When Finance Moves Inside the Platform, Risk Moves With It
BigTech in finance is not simply a story about technology companies selling a few financial products. It is a story about large, data-centered platforms extending an existing relationship with users and merchants into payments, credit, insurance, investment products, and sometimes a much broader SuperApp. This opening chapter focuses on what makes that expansion different from ordinary fintech and why the important unit of analysis is often the wider group rather than the product carrying a particular license. The practical understanding this series develops is straightforward: follow the dependencies that connect the platform, its users, its data, its financial partners, and the infrastructure on which others rely.
That question matters because the same connection can create value and vulnerability. A platform may reach people who have thin credit files, lower the friction of a payment, or tailor a product using information that a traditional lender cannot easily observe. It may also make a user dependent on one interface, give one company unusual knowledge of behavior, or connect a technology failure to many regulated firms. These are not automatically opposite outcomes. They can be produced by the same architecture.
BigTech is a group with a financial extension
The IMF describes BigTech as a large technology conglomerate whose platform-based business model is organized around interactions among a large number of mainly retail users. Those interactions generate data, and the data support more services, which in turn generate more activity. The core business is usually outside finance. That distinguishes BigTech from a technology-enabled firm whose main business is already financial, even if that firm has become large and digitally sophisticated.
The distinction is more than a matter of vocabulary. A bank begins with a balance sheet and a set of financial obligations. A fintech usually begins with a financial service delivered through technology. BigTech can begin with search, social communication, e-commerce, hardware, advertising, mobility, or another nonfinancial activity, then use its existing distribution and information advantages to enter finance. The financial service is connected to a customer relationship that was built somewhere else.
The IMF identifies three forces behind this expansion. The platform has commercial reasons to add payments for transactions it already facilitates, lending for buyers or merchants, or insurance for users of its services. It has technological advantages in scale, application programming interfaces, cloud computing, and machine learning. And incumbent financial institutions may want to outsource parts of the customer relationship or distribution layer while retaining the balance sheet. This is the logic of embedded finance and banking as a service: the visible financial experience can sit inside a nonfinancial platform even when a licensed institution performs an important regulated function behind it.
BigTech firms do not all arrive with the same business model. Some rely heavily on data, some on fees and commissions, and some on hardware or a large installed device base. A social or messaging platform has a natural path into person-to-person payments; an e-commerce platform may have a stronger reason to offer merchant credit; a hardware ecosystem may control a convenient wallet interface. These differences change the way scale is acquired and the kind of dependency that can form. They also explain why a single rule aimed at “technology companies” may be too broad for one model and too narrow for another.
That arrangement makes a product label incomplete. Saying that a company operates a wallet does not reveal whether the wallet is merely an interface over a card network, a major domestic payment channel, or the entry point to a lending and investment ecosystem. Saying that a bank made a loan does not reveal who selected the customer, supplied the behavioral data, presented the offer, or shaped the incentives to approve it. The legal entity remains important, but it does not by itself describe where control and dependence sit.
Why the payment relationship changes the picture
Payments are usually the first bridge between a platform and finance because platforms already need users to exchange value. An e-commerce site needs a trusted checkout; a messaging service can make person-to-person transfers convenient; a mobility platform needs a low-friction way to pay. In advanced economies, BigTech may mainly provide a digital wallet that stores tokenized payment information while relying on existing card infrastructure. In many emerging market and developing economies, where card-based services are less pervasive, mobile money, QR payments, and real-time rails can make the platform a much deeper part of everyday exchange.
The payment relationship does two things at once. It gives the platform a distribution channel that users return to frequently, and it creates a stream of observations about economic behavior. Payment data can be combined with e-commerce or social information to build a more detailed picture of a user or merchant. That picture may help extend credit to people who lack a conventional credit history. The IMF associates services such as M-Pesa with wider access to financial services in Kenya, while also warning that inclusion benefits do not remove the risks of fraud, privacy loss, market concentration, or exclusion from a critical service.
The cases show why the local setting matters. The IMF reports that M-Pesa’s launch in Kenya was associated with financial-service access rising from 14 percent to 83 percent of the population between 2006 and 2019, while it describes Alipay and WeChat Pay as together holding close to 95 percent of China’s retail mobile-payment transaction volume. Those figures come through the note’s cited case studies and market sources, so they are illustrations rather than a universal measure of BigTech success. Their analytical point is sturdier than the rankings: where a platform becomes a routine route for value exchange, its payment role can shape later access to other financial services.
Scale is therefore not just a measure of popularity. It can affect substitutability. If a user can move easily to another payment method, a platform’s failure or bad decision may be inconvenient. If the platform is the practical route to payments, credit, or other daily services, the same decision may have consequences that are much harder to avoid. A product that is small when measured by its own balance sheet can still matter because many people depend on it, because it is connected to other services, or because few alternatives can replace it quickly.
This is where financial and operational interconnectedness enter the story. The IMF identifies four ways that BigTech can become relevant to financial stability: a single activity can become systemically important; several individually modest services can become important in combination; BigTech can become financially connected to incumbents through ownership or joint intermediation; and financial firms can become operationally dependent on a few BigTech providers of cloud, data, artificial intelligence, or other technology. The note’s detailed survey centers on consumer-facing finance, but it treats those technology services as part of the wider dependency map.
The four routes are related but not interchangeable. A payment service can be critical because users have nowhere else to go. A bundle can be critical because payment data makes lending more useful, while lending makes the payment account harder to abandon. A partnership can transmit a problem through contracts, ownership, or shared incentives even if the platform does not hold all the assets. And a technology service can create concentration without ever appearing in the customer’s financial product. A careful assessment must therefore ask both how large an activity is and what would happen if users or firms had to replace it quickly.
The Bank for International Settlements offers a complementary way to understand the mechanism. In its framework, digital technology can improve credit screening by turning behavioral information into a signal of repayment ability. A shared payment ledger might also route future sales toward repayment. But a ledger is not self-enforcing if a borrower can sell for cash or use an anonymous alternative outside it. The practical question is not only whether a platform knows more. It is whether the platform controls enough of the place where transactions occur to make leaving, hiding, or defaulting costly. That control may support access to unsecured credit, but it also gives the platform leverage over users.
The useful question is where dependence is created
It is tempting to ask whether BigTech in finance is good or bad. The evidence supports a more useful question: which dependency has been created, who controls it, and what happens if it fails or becomes difficult to replace?
This lens changes what counts as evidence. A user-count figure tells us about reach, but not whether another provider can substitute for the service. A loan-volume figure tells us about activity, but not whether the platform selected the borrower, a bank supplied the funds, or a payment account enforced repayment. A claim about efficiency says little about who receives the gain and who absorbs the loss when an algorithm is wrong, a fund cannot meet redemptions, or a system outage interrupts ordinary transactions. The important facts are relational: the contract, the data flow, the incentive, the fallback, and the authority able to intervene.
A payment interface can create convenience and a data advantage. The same interface can also become a gate through which a user must pass to reach credit or commerce. A bank partnership can bring a regulated balance sheet and broader distribution. It can also split incentives, leaving a platform rewarded for referrals while a bank carries much of the credit risk. A shared technology service can lower costs for many firms. It can also create a common operational point of failure. None of these outcomes follows from the product name alone; each depends on the relationships around it.
The IMF says that current financial-stability implications remain limited in most jurisdictions, even as BigTech grows quickly, especially in emerging markets. That qualification matters. The case for closer attention is not that every platform has already become a systemic institution. It is that rapid growth can move a service from optional to critical faster than supervisors, competitors, and users can adjust. Activity-based rules remain necessary, but they may miss the cumulative effect of a group that combines financial and nonfinancial activities.
The central map for this series is therefore a chain rather than a catalogue:
The chain does not say that every BigTech firm completes every link, or that every link produces the same result. It says where to look. The relevant boundary may be between a wallet and a payment system, commerce and credit, a platform and a bank, customer data and underwriting, or a local service and a cross-border group. The more links share one interface or one source of data, the less informative a product-by-product description becomes.
That is why the subject is neither “technology” in the abstract nor finance detached from technology. It is the changing organization of financial relationships around BigTech platforms. The benefits are real enough to explain adoption, and the risks are real enough to require more than a narrow license check. The appropriate response is not to assume that every platform should be treated like a bank, or that every partnership is harmless because a bank remains legally responsible. It is to match oversight to the dependency actually created, while keeping the evidence and the limits of the claim visible.
The first place to see both sides at once is the payment relationship: it is where a platform’s existing reach begins to shape who can exchange value, what the platform can learn, and which financial services can follow.