Embedded finance has been hailed as a paradigm shift in financial services, moving from being product-centric to customer-centric. Embedded finance generally refers to financial services integrated into nonfinancial Internet platforms, most notably in areas of lending and payment services (e.g., buy-now-pay-later [BNPL] services available as payment options on ecommerce platforms). Its development in recent years involves several players:
- Big Tech—Large technology enterprises with extensive, established customer networks. Their expansion into financial services has benefited enormously from having substantial existing customer bases and collecting and analyzing their customer data, which puts them on track to achieve scale rapidly.
- Independent fintech—Technology start-ups that are not affiliates of Big Tech and focus on cloud-based technology solutions to enable digital financial services. This can include payment gateway services such as Paypal and other financial technology enterprises unaffiliated with Big Tech. Unlike banks’ traditional technology vendors, these newly emerged fintech enterprises provide out-of-the-box solutions and extensively leverage application program interfaces (APIs), which support banks in their efforts to extend financial services beyond their branch networks more flexibly and effectively.
- Banks—Provisioning financial services in most countries, if not all, requires service providers to obtain a specific form of license in advance. However, these licenses usually come with regulatory responsibilities from which Big Tech or fintech enterprises aspire to be free (e.g., the US Truth in Lending Act [TILA, Regulation Z]). As such, banks are usually eager to fill the regulatory gaps in between (e.g., banks act as the underwriter to loans facilitated by Big Tech/fintech enterprises that do not hold any lending licenses).
The interrelations among these main players can fit into three different business models, as depicted in figure 1:
- Big Tech and bank model—Big Tech enterprises partner with banks to offer tailor-made financial services for users on Big Tech platforms. Data on user profiles and behaviors on these platforms, including nonfinancial data, can provide better customer insights for calibrating the financial services deployed on these platforms. One example of this is the Apple card, which is issued by Goldman Sachs exclusively to Apple users. Certain parts of users’ transactions are experiences with Apple that will be used for assessing and granting credit for the Apple card.1
- Fintech and bank model—Fintech enterprises provide banks with digital-native technology solutions to tap into unserved or underserved needs arising from nonfinancial ecosystems either online or offline. One example is Affirm’s buy now, pay later (BNPL) option. Affirm integrates installment loan servicing capability into its payment solution, which is then embedded into a third-party merchant platform. Customers request loans via Affirm, which performs identity verification and credit eligibility checks2 before passing the loans to banks for issuing.3
- Big tech and fintech and bank model—Some fintech enterprises develop standardized application programming interfaces (APIs) as intermediaries between Big Tech and banks (i.e., open banking). With these APIs, banks can be relieved from the burden of adapting to myriad legacy banking architectures, which becomes the responsibility of specialized fintech enterprises. One example is the relationship between Google Pay and Plaid. Upon user consent, Google Pay can work with Plaid to obtain transaction and balance information. This gives Google Pay more user insights to use to become a one-stop solution for managing users’ personal finances. To facilitate this, Plaid creates a secure connection to financial institution login information to access the accounts.4
Embedded Finance Risk Areas for Banks
Embedded finance seems to benefit all parties involved. But no business model is perfect, and embedded finance is no exception. More deliberate consideration is warranted for banks and other financial institutions involved, especially for the three main areas of risk: strategy, credit and reputation.
Strategy Risk: Overdependence and Accountability Asymmetry
The nature of partnership between banks and technology enterprises can sometimes be uncertain in terms of who is accountable for the financial services being offered.
For example, from a regulatory perspective, BNPL services are, ultimately, installment loans issued by the bank. But from the customer’s perspective, banks are less visible than the Internet platform facilitating user acquisition and credit assessment. In this case, the question for banks is whether they can be accountable for such loans on the balance sheet without fully owning the customer relationship and risk management process.
Accountability asymmetry of this kind can potentially introduce systemic risk to the financial sector if the partnership between the bank and Big Tech/ fintech scales up to a more dominant role in the sector. This is especially the case for community banks and credit unions. Given their relatively smaller scale, concentrating on partnering with third-party Internet platforms could put a significant portion of their balance sheet at risk when adverse events from partnering platforms cause contagion effects beyond control of the bank. Even when these effects are not imminent, disproportionate exposure to third parties still calls into question the independence and sustainability of the bank.
The nature of partnership between banks and technology enterprises can sometimes be uncertain in terms of who is accountable for the financial services being offered.
Credit Risk: The Untested Powers of Data and Algorithms
Despite market differences, lending is typically considered one of the most common forms of service offered in embedded finance. For example, Ant Group, known for its full range of financial services embedded into the ecosystem of Chinese ecommerce giant Alibaba, disclosed in its 2020 IPO prospectus that its biggest source of revenue was from credit-related services (39.5 percent of total revenue in the first half of 2020), which was even higher than the revenue from its digital payment services (35.9 percent).5 Meanwhile, it managed loan delinquency within reasonable levels (e.g., only 2.23 percent of its small business loans and 2.7 percent of its consumer loans were more than 30 days past due as of 30 September 2020, despite the negative impact of the COVID-19 pandemic).6
Embedded finance platforms, including Ant Group,7 prefer to boost their success with risk management capabilities driven by big data and proprietary algorithms. However, their success can also be attributed, at least to some extent, to the sizeable unserved or underserved market demand for credit in China. One way to understand this impact is to look at credit registry coverage as a proxy for existing credit service penetration by banks, which rely on credit registry data for underwriting. Credit registry coverage reports the number of individuals and firms listed in public/private credit registry with current information on repayment history, unpaid debts or outstanding credit. In figure 2, this number is expressed as a percentage of the adult population.
Data and algorithms have played a large role in the rise of embedded finance platforms such as Ant Group.
According to World Bank statistics, in 2013 (the year before Ant Group launched credit services), credit registry coverage was 30.2 percent for China and 100 percent for the United States.8 This metric implies that Chinese technology enterprises such as Ant Group can start with an untapped and larger share (69.8 percent) of the market, while banks, due to reliance on credit registry data, mainly focus on the smaller 30.2 percent share of market. Given the sheer volume of untapped customers, Chinese technology enterprises have a higher chance of success of choosing good customers (i.e., those with a higher willingness and ability to repay debts) among this population. This is where big data and proprietary algorithms can play a greater role than in a crowded market already well served by banks.
Data and algorithms have played a large role in the rise of embedded finance platforms such as Ant Group. But as previously untapped populations have been served by embedded finance platforms, which also report to centralized credit registry agencies, this population has been gradually covered by public/ private credit registries, putting embedded finance platforms and banks on a more level playing field. This demonstrates the real power of proprietary data and algorithms as they go through different economic cycles. As Warren Buffet, chief executive officer of Berkshire Hathaway and one of the world’s richest men, said, “It’s only when the tide goes out that you learn who’s been swimming naked.”9
Even though customers are acquired through partnering with Internet platforms, banks should seek to gain more control in managing customer relationships of embedded finance.
Reputation Risk: Uneven Expectations Between Banks and Tech Organizations
Customers often have higher expectations for licensed financial institutions such as banks than for technology organizations.
Based on these expectations, when customer dissatisfaction is encountered in an embedded finance context, the complaint is often more likely to be geared toward the banks involved rather than the technology organization. This can negatively affect the banks if disgruntled customers spread negativity across social media.
In 2019, Goldman Sachs was accused of discriminatory behavior regarding its credit card services for Apple users (i.e., Apple Card), where some Apple Card customers claimed they received longer lines of credit while others did not. Allegations that the discrepancy was gender based began appearing on social media and later were reported by mainstream media. Though Apple Card was co-developed with Apple, Goldman Sachs, as the licensed card issuer, seemingly took a bigger share of the blame.10
Risk Mitigating Approach
Despite the challenges banks face, there are some high-level takeaways to help them steer through the uncharted waters of partnership with technology enterprises.
Know the Partner
Knowing the customer is a well-known principle for the customer onboarding process in banks. But it is equally important, if not more so, for banks to know their partners in running embedded finance, given that financial services are now premised on partner platforms. The checklist of actions banks can take to get to know their partners should include:
- Performing due diligence for partnering platforms (similar to banks onboarding a business partner rather than a tech vendor)
- Calibrating business plans with reasonable expectations for all parties involved, and defining the appropriate risk appetite and exit strategy for different contingencies
- Closely monitoring the operational performance and strategic developments of partnering Big Tech/fintech platforms and promptly following up on unexpected events
Reclaim Customer Relationships
Even though customers are acquired through partnering with Internet platforms, banks should seek to gain more control in managing customer relationships of embedded finance, including having greater control over customer acquisition strategy and access to data platforms collected about users. To reach this goal, banks can build on their experience with regulatory requirements and public confidence in licensed financial institutions to exert greater influence on their Big Tech/fintech partners. This could prove even more valuable amidst growing public distrust and regulatory scrutiny over Big Tech.
Diversify Partner Base
The scale of Big Tech’s ecosystem makes the opportunity to partner with any Big Tech enterprise seem irresistible to banks. But the more dominant the role that Big Tech plays in its sector, the less bargaining power banks may retain in such partnerships. As a mitigation approach, banks should explore possibilities to partner with challengers in each target technology sector. As challengers rise to outcompete market leaders, they will need partnerships with banks as catalysts for growth. This could potentially create more leeway for banks to gain greater control over product design and customer visibility when offering financial services through these Internet platforms.
Build a Digital Team
Digital transformation within a bank is driven by people rather than technology. Most technology enterprises have a deeply rooted agile culture that provides for faster time to market at a smaller cost (i.e., making small, incremental changes [iterations] to products and services, but at a much faster pace [e.g., several iterations a week]). To enable financial services to run seamlessly on nonfinancial platforms, banks may need to internally adopt this agile culture in some way, at least within the team responsible for running embedded finance. This may not happen overnight, but banks can start by building dedicated teams that consist of both existing resources from internal teams and seasoned professionals recruited from the technology sector. Further, such teams can serve as champions for the bank’s digital transformation journey and provide testing grounds for banks to learn how to deploy agile culture to other parts of the business.
Go Digital From the Core
Embedded finance is not the end of a bank’s digital transformation journey; it is simply a means to an end. For banks, this is a unique opportunity to closely observe how technology start-ups develop and launch digital-native services in a fast-changing environment, as opposed to banks’ conventional approach of digitizing existing services (i.e., bringing services online). But there is more that can be done by banks on this journey. Embedded finance so far has only scratched the surface of banking. Many parts of banking services (e.g., private banking) still run with human-to-human interactions, but these areas call for more digital elements as customers grow accustomed to digital and seamless financial services. In addition, banks must build the digital- native mindset into their core competence. Driven by short-term incentives, banks may leverage embedded finance as an engine for achieving key performance indicator (KPI) targets, rather than as an incubator for nurturing the digital-native mindset within the organization. Failure to accomplish the latter also may gradually put banks at risk of being a pure bank- license provider for embedded finance.
Conclusion
Embedded finance platforms have expanded into financial services, but they are not ready to completely replace the role banks play today. So far, Big Tech and other technology enterprises have worked with embedded finance to magnetize and monetize the network effects they have already built while trying to avoid being regulated as banks. These regulatory gaps are expected to continue, at least for the foreseeable future and, thus, make room for banks to survive and even thrive in the space of embedded finance. Among these banks, those that embrace embedded finance from the core in a well-informed and prepared manner will be better positioned for the future.
Embedded finance platforms have expanded into financial services, but they are not ready to completely replace the role banks play today.
Endnotes
1 Apple, “Apple Card and Privacy,” 26 July 2021, http://support.apple.com/en-us/HT210662
2 Affirm, “Affirm Privacy Policy,” 18 August 2020, http://www.affirm.com/privacy
3 Affirm, “Affirm’s Lending Partners,” http://www.affirm.com/lenders
4 Google, “Common Questions About Plaid,” http://support.google.com/googlepay/answer/10193349
5 Ant Group, Stock Code: 6688, H Share IPO, Hong Kong, China, 27 October 2020, http://www1.hkexnews.hk/listedco/listconews/sehk/2020/1026/2020102600165.pdf
6 Ibid.
7 Ibid.
8 World Bank, “Private Credit Bureau Coverage,” Doing Business Project, 2004–2019, http://data.worldbank.org/indicator/IC.CRD.PRVT.ZS
9 Maxfield, J.; “Warren Buffett: How to Avoid Going Broke,” The Motley Fool, 2 August 2014, http://www.fool.com/investing/general/2014/08/02/warren-buffett-broke.aspx
10 Vigdor, N.; “Apple Card Investigated After Gender Discrimination Complaints,” The New York Times, 10 November 2019, USA, http://www.nytimes.com/2019/11/10/business/Apple-credit-card-investigation.html
Terrence Cai, CISA, CIA, FRM
Is an internal auditor at a mid-sized commercial bank in China. He helps implement innovative analytic solutions to address business and risk challenges, supporting governance of the bank.