The Bank for International Settlements issued a warning on August 28 that reads less like a forecast and more like a ledger entry. Pablo Hernández de Cos, the BIS chief, stated that stablecoins could make borrowing more expensive. The market cap of these digital assets has crossed $304 billion, with Tether holding $183 billion and USDC at $74 billion. The Federal Reserve's own researchers now classify these tokens as direct competitors to traditional transaction accounts. This is not a prediction. It is a balance sheet statement waiting to be reconciled.
For years, the banking sector dismissed stablecoins as a crypto-native experiment. That dismissal is no longer operationally viable. Arthur Firstov, Chief Business Officer at Mercuryo, articulated the shift with clinical precision: stablecoins have stopped being a crypto product and become a payments product. When tokens are used for treasury operations, cross-border settlement, merchant payouts, and institutional clearing, they are no longer adjacent to the banking system. They are inside it, competing directly with the most valuable product a bank owns: the transaction account.
A Federal Reserve survey from September 2025 confirms the industry's response. Roughly half of respondents prioritized growth in at least one stablecoin or digital-asset area over the next three years. The question is no longer whether banks will enter this market. The question is what happens to their existing liabilities when they do.
The Liability Transformation Problem
The distinction between a tokenized deposit and a bank-issued stablecoin is not semantic. It is structural. J.P. Morgan's JPM Coin represents a bank deposit on a blockchain. Société Générale-FORGE's CoinVertible is a MiCA-regulated stablecoin backed by segregated collateral. Both use similar technology. Both carry fundamentally different promises to the customer and, more importantly, to the bank's own balance sheet.
Nitin Gaur, Head of Institutions at Nethermind, framed the issue correctly. The interesting question stopped being whether a bank can issue and became what a bank is issuing. A tokenized deposit remains bank funding. It is a liability the bank can lend against. A stablecoin issued under the US GENIUS Act pathway is not a deposit. It is a payment instrument backed by segregated reserves that the issuer cannot touch. The Treasury proposed implementation rules on August 17, and those rules carry a specific consequence.
When a corporate treasurer moves $100 million from a demand deposit into the bank's own stablecoin, the bank has converted a funding source into a matched, non-lendable reserve pool. The money is still on the balance sheet, but it is no longer available for credit creation. This is the core mechanism by which stablecoin adoption raises borrowing costs. It is not about competition for customers. It is about the destruction of lendable capital.
I have spent years auditing DeFi protocols where this exact dynamic plays out in miniature. Liquidity pools that look deep on the surface often have a structural flaw: the assets are there, but they are not available for productive use. The same logic applies to bank balance sheets. A reserve pool is not a loan book. It is a parking lot.
The Funding Cost Arithmetic
The wider effect depends on where the reserves ultimately reside. If stablecoin reserves are deposited back at banks, they can still provide funding, although the concentration risk increases. Money that arrives via a stablecoin issuer is more concentrated and quicker to leave than a diversified retail deposit base. Adrian Wall, Managing Director of the Digital Sovereignty Alliance, identified the systemic risk: if stablecoin adoption shifts funding away from bank deposits rather than recycling those funds back into the banking system, banks face higher funding costs and reduced capacity to extend credit.
The math is straightforward. Banks earn the spread between what they pay for deposits and what they charge for loans. If stablecoins replace a portion of the deposit base, the bank loses a cheap, stable source of funding. It must replace that funding with wholesale markets, which are more expensive and more volatile. The cost of that replacement is passed on to borrowers. This is not a hypothetical scenario. It is the mechanical consequence of converting demand deposits into segregated reserve pools.
My own work on on-chain liquidity analysis has shown me that capital efficiency is the hidden variable in every market. The same principle applies here. A bank with $10 billion in stablecoin liabilities and $10 billion in segregated reserves is not the same as a bank with $10 billion in demand deposits. The former has zero lendable capital from that pool. The latter has perhaps $9 billion. The difference is the credit that never gets created.
The Fragmentation Vector
There is a second-order problem that the industry is only beginning to confront. As more banks issue their own tokens, the market fragments into dozens of thin, incompatible liquidity pools. Users must exchange one bank's token for another, and conversion at face value is not guaranteed during market stress. This is not a technical problem. It is a liquidity problem with technical symptoms.
Europe's Qivalis consortium is attempting to solve this by construction. The group has assembled 37 banks across 15 countries around a single planned euro stablecoin, targeting a launch in the second half of 2026. Ernesto Olmedo Pereira, Head of Strategy & DeFi at Qivalis, articulated the rationale: if every bank launches its own token, you get dozens of thin, incompatible pools instead of one deep, liquid euro instrument. The consortium exists precisely because its member banks decided to build one shared rail rather than compete with 37 separate ones.
This is the correct architecture. A shared coin creates a deep, liquid market. Banks then compete through the services surrounding that money—foreign exchange, corporate lending, treasury management—rather than through the token itself. The coin becomes a utility. The services become the product.
But there is a catch that the consortium model does not fully address. The shared coin still represents a liability transformation. When a customer moves money from a demand deposit into the shared stablecoin, the issuing bank loses lendable capital. The fact that 37 banks share the infrastructure does not change the underlying balance sheet mechanics. It only changes who bears the cost.
The Contrarian View: Correlation Is Not Causation
The prevailing narrative is that stablecoins are a threat to bank profitability. The data supports this, but only partially. The real risk is not the existence of stablecoins. It is the structure of the liabilities they create. A tokenized deposit that remains on the bank's balance sheet as lendable funding is not a threat. It is an efficiency gain. A stablecoin that requires segregated, non-lendable reserves is a different animal entirely.
The GENIUS Act pathway creates a specific kind of liability. It is a payment instrument, not a funding source. The bank cannot lend against it. The reserves sit in a matched pool, earning perhaps a Treasury yield, but generating no credit. This is the mechanism by which borrowing becomes more expensive. It is not about competition. It is about the destruction of lendable capital.
But here is the counter-intuitive angle. The banks that embrace stablecoins may be making a rational short-term decision. They are trading a portion of their lending capacity for a new revenue stream. The fees on stablecoin issuance and payment processing can be substantial. The question is whether those fees compensate for the lost interest margin on the converted deposits. In most cases, they do not.
This is where the data gets interesting. J.P. Morgan reports around $7 billion in daily activity across its Kinexys products. CoinVertible reports €156.6 million of euro tokens and $12.55 million of dollar tokens outstanding. These figures measure different things—transaction volume versus circulating supply—so they cannot establish which model is winning. But they do establish that the volume is real. The question is whether it is profitable.
The Takeaway: Watch the Balance Sheet, Not the Headlines
The stablecoin race is not about technology. It is about the structure of bank liabilities. The banks that understand this will design products that preserve lendable capital. The banks that do not will find themselves with large stablecoin businesses and shrinking loan books.
The next signal to watch is not the market cap of Tether or USDC. It is the ratio of stablecoin reserves to bank deposits at the largest issuers. If that ratio rises, expect funding costs to rise with it. If it stabilizes, the market may have found an equilibrium.
Check the logs, not the tweets. The balance sheet is the ultimate ledger, and it does not lie. Code is law; hype is just noise. The banks that treat stablecoins as a liability transformation problem will survive. The banks that treat them as a marketing opportunity will not. The data will tell us which is which, but only if we are willing to read it.