Helping or Ruining? An Ethical Examination of Alipay’s Credit Business and Its Impact on Personal Credit
Strict sanitary measures resulting from the COVID-19 have forced businesses to operate online to the largest extent possible. In China, the switch from offline to online was greatly facilitated by platform-like tech giants who have penetrated almost every aspect of Chinese urban life. From the financial perspective, businesses of all sizes in China have been empowered to continue to make smooth transactions with their customers thanks to innovative FinTech companies. Billions of nonbank online payments were generated on a daily basis (the figure had already reached 720 billion worthy of US$383 trillion in 2019, well before the COVID-19, averaging around 2 billion and US$1 trillion per day). Due to the network effect and increasing returns inherent in the modern IT industry, it’s natural that the surging demand catalyzed by the pandemic would foster exponential growth of leading players. The harvest seemed to be on the horizon as the largest ever IPO in human history was well scheduled to be launched in the fall. Had nothing gone wrong, Ant Group, the world’s most valuable unicorn under the affiliation with Alibaba Group (the world’s largest retailer and e-commerce company, as of 2020), would have raised US$34.5 billion in its IPO at Shanghai and Hong Kong Exchange in October 2020.
Things did go wrong. Just two days before the big date, the Chinese authorities abruptly halted the IPO, stating that “the company may not meet listing qualifications or disclosure requirements” due to “some major issues such as changes in the financial technology regulatory environment” which were “reported by Ant Group” itself.
No details about the exact “major issues” were disclosed at the time of halt, but days later at the end of December, the People’s Bank of China (PBC, the central bank) publicized that Ant Group was experiencing “defects in corporate governance, negligence of the regulation and compliance requirements, illegal practices of arbitrage, abuse of its market monopoly position, and infringement of the rights of consumers”, and urged it to “increase transaction transparency and end its unfair competition; operate personal credit business based on legal compliance and better protect consumer data privacy; ensure sufficient funds for its associated transactions; rectify its financial businesses including loaning, insurance and wealth management; impose stronger institutional governance over its asset securitization business”.
The statement more or less verified what the media had previously guessed at the sudden halt. Several articles had emphasized the systematic financial risks inherent in Ant’s business model with caution drawn on Ant’s resemblance to the mortgage originators who fueled the 2007~2010 subprime mortgage crisis. It is Ant who originates loans to individuals and SMEs using its assessment algorithms, but the underlying cash mainly comes from partner banks with Ant funding merely 2%, making the banks instead of Ant the major bearer of the default risks. That is, Ant can make loans and charge commission fees at almost zero marginal cost and doesn’t need to worry much about how to get the principals and interests back. As such, Ant is disincentivized to filter out borrowers with weak repaying ability and high default risks. It’s hard not to recall the 2007~2010 catastrophe if the size of the borrowing population gets too large for the financial system to undertake.
The problem is: the borrowing population is indeed growing too large. Ant Group stated in its IPO prospectus that its one-stop mobile application Alipay had 711 million monthly active users (more than half of the Chinese population) and a standing US$320 billion in its micro-lending services as of June 2020. Third-party research also showed that as high as 62.4% of its respondents used nonbank consumption credit services (i.e. micro-lending to consumers) as their primary payment method. If access to micro-lendings is not granted under proper selection criteria, such a huge user base can incur severe credit risks. The lack of scrutiny over the criteria in the past is probably what motivated the authorities to step in with a series of new regulations aiming at online micro-lenders.
The impacts don’t just stop here. Functioning as an intermediary between financial institutions and end-users, Ant does not transfer risks only to the institutions and the macroeconomy, but also to each user who participates in the micro-borrowing programs. To understand why this is the case, we need to first investigate the relevant credit businesses of Ant Group.
Ant’s dominant market position stems from its close connections with the huge business empire of Alibaba which spans online retail (e-commerce), food delivery, health care, travel, transportation, housing, and many other fields. A Chinese can basically satisfy all his/her everyday needs through Alibaba and pay with Alipay. The broad business scope combined with innovative technologies helped Ant gain incredible penetration. Nowadays Alipay is frequently used as the primary method for paying and receiving money. However, payment services are limited in profitability as easing transactions doesn’t necessarily help much in motivating people to consume more. In economic terms, the budget constraint remains unchanged. Ant then thought it would be a great idea to increase the current purchasing power of its users by allowing them to delay the payments until some time later at zero interests or adopt an installment scheme when they purchase goods or services from Alibaba or its contracted merchants, in order to lure people into buying more. The service is called Huabei (meaning “just spend” in Chinese) and its rationale is pretty similar to that underlying credit cards. There are still areas, however, where Alibaba doesn’t have a presence and Huabei is not accepted. Hence Ant launched another service called Jiebei (just borrow) which allows users to borrow a small amount of cash and benefit from a certain interest-free period. With the enormous mass of data it gathers from all lines of businesses, Ant is able to build a meticulous digital profile of each user and quantify their credibility with a score called Sesame Credit. Both the lending limit for Huabei/Jiebei and the Sesame Credit score are dynamically readjusted on a monthly basis as new behavioral data stream in. At least in the realm of Alibaba, the Sesame Credit score acts like a passport that grants different users with different access to its product and service offerings.
Out of the 900 million Chinese working-age population, 400 million currently don’t hold a single credit card. Huabei and Jiebei provide an alternative method of accessing credit services for these people who are either unqualified for a credit card or are simply unwilling to apply for one. Ant utilizes its technology excellence to complement the current credit market by making it more inclusive. Both individual users and SMEs benefit from a higher level of liquidity and financial sustainability when they run out of cash. In this sense, it is indeed doing good as it increases consumer welfare. Ant is able to offer lending that would have been rejected by banks because it possesses much richer data related to personal credibility and has kept iterating on the risk assessment model to guarantee its predictive power.
Technically, what essentially distinguishes Ant or Alipay from traditional banks is simply its better knowledge of a person’s credibility. It is supposed to make better informed and less biased decisions. Unfortunately, the profit-maximizing nature of private firms has caused Ant to diverge from ethical good. The issues mainly center around the unjust design of its assessment algorithms, the contradictory nature between scoring and user data privacy, intransparency in disclosure, and the problematic granting of credit services.
Unlike heavily regulated financial institutions, Ant’s self-positioning as a technology company had got it away with strict scrutiny. Ant, therefore, enjoys a large room in modifying its credit scoring and lending allowance algorithms in order to best serve its businesses. In the traditional capital market, credit is measured by an entity’s past repayment records, asset holdings, liquidity health, etc. These all ensure that the debtors could and would honor the outstanding liabilities. Apart from these factors, however, one can find in the Sesame Credit page in Alipay many less related contributing elements. These are mainly about new products or services launched by Alibaba, like returning power banks on time, adhering to a fitness campaign held by Alibaba, collecting and consuming Sesame Points (virtual points that can be exchanged for credit services) from friends, etc. Alipay actually states explicitly that one can increase his/her Sesame Credit score by trying out these new offerings more frequently. There is no social consensus to the relevance between these factors and one’s credibility, and what’s more important, the scoring is actually biased towards heavy users when they are not necessarily more credible. One can argue that Sesame Credit is not trying to mimic the credit scoring in a capital market, but merely to govern the Alibaba realm and help it grow stronger. This argument is losing its validity since Alibaba’s influence on Chinese daily life is so deep-rooted and far-reaching, to the extent that it’s not easy to find a comparable alternative to its products and services (perhaps JD.com is one of the few exceptions). Alibaba’s market monopoly position in return eases its nerves when modifying the algorithms. In the meanwhile, more and more outside agents including Canadian & Latvian tourist visas and Airbnb are adopting Sesame Credit score as a credit testimony. Hence the scoring algorithm necessarily needs to be in accordance with the public will and its proximity to commercial purposes should be carefully investigated and re-evaluated by the authorities. It’s getting urgent that the score truly reflects a person’s performance record and solvency rather than involvement in the promotion campaigns of Alibaba’s new businesses.
A second issue is the contradictory nature between the scoring mechanism and privacy protection. Building trustworthy digital profiles requires as much detailed information about the users’ behaviors as possible, meanwhile as Alipay requires real-name authentication, the collected data are not anonymous. On the contrary, the protection over user data privacy requires “data minimization” and “de-identification”, as indicated by the GDPR principles. It’s simply impossible to produce a reliable credit score if the users choose not to allow Alipay to collect so much data, but accepting it puts their data privacy in danger.
Ant may find it painful to do trade-offs to resolve the above two issues, but at least it can easily increase its transparency in public disclosure. At present, Alipay users can only have a vague idea of how their Sesame Credit score and Huabei/Jiebei allowance are determined without knowing the exact ways in which their behaviors lead to the results. The insufficient transparency hinders users from properly adjusting their borrowing and consumption behaviors. From a Data Science point of view, this is a common pain point as the state-of-art algorithms generally lack interpretability. Nonetheless, unless the calculation is clearly understood, users can reasonably suspect that they are facing discriminatory treatment.
The final issue concerns Alipay’s problematic granting of credit services. Compared to conservative banks, Alipay seems much less selective and grants access to Huabei/Jiebei to basically anyone older than 18 years old. It actually simplified the application procedures and provided subsidies to attract users during the promotion. This results in a large number of users with low to no income relying on Huabei/Jiebei to satisfy their extravagant needs. Alipay doesn’t take much effort to inform users of the consequences of failing to repay the debts but instead uses beautifying commercial ads to lure people into using Huabei/Jiebei. As a consequence, many users have to go extreme ways (like loans for nudes and gangdom loans) to repay their high Huabei debts.
There is no doubt that Ant is infusing vigor to the financial system and increasing consumer welfare with its innovative technologies. However, we also need to pay extra attention to the risky practices and ethical issues in its current businesses, so as not to let the new business model transfer unbearable risks to both institutional and personal credit.