact on the traditional lending business (especially small- and medium- sized banks and non-bank financial institutions) is worth further investigation.
Our paper highlights the competitive impact of big-tech credit and its implications on the market structure of the financial industry, such as the increasing concentration. In a different paper, we explore another possibility of fintech transformation by non-bank traditional lenders. Overall, it is a promising and fruitful research avenue towards a more comprehensive understanding of the opportunities and challenges presented in the era of big tech.
- This blog is based on In the Shadow of Big Tech Lending, China Economic Review.
- The post represents the views of its author(s), not the position of LSE Business Review or the London School of Economics.
- Featured image by Scott Graham on Unsplash
- When you leave a comment, you’re agreeing to our Comment Policy.