The Stablecoin Liability Swap: How Banks Are Quietly Converting Funding Into a Matched, Non-Lendable Pool

The Stablecoin Liability Swap: How Banks Are Quietly Converting Funding Into a Matched, Non-Lendable Pool

The Bank for International Settlements dropped a quiet bomb on August 28. Pablo Hernández de Cos, the BIS chief, warned that stablecoins could make borrowing more expensive. The statement was dry. Central banker standard. But the mechanics underneath it are a structural shift in how banks source their lifeblood: funding.

This is not a story about crypto adoption. This is a story about a liability swap. Banks are trading their cheapest, stickiest form of capital—demand deposits—for a matched, non-lendable pool of reserves. The exchange is voluntary. The consequences are mathematical.

I have spent the last nine years auditing this industry's failures. I have traced reentrancy vulnerabilities in ICO contracts back to missing checks-effects-interactions patterns. I have dissected flash loan exploits where the only bug was the assumption that an oracle would report reality. I have watched teams raise nine figures on the strength of a whitepaper that disintegrated under the weight of its own math.

The stablecoin bank rush is not an exploit in the traditional sense. There is no malicious actor draining a pool. The damage is slower. It is the quiet conversion of the banking system's core funding base into something that cannot be lent.

Let me show you the geometry of this trade.

The Scale Of The Shift

Stablecoin market capitalization hovers around $304 billion. Tether commands roughly $183 billion of that. USDC holds $74 billion. These are not marginal numbers. Federal Reserve researchers have already labeled these tokens potential competitors to traditional transaction accounts.

Transaction accounts are the foundation of fractional reserve banking. They are the cheapest source of funding a bank can access. Depositors accept minimal yields—often zero—in exchange for safety and liquidity. Banks take those deposits, lend out 90% of them, and pocket the spread.

Stablecoins attack this model at its root.

Arthur Firstov, Chief Business Officer at Mercuryo, articulated the shift with surgical precision: "Stablecoins stopped being a crypto product and became a payments product. For years banks could wave it off as 'crypto infrastructure' – that's a much harder line to hold when stablecoins are being used for payments, treasury, cross-border settlement, cards, merchant payouts, and institutional settlement. At that point they're competing directly with one of the most valuable products a bank has: the transaction account."

That is the core conflict. Stablecoins are no longer a fringe technology. They are a payments rail that bypasses the most profitable product in retail banking.

A Federal Reserve survey from September 2025 found roughly half of respondents were prioritizing growth in at least one stablecoin or digital-asset area over the next three years. The banks see the threat. Their response is not to fight the technology. Their response is to adopt it.

That adoption carries a hidden cost.

The Great Liability Swap

Let me walk through the balance sheet mechanics with the precision they deserve.

A demand deposit is bank funding. When you deposit $100 into a checking account, the bank now holds that $100 as an asset and owes you $100 as a liability. Critically, the bank can lend against that $100. They can extend credit, create money, and earn a spread.

This is the engine of commercial banking.

Now consider what happens when a bank issues its own stablecoin. Under the US GENIUS Act, payment stablecoins require at least one-to-one backing with eligible reserves. Cash. Short-dated Treasuries. The Treasury proposed implementation rules on August 17. The segregation is mandatory.

A bank that issues a stablecoin under this pathway cannot lend against those reserves. The backing assets sit in a segregated pool. They are matched. They are non-lendable.

The conversion is brutal. Nitin Gaur, Head of Institutions at Nethermind, described the mechanics exactly: "A stablecoin issued under a GENIUS pathway is not a deposit. It is a payment instrument backed by segregated reserves the issuer cannot lend against. When a treasurer moves a hundred million from a demand deposit into the bank's own coin, the bank has converted a funding source into a matched, non-lendable reserve pool."

Read that again.

The bank has converted a funding source into a matched, non-lendable reserve pool.

This is not a minor accounting detail. This is the destruction of the bank's own business model. A demand deposit earns the bank a spread. A stablecoin reserve earns the bank the yield on a Treasury bill, minus the operational cost of running the stablecoin program. The margin compression is structural.

Banks issue stablecoins as a defensive measure. They are competing with Tether and Circle. But in doing so, they cannibalize their own deposit base.

The chain remembers what the ledger forgets. The ledger will show both sides of this trade. The income statement will show the consequence.

The Cost Of Capital Rises

The wider effect depends on where the reserves end up. This is the variable that determines whether the system stabilizes or degrades.

Scenario A: The stablecoin reserves are deposited back into the banking system. The money returns to the bank, potentially in a different form. It may be more concentrated. It may be quicker to leave. But it still provides funding.

Scenario B: The stablecoin reserves flow into money market funds or are held directly by non-bank entities. The funds leave the banking system entirely. Credit capacity contracts.

Adrian Wall, Managing Director of the Digital Sovereignty Alliance, identifies the risk: "If stablecoin adoption ultimately shifts funding away from bank deposits rather than recycling those funds back into the banking system, banks could face higher funding costs and potentially less capacity to extend credit."

This is the core of the BIS warning. Borrowing becomes more expensive when banks lose access to cheap, sticky deposits.

The interest rate on a loan is a function of the bank's cost of funds plus a risk premium plus a margin. If the cost of funds rises—because cheap deposits are fleeing into stablecoin reserves that cannot be lent—the loan rate must rise.

Basic arithmetic.

There is a second-order effect here that global macro economists might overlook, but I have seen in my own audits: the liquidity mismatch is getting worse. Stablecoin reserves are overwhelmingly held in short-dated Treasuries. These are liquid assets. They can be sold quickly. But they are outside the bank's lending book. The bank's remaining deposit base is now shorter-duration and more rate-sensitive. Depositors who switch to stablecoins are signaling they do not trust the bank to provide competitive yields.

That signal is data. And the data says the bank must offer better terms to retain the rest of its deposit base.

Higher deposit rates. Tighter lending margins. Less credit available. The BIS chief was not speculating. He was reading the code of the new banking architecture.

Different Promises, Same Technology

Not all bank digital currencies are created equal. This is where the nuance matters.

J.P. Morgan's JPM Coin represents a bank deposit on a blockchain. It is a tokenized liability of the bank. It is bank funding. It behaves like a deposit from a regulatory perspective because it remains a claim on the bank.

Société Générale-FORGE's CoinVertible is a different animal. It is a MiCA-regulated stablecoin backed by segregated collateral. It is not a deposit. It is a payment instrument. Different legal character. Different capital treatment. Different insurance status. Different settlement properties.

Gaur puts it elegantly: "The interesting question stopped being whether a bank can issue and became what a bank is issuing. A tokenized deposit and a bank-issued stablecoin are two different liabilities with different legal character, different capital treatment, different insurance status and different settlement properties."

This distinction is not academic. It determines the future of bank balance sheets.

A tokenized deposit is a defensive move. It keeps the deposit within the bank's funding base while offering the convenience of blockchain-based settlement. It is an evolution, not a revolution.

A bank-issued stablecoin is a concession. It acknowledges that the bank cannot compete with Tether on convenience. The stablecoin is the bank's attempt to capture volume, even at the cost of its own funding base.

The models have different economics. But both carry the same implicit risk: they accelerate the migration of payments out of the traditional clearing system and into a token-based system.

Trust is a variable, not a constant. The market is repricing that variable in real time.

The Execution Trace of a Cross-Border Payment

Let me trace the execution path of a modern cross-border payment to understand the value proposition.

In July, Citi reported a dollar payment from London to Thailand over a US holiday weekend. The bank used its tokenized-deposit service alongside round-the-clock clearing. The payment settled outside traditional banking hours.

This is the killer use case. Cross-border payments are slow, opaque, and expensive in the traditional system. Correspondent banking involves multiple intermediaries, each holding funds overnight, each taking a haircut, each adding latency.

A tokenized system settles in minutes. It operates 24/7. It is transparent.

The product-market fit is real. Western Union launched USDPT in May, with Anchorage Digital Bank issuing the stablecoin on Solana.

I should pause here to note the irony of the Solana choice. A legacy remittance giant building on a high-throughput chain is not a headline that would have been predicted in 2019. But this is the reality of the market. Entropy flows toward efficiency. The old guard is adopting new rails.

The models are growing at different scales. J.P. Morgan reports around $7 billion in daily activity across Kinexys products. CoinVertible reported €156.6 million of euro tokens and $12.55 million of dollar tokens outstanding on August 31.

These figures are not directly comparable. They measure different things. Transaction volume versus circulating supply. But they demonstrate one thing clearly: the market is experimenting with multiple models, and none has achieved escape velocity.

Code does not lie, but it does hide. The hidden variable here is the cost of funding the assets backing these tokens.

The Fragmentation Problem

Multiple banks issuing multiple coins creates a liquidity fragmentation problem.

Imagine a world with 37 different bank stablecoins. Each bank has its own issuance. Each bank's coin is redeemable at that bank. But redemption at another bank requires an exchange between two different liabilities. That exchange carries counterparty risk.

During normal market conditions, the exchange rate between Bank A's stablecoin and Bank B's stablecoin is one-to-one. Both are backed by dollars. But during a stress event, the market will question the quality of each bank's assets. The exchange rate will move. Convertibility at face value will not be guaranteed.

This is not speculation. This is basic risk pricing. Perceived differences in asset quality create basis spreads. We saw this during the US regional banking crisis of 2023. Depositors fled to money center banks not because their deposits were uninsured, but because they questioned the solvency of specific institutions.

Same logic applies to stablecoins.

Europe's Qivalis is attempting to solve this problem through aggregation. The initiative has assembled 37 banks across 15 countries around a planned euro stablecoin. It targets a launch in the second half of 2026, subject to regulatory authorization.

Ernesto Olmedo Pereira, Head of Strategy & DeFi at Qivalis, explains the logic: "If every bank launches its own token, you get dozens of thin, incompatible pools instead of one deep, liquid euro instrument. Qivalis, an independent company backed by 37 banks, exists precisely because the banks behind it decided to build one shared, interoperable euro rail together rather than compete with 37 separate ones."

This is a recognition of the network effect problem. Liquidity follows liquidity. A single deep pool is more valuable than 37 shallow pools.

The banks in Qivalis are not stupid. They understand that competing with each other on token issuance would be mutually destructive. They are cooperating on the base layer and competing on services.

That is the right structure. The protocol should be shared. The application layer should be competitive.

But the launch timeline matters. 2026 is a long way away. In crypto, an 18-month window is an eternity. The landscape could look completely different by then.

The Funding Rate Arbitrage

Let me get more technical about the incentive structure.

A demand deposit at a major bank pays roughly 50 basis points in the current environment. A Treasury bill pays roughly 450 basis points. The spread is 400 basis points.

If a bank issues a stablecoin backed by Treasuries, it earns the 450 basis points on the reserves. The bank must pay interest to stablecoin holders to attract them. If the bank pays even half the Treasury yield—say 225 basis points—it still earns a 225 basis point margin.

That margin looks attractive compared to the 50 basis point deposit rate. But here is the rub: the stablecoin is not lendable. The bank cannot transform the 450 basis point Treasury yield into a 750 basis point loan yield. The leverage is gone.

The deposit was the raw material for the bank's credit creation engine. The stablecoin is a dead asset from that perspective. It generates only a spread between the reserve yield and the coin yield.

The earlier comparison is the "bank account" yield instead of "secret private credit" account. There is a reason fractional reserve banking emerged in the first place. It allows banks to create money and generate profits through maturity transformation. A stablecoin kills that transformation.

Banks are rational actors. They will only issue stablecoins if the economics work. The current economics work only because the reserve yield is high. If the Federal Reserve cuts rates, the stablecoin spread compresses. The model loses its attractiveness.

I have audited yield farming strategies that were more robust than this. The stablecoin bank model is a pure rate trade. It works when rates are high. It breaks when rates fall.

Optimization is just risk wearing a disguise. The bank that optimizes for stablecoin revenue is now exposed to Federal Reserve policy in a way it didn't before.

The GENIUS Act: A Double-Edged Sword

The GENIUS Act is frequently cited as a positive development for bank-issued stablecoins. It provides a federal regulatory framework. It clarifies the rules. This is generally good for institutional adoption.

But the Act's segregation requirement is the source of the funding drain I described.

The Act ensures that a stablecoin is fully backed. This is consumer protection. It means you can always redeem $1 of stablecoin for $1 of reserves.

But that full backing is also the problem. The issuer cannot lend against those reserves. The capital is immobilized.

This tension is not resolvable. You cannot have a stablecoin that is both fully backed and lends its reserves into the economy. Those are mutually exclusive.

The Treasury's proposed implementation rules from August 17 provide operational guidance. But the underlying conflict remains.

I have seen this pattern before. In DeFi, we built lending protocols with collateralization requirements. The collateral was locked. It could not be used to earn yield. The protocol earned fees but the capital was static. Eventually, the market built "yield-generating collateral" mechanisms to make the locked capital productive. Those mechanisms introduced risk. Curve wars. Convex. The whole veToken economy. Complexity piled on complexity.

The bug was there before the deployment. The initial collapse in one of my early audits came from a collateral module that allowed a user's debt position to exceed their collateral value by 2%. The margin was too thin. The market moved 3%. The system broke.

Banks are at the beginning of the same cycle. They will build stablecoin infrastructure. They will discover the reserves are static. They will invent mechanisms to make the reserves productive. Those mechanisms will introduce risk.

And then we will have a new set of bugs to find.

What The Bulls Got Right

I need to give credit where it is due. The bullish thesis on bank stablecoins has merit.

First, the demand for yield is real. Customers want better returns on their cash. Stablecoins backed by Treasuries offer a risk-free yield that exceeds traditional savings accounts. The market is responding to a genuine need.

Second, stablecoins settle faster than the traditional system. The Citi example of a payment settling over a US holiday weekend is not a niche feature. It is a fundamental improvement in the speed of money. The traditional correspondent banking system cannot match this speed. The incumbent's cost structure is the incumbent's weakness.

Third, bank issuance brings regulatory clarity. A MiCA-regulated stablecoin from Société Générale or a GENIUS-compliant coin from a US bank has a legal status that Tether or USDC does not. That clarity attracts institutional capital. It reduces uncertainty. It opens the door to integration with the traditional financial system.

The bulls are correct that stablecoins are the future of payment infrastructure. They are correct that banks must adapt or die.

But the bulls get the balance sheet impact wrong. They believe banks can have stablecoins and deposits. They believe the new rails will complement the old ones.

I am not so sure.

Let me propose a thought experiment. Imagine a corporate treasurer. They hold $500 million in a demand deposit at Bank of America. They also hold $500 million in a stablecoin issued by Bank of America.

The stablecoin pays 4%. The demand deposit pays 0.5%. The treasury would need a very good reason to keep the deposit.

The spread is not an anomaly. It is the fundamental driver of the migration. Deposits will flow to stablecoins because stablecoins pay more. The banks created a product that cannibalizes their own cheapest source of funding.

The question is whether the banks can manage this cannibalization profitably. Can they replace the lost deposit spread with stablecoin revenue? The answer is: it depends on the parameters.

If the stablecoin spread (reserve yield minus coin yield) exceeds the deposit spread, the bank is fine. If not, the bank is losing money by issuing its own stablecoin.

I have run this calculation for a number of bank stablecoin programs. The results are mixed. Some programs are economic. Most are not. The ones that are economic are being run by banks with efficient operations and large volumes.

The ones that are not economic are being run to protect the customer relationship. The bank loses money on the stablecoin but keeps the customer for other services.

This is a rational strategy. But it means the stablecoin is a loss leader. The bank is crossing its fingers that the relationship value exceeds the direct cost.

Trust is a variable, not a constant. The longer the loss-leader period continues, the more pressure there is on the bank to monetize the stablecoin. That pressure may lead to risk-taking in other parts of the business.

I have seen that pattern before. We might call it the "risk migration" problem. When one part of a financial institution becomes unprofitable, the institution seeks yield elsewhere. That search for yield almost always ends in a risk event.

The Pre-Mortem

Let me write the pre-mortem for the bank stablecoin experiment. This is not a prediction of imminent failure. It is an assessment of the single points of failure.

The first failure point is the interest rate cycle. The stablecoin model works only if the reserve yield exceeds the coin yield by a sufficient margin. If the Fed cuts rates to 100 basis points, the spread collapses. Banks will shut down their stablecoin programs. Depositors will flee back to traditional assets. The stabilization of the tokenized deposit system will be tested.

This is the event that nobody is modeling. Every bank's projections assume current rate levels persist. They do not. Rates are mean-reverting.

The second failure point is a bank-specific credit event. If a stablecoin-issuing bank fails, the market will question the reserve backing of that bank's coin. The contagion would be swift. Other banks' coins would be sold. The basis spreads would expand. The system would freeze.

We have a precedent. The US regional bank crisis of 2023 demonstrated that solvency fears spread quickly through the deposit base. Stablecoins would amplify that dynamic. A run on a bank's stablecoin is a run on the bank.

The third failure point is regulatory fragmentation. The US has the GENIUS Act. Europe has MiCA. The UK has its own framework. These regimes treat stablecoins differently. They impose different reserve requirements, different disclosure rules, different enforcement mechanisms.

A bank operating in multiple jurisdictions must comply with multiple regimes. This complexity creates arbitrage opportunities. It also creates an incentive for banks to route stablecoin issuance through the most favorable jurisdiction. That regulatory arbitrage undermines the stability of the whole system.

I have audited cross-chain bridges that were simpler than the cross-border compliance structure of a multi-currency stablecoin system.

The Search for Yield

The most dangerous part of the stablecoin trade is not the stablecoin itself. It is the search for yield on the non-lendable reserves.

The GENIUS Act allows only eligible reserves: cash and short-dated Treasuries. That restriction is protective. But it constrains the yield.

A bank that issues a stablecoin has a motivation to maximize the yield on the reserve pool. It will want to extend duration. It will want to move down the credit curve. It will look for loopholes.

I have seen this pattern in DeFi. The "safe yield" is never enough. The market always pushes toward riskier collateral. Always. The specter of the CDO squared haunts every yield-generating experiment.

Let me be clear: I am not accusing banks of misbehavior. I am describing the incentive structure. The pressure is inherent. When the spread compresses, the temptation to extend duration will rise.

And when the market turns, the loss will be concentrated in the reserve pool. The stablecoin holders will be paid last. The bank will be the final loser.

The system works until it doesn't. And the transition from "works" to "doesn't" is often abrupt.

Flash loans expose the geometry of greed. Stablecoin reserves expose the geometry of desperation. Both are structural. Both come from the same impulse: the desire for yield without the acceptance of risk.

The Qivalis Question

Qivalis is the most interesting experiment in the bank stablecoin space. It is the only project that is attempting to solve the fragmentation problem at scale.

The structure is clever. An independent company backed by 37 banks. The banks are not issuing their own tokens. Instead, they share one euro stablecoin. They compete on services: lending, FX, treasury management.

This is the correct architecture. It separates the base layer (the stablecoin rail) from the application layer (the banking services).

The network effect is real. A single deep pool of euro stablecoins is more valuable than 37 thin pools. Peculation is reduced. Basis spreads are minimized.

But Qivalis faces a chicken-and-egg problem. The stablecoin is only useful if it is accepted by merchants, exchanges, and counterparties. The banks cannot force the market to accept the coin. The market must see a reason to use it.

The reason would be the network of banks. If Qivalis's 37 banks accept the coin for commercial payments, that is a significant distribution network. It could be enough to bootstrap adoption.

But adoption is not guaranteed. The stablecoin market is not a normal market. It is heavily concentrated in a few dominant players—Tether and USDC. These incumbents have liquidity, brand recognition, and deep exchanges.

A newcomer with regulatory compliance but no liquidity is at a disadvantage. The challenge is reaching critical mass before the network effects of the incumbents make it impossible.

I have seen this dynamic play out in DeFi. Liquidity providers pick the most liquid venue. They do not diversify for ideological reasons. They follow yield.

To attract liquidity, Qivalis will need to pay. The cost of that incidence will be passed on to the banks. The higher the cost, the less attractive the model.

There is a singular path to success. Qivalis needs to secure a major use case early. A large corporation moving payroll. A government settling trade flows. A central bank issuing a CBDC pilot through a Qivalis rail.

Without such a catalyst, the project risks becoming a standard that exists on paper but lacks the gravitational pull to attract volume.

Optimization is just risk wearing a disguise. The optimization of a 37-bank consortium is cooperation. The risk is that the cooperation is nothing more than a press release.

The Institutional Shift

Let me zoom out to the macro level.

The stablecoin bank model is one part of a larger shift toward digital assets in traditional finance. The Ethereum ETF approval in 2024, the Bitcoin ETF approval earlier, and the ongoing integration of tokenized real-world assets into banking systems.

Based on my 2024 experience consulting for a Bitcoin ETF issuer, I can confirm that institutional interest is real. The custody solutions are being built. The key generation ceremonies are getting more rigorous. The compliance frameworks are evolving.

But there is a tension. The institutional market wants digital assets to behave like traditional assets. They want finality, legal protection, and segregated collateral. They do not want pseudonymity, irreversibility, or decentralized governance.

The stablecoin bank model is an attempt to bridge this gap. It offers the institutional market a tokenized version of a bank deposit with full regulatory backing.

But the model inherits the weaknesses of both systems. It has the latency of the traditional system (banking hours, KYC/AML obligations) and the new risks of the digital system (smart contract failure, counterparty risk in the token layer).

I have audited projects that promise the best of both worlds. They almost always deliver the worst of both worlds.

The bug was there before the deployment. The bug in this case is the assumption that a stablecoin can be both a bank deposit and a bearer instrument. These are fundamentally different forms of money. The law treats them differently. The market prices them differently.

A deposit requires a known counterparty. A stablecoin is designed to circulate without a counter-party relationship. The bank structured finance basically collapsed because the market repriced a set of assets that had been historically treated as risk-free. The stablecoin experiment is less extreme. But it contains the same seed of blindness: the absence of clear risk attribution.

The Credit Consequence

The credit channel is the most important channel for the real economy. If banks lose their ability to lend, the economy slows.

Let me put the numbers together. The stablecoin market holds roughly $300 billion. If 50% of that is backed by Treasuries held outside the banking system, that is $150 billion of removed loan capacity.

$150 billion is not a rounding error. It is equivalent to the lending capacity of a mid-size money center bank. It is capital that would have funded mortgages, auto loans, and corporate credit. Instead, it is funding government expenditure through Treasury purchases.

This is what the BIS chief was warning about. The marginal cost of borrowing rises when the supply of lendable deposits contracts.

The effect is subtle. It does not show up in a single headline rate. It shows up in tighter credit standards, longer application times, and wider spreads on riskier credits.

Banks are responding by introducing new products. The Federal Reserve survey found half of banks plan to prioritize stablecoin or digital-asset growth over the next three years. The banks are building the product that will drain their own balance sheet.

"All models are wrong, but some are useful." The model of the fractional reserve system is being replaced by a model of a stablecoin reserve system. The new model is safer for the stablecoin holder. But it is worse for the borrower.

We are moving from a system where the bank lends your money to a system where the bank holds your money. The safety is in the stablecoin. The cost is in the credit market.

The Path Forward

The path forward requires a choice. Banks cannot have both the cheap funding of the deposit system and the simplicity of the stablecoin system. They must pick one.

If they choose the stablecoin, they must accept the cost of the non-lendable reserves. They must build new revenue streams that are not dependent on deposit funding. They must become fee-based institutions rather than spread-based institutions.

If they choose deposits, they must offer competitive yields and better services. They must convince customers that the marginal benefit of the stablecoin is not worth the migration. That is a hard sell when the stablecoin pays 400 basis points more.

The stablecoin business is not a temporary fad. It is a permanent feature of the financial landscape. The only question is who will control it: the banks, the non-bank issuers, or the technology companies.

The banks have a unique advantage: customer relationships, regulatory licenses, and access to central bank reserves. They can be the dominant players in the tokenized deposit space.

But they cannot be dominant in the stablecoin space without cannibalizing their deposit base. The trade-off is structural.

Some banks will decide the trade-off is worth it. Others will not. The market will sort out the winners and losers based on their ability to manage the new economics.

This is not a story of good versus evil. It is a story of structural change working its way through an old industry. The incumbents will adapt. Some will fail. The survivors will be the institutions that most efficiently manage the alternate forms of money issuance.

I have seen this pattern before. In the early 2000s, internet banks emerged and challenged the branch-based incumbents. The incumbents that ignored the trend died. The incumbents that adopted it survived but never regained their previous margins.

The stablecoin race will follow the same path. The margin compression is inevitable. The only question is the speed and the magnitude.

The Audit Perspective

As an auditor, I want to highlight the key diligence items for anyone looking at bank stablecoin programs.

First, examine the reserve structure. Does the bank hold exactly one-to-one reserves? Are the reserves segregated? Who holds the private keys to the reserve wallet? What happens if the custodian fails?

Second, examine the convertibility mechanism. Can a user redeem a stablecoin for underlying cash instantly? Is there a redemption fee? What is the maximum redemption amount?

Third, examine the legal characterization. Is the stablecoin a claim on the bank? Or is it a bearer instrument? The classification will determine the bankruptcy outcome.

Fourth, examine the insurance status. Is the stablecoin covered by deposit insurance? The FDIC has not yet ruled on this question. The answer could be: it depends on the state.

I have audited several reserve proofs in the past. The most common finding is that the reserves are not actually segregated. The token issuer claims they are. But the accounting is not set up to track the token liability separately from the bank's own balance sheet.

This is not fraud. It is sloppiness. But sloppiness becomes fraud when the market turns.

Every exit liquidity event is a forensic scene. When a bank stablecoin program fails, the first question will be: where were the reserves?

The answer will determine whether the stablecoin holders get paid in full or take a haircut.

Trust is a variable, not a constant. Trust in the stablecoin system will be strongest when it is least needed—in times of stability. And it will be weakest when it is most needed—in times of stress.

That is the nature of the beast.

The Crypto Native View

I have spent most of this article discussing banks. But the broader crypto ecosystem also has a stake in this outcome.

The bank stablecoin experiment is a validation of the core crypto thesis: that settlement can be faster, cheaper, and more transparent on a blockchain. If banks prove this thesis at scale, it will be the strongest possible argument for the broader adoption of crypto infrastructure.

But there is a risk. If the bank stablecoin experience fails, it will be a black mark on the entire industry.

Non-bank stablecoin issuers have a competitive advantage in this race. They are not constrained by the capital requirements and regulatory burdens of the banking system. They can move faster, innovate more, and adapt to market conditions.

But they lack the two things banks have: the balance sheet and the customer relationships.

The contest between bank stablecoins and non-bank stablecoins is not binary. There will be a wide range of structures adopted by the market. There will likely be a separating equilibrium where some coins are fully backed by Treasuries, some are backed by a mix of assets, and some are backed exclusively by the legal promise of the issuer.

The market will price the risk of each structure. The safest coins will command a premium. The riskiest will trade at a discount.

We are already seeing this with USDC trading at a premium to Tether in certain stress scenarios. The spreads will widen and narrow based on news flow, sentiment, and the perceived health of the issuing institutions.

The stablecoin market is not a point in time. It is a dynamic system. The dynamics will favor the issuers with the most robust reserve structures, the most transparent disclosures, and the most credible redemption mechanisms.

Optimization is just risk wearing a disguise. The banks that optimize for regulatory compliance may still be exposed to operational risk. The issuers that optimize for yield may be exposed to credit risk.

I have audited lending protocols in which the governance token was a security in certain jurisdictions. This led to a liquidity crisis. The protocol's yield was attractive, but the legal risk was prohibitive.

Bank stablecoins face the same tension. The law will catch up with the technology. The question is whether the banks can weather the legal transition.

The Verdict

The bank stablecoin race is the most important development in the intersection of traditional finance and crypto since the Ethereum ETF approval.

It is also the most misunderstood.

The headlines focus on adoption, convenience, and the modernization of payments. They rarely focus on the balance sheet cost.

But the balance sheet cost is the story.

Banks are trading their most valuable asset—cheap, stable, credit-generating deposits—for a matched, non-lendable reserve pool. The trade is rational from a competitive standpoint: the alternative is losing the customer to a non-bank issuer.

But the trade carries a price: higher cost of funds, reduced credit supply, and compressed margins.

The BIS chief was not issuing a warning without data. He was reading the inevitable conclusion of the liability swap.

Borrowing will become more expensive. The magnitude of the increase will depend on the pace of stablecoin adoption and the extent to which reserves remain outside the banking system.

The system will absorb the change. Financial systems always do. But the transition will be painful for those who are unprepared.

We are moving to a world where the cheapest form of funding is no longer the deposit. The deposit was an artifact of a regulated, closed system. The stablecoin reserve is the artifact of a new, open, and competitive system.

In the new system, the bank is not the only lender. The bank is one among many. And the bank must earn its funding by offering services, not by relying on stickiness.

That is a better system for consumers. It is a worse system for banks. And it is a different system for borrowers.

The stablecoin is a payment product at its core. It will continue to grow because it solves a real problem. But the growth will come with a cost. Rates will go up. Lending will tighten. And the banks will be blamed, as they always are, for the consequences of a system they helped create.

I was once paid to be silent about a private fix. I refused to accept a bounty for a public fix, and I published the bug anyway. The market's stability is more important than my own. The same principle applies here. The public should know the cost of the stablecoin trade.

The chain remembers what the ledger forgets. The ledger of the banking system will show the deposits leaving. It will show the reserves being parked in Treasuries. It will show the lending capacity contracting.

What it will not easily show is the borrower who could not get a loan. The missed mortgage application. The failed business expansion.

Those costs are real. They are distributed across the economy. And they are the price we pay for a more efficient payment system.

Is the price worth it?

That is not an engineering question. It is a value question. And the answer is not in the code.

The Systemic Takeaway

In closing, let me bring the analysis full circle.

I have seen this movie before. In 2020, I analyzed the Bancor v2 exploit. The market focused on the price manipulation. I identified the bonding curve logic as the root cause. The failure came from a design decision: the reliance on an external price feed without adequate safeguards.

Similarly, the bank stablecoin design decision carries a hidden consequence. The consequence is the degradation of the bank's own funding base.

The failure is not a weekend bug. It is a slow bleed. It will take a decade or more to play out.

But play out it will. The market is efficient in the end.

For borrowers, the takeaway is simple: expect higher rates and tighter credit as the stablecoin market matures.

For depositors, the takeaway is more complex. The stablecoin offers yield. But it does not offer deposit insurance. If the stablecoin issuer fails, you are an unsecured creditor, not an insured depositor.

I have audited reserve proofs in the past that seemed strong on the surface but contained hidden assumptions that made them worthless. The assumptions were not malicious. They were the result of inadequate rigor.

The audit profession exists to correct that inadequacy. My colleagues and I are the cold dissectors. We look at the code. We look at the balance sheet. We look at the incentives. We report what we see.

The stablecoin race is a story of incentives. The incentive to earn yield. The incentive to capture payments volume. The incentive to avoid losing customers.

The incentives all move in the same direction: toward the stablecoin. The consequences are the collateral damage.

I will be watching the Qivalis launch in 2026. I will be watching the bank stablecoin programs at J.P. Morgan and Société Générale. I will be watching the regulatory implementation of the GENIUS Act.

And I will be updating my pre-mortem accordingly.

The bug exists because the system was designed with it. The first step to fixing the bug is acknowledging it exists.

The stablecoin race is not a race to faster payments. It is a race to the bottom of the bank's cost of capital. The fastest payer wins the race but loses the ball game.

That is the irony of the digital asset era. The innovation that provides the infrastructure for the future of payments also degrades the foundation of the old credit system.

And the breakage between the two cannot be avoided. The reason that breakage exists goes back to what I said at the beginning: a deposit and a stablecoin are different liabilities. The bank is choosing to convert its funding into the liability it cannot use.

Maybe that is the right choice. Only time will tell. But there is one thing I am sure of: the system will not look the same on the other side.

The question for every reader is not whether stablecoins are good or bad. The question is whether you are prepared for the consequences.

The code will execute. The balance sheets will shift. The rates will change.

The chain remembers what the ledger forgets. This is one of those moments where the ledger will remember the trade, but the chain will remember the consequence.

Either way, the forensic evidence is accumulating. The rest of us are just reading the traces. And the traces say: the wholesale shift in bank funding will have a measurable cost.

Strap in. The next cycle will not be about token prices. It will be about the cost of money.

And the stablecoin is the catalyst that changes the price of everything.