The Q1 2025 attestation reports across the stablecoin sector have resolved a question the market has avoided phrasing directly. The most profitable businesses in digital assets are no longer exchanges, and they are no longer DeFi protocols. They are stablecoin issuers monetizing the reserve balances their users deposited in exchange for tokens redeemable at par. Tether's most recent disclosure showed net profits north of $4.5 billion, with the overwhelming share attributable to interest on U.S. Treasury holdings. Circle, which runs a reserve book concentrated in short-dated government instruments, operates on the same economics even if its disclosure windows differ. These are documented facts. They appear in signed attestation reports and monthly transparency statements. They are not extrapolations or price forecasts.
The same signal is visible on the asset management side. BlackRock's tokenized fund product, BUIDL, crossed the half-billion-dollar threshold within weeks of its launch. Franklin Templeton's OnChain U.S. Government Money Fund has been live since 2021 and now manages a multiple of its early-stage assets. Both products treat the blockchain as a settlement rail for a money market instrument that would otherwise exist in a conventional brokerage account. The yield is the product. The chain is the distribution channel. The phrase the industry uses for the overall shift is 'stablecoin reserve yield.' A more accurate phrase, from a forensic accounting standpoint, is banking.
The mechanics of a stablecoin are simple. A user deposits one dollar. The issuer holds the dollar in a reserve account. The user receives a token that the market treats as worth one dollar. For the first several years of the industry, this arrangement was a cost center for the issuer. Reserve assets were held in cash or short-term instruments. The yield was negligible. The economic value was the settlement utility of the token, not the income on the reserve.
That changed when the Federal Reserve raised its policy rate from near zero to above five percent over roughly eighteen months. A reserve book that formerly earned a few basis points now earns the full policy rate. For an issuer with a market capitalization in the tens of billions of dollars, the annualized income on reserves jumps from tens of millions to billions. This is the net interest margin of a bank. A bank accepts deposits, deploys them into earning assets, and retains the spread. A stablecoin issuer does the same, with one difference. The issuer does not carry the capital requirements, the liquidity coverage ratio, or the lender-of-last-resort backstop that form the regulatory contract of a commercial bank.
Ten years ago, stablecoin reserves were treated as a formality. The market assumed they existed. The market rarely audited them. The Terra collapse in 2022 shifted the spotlight to the audited reserve model, and regulatory pressure forced the largest issuers to publish increasingly granular reserve breakdowns. By 2024, the monthly reserve report had become the most read financial document in the sector, not because users suddenly cared about the balance sheet, but because the profitability of the industry had become legible in it. When the reports began showing billions in interest earnings, the banking analogy ceased to be a metaphor.
The average user of a stablecoin is lending the issuer their dollars without requiring the issuer to pay interest, and the issuer is converting that free liability into Treasury yields. This single transfer explains why the industry's center of gravity has moved from trading revenue to asset management fees.
A tokenized money market fund is a distribution upgrade, not a new instrument. The underlying asset remains the Treasury bill or the repurchase agreement. The chain carries the ownership record. Yet the consequences for the liquidity structure of the crypto market resemble a bank issuing claims against a reserve pool. The token is a bearer instrument for a fractional claim on a pool of assets, and the speed of settlement runs far ahead of the speed of verification.
The verification cadence for the reserve pool is measured in days or weeks. The settlement cadence for the token is measured in seconds. During a crisis, a fund that advertises T+0 settlement on the front end will discover a back-end queue if the underlying redemption mechanism, connected to traditional market infrastructure, takes the customary T+1 or T+2 cycle. The friction between on-chain settlement time and off-chain asset settlement time is a liquidity gap. That gap is exactly the mechanism that produces a bank run in the conventional financial system.
The entire stability argument for the banking-ification of crypto rests on the reserve attestation stack, and that stack is not strong enough for the payload it carries. The dominant format remains the monthly or quarterly signed statement from an accounting firm, followed by a spreadsheet of reserve assets. This is a document-based verification system applied to an asset class whose defining feature is continuous, real-time transaction flow. The mismatch matters because the worst-case scenario in this business is not a slow deterioration. It is a sudden mass redemption event. A monthly attestation does not de-risk an intraday panic.
In 2022, the FTX collapse produced a wave of exchange proof-of-reserves commitments that were inconsistent and easily gamed. Some implementations excluded liabilities from the tree. Some included illiquid assets at face value. Some published no custody list at all. The exercise lowered trust rather than raising it because the standards were ambiguous. The identical failure mode is now visible in the stablecoin reserve industry. A Merkle-tree proof of liabilities can be produced in seconds. A signature from a custody bank can be timestamped on chain. The asymmetry between what the user sees and what the issuer holds is technically bridgeable. It is not bridged because the incentives of issuers favor minimal disclosure.
A reserve report is a snapshot, not a promise. From my own technical verification experience in this market, building wallet-tracking scripts throughout the NFT wash-trading cycle of 2021, the lesson was that the transaction graph reveals what the summary statistics hide. A floor price chart looks healthy while sixty percent of the volume is circular trades between wallets owned by the same operator. A reserve report has the same blind spot: a haircut schedule that assumes perfect liquidity; a custody list where one name carries the majority of the assets; a liquidity calculation that never runs the scenario where the counterparty fails. The summary statistic hides the concentration. The attestation certifies the snapshot. Nothing certifies behavior under the next hour.
Tokenized securities require a token standard that embeds compliance at the protocol level. ERC-3643, finalized in 2022, implements identity management, investor whitelisting, and transfer restrictions as enforceable logic. Adoption remains partial. Several major tokenized fund products instead maintain compliance in the issuer's back office and use the token as a receipt for an off-chain record. This works for a buy-and-hold portfolio. It fails for composable financial use.
A fund token that is a database receipt cannot be posted as collateral in a DeFi lending protocol without substantial legal uncertainty about the enforceability of the custody arrangement. The result is a two-tier market. One tier is fully integrated with open finance: tokenized Treasuries that can be borrowed, lent, and utilized as collateral in on-chain money markets. The other tier is a closed token: compliant, secure, but inert. The industry is building both tiers simultaneously. Whether tokenized funds genuinely expand the scope of on-chain credit depends on which tier ultimately wins the liquidity.
The Howey test shadows every tokenized fund product. A token that represents a share of an actively managed money market fund is a security under existing precedent. That classification triggers registration requirements, investor accreditation filters, and custody obligations that most crypto-native platforms are not logistically prepared to execute. The gap between the user experience and the legal requirements is growing even as the products gain traction. This is not a reason to avoid the sector. It is a reason to demand that the legal layer be designed into the token from day one, rather than bolted on after the first regulator inquiry.
The phrase 'balance sheet management' entered crypto commentary as a fashionable term, but its technical weight is substantial. If an issuer manages duration, rehypothecation, and liquidity coverage on the ledger, the ledger must understand the asset metadata. A smart contract does not know the difference between a ten-year Treasury and a seven-day repurchase agreement unless that metadata is fed into the contract. If the metadata feed is incorrect, the risk model collapses and the collateral ratios become fiction.
During my contract-audit work in 2020, I found a lending protocol whose interest rate calculation appeared correct on first read but suffered precision drift in low-probability branch conditions. The error surfaced only through a deep trace of the transaction history. The reserve balance sheet problem shares this property. The failure mode lives in a composability edge case, not the happy path shown in the dashboard. The interface between traditional accounting systems and chain-native data is the least discussed component of the banking-ification stack and the one with the largest gap between what the products promise and what the infrastructure delivers. A programmatic balance sheet without a verified data feed is a spreadsheet with better market access.
The conventional reading of this trend is maturation. The more accurate reading is that crypto is importing the risk profile of banking without importing the safeguards of banking. A bank faces a run when depositors panic about the quality of the asset book. The depositor's guarantee comes from federal deposit insurance and, behind it, the central bank's discount window. No such guarantee exists for a stablecoin holder. The token is redeemable only to the extent the issuer's reserve assets are liquid at that moment. There is no institution standing behind the issuer as a lender of last resort. There is no insurance fund.
The March 2023 USDC depeg is the clearest evidence of what this asymmetry does in practice. The market did not panic about the smart contract code. It panicked because a custody bank held part of the reserve pool and the deposit insurance limit was a quarter of a million dollars. The token depegged to $0.87 in two days. The audit trail was complete. The asset's liquidity was not. The redemption mechanism is designed for normal markets. Its vulnerability is a crisis in the traditional settlement layer, which is precisely the scenario that triggers a banking panic.
Picture the next sequence in concrete terms. The federal funds rate eases by fifty basis points. The forward curve reprices. A fund that invested user deposits in longer-dated instruments to capture a pickup over the reserve rate now faces mark-to-market losses. The redemption request arrives in the same hour that the bond market is selling off on a liquidity shock. The stablecoin issuer must sell assets to meet the redemption wave, but the assets it holds are exactly the assets that are losing value. It becomes a forced seller into a falling market, which accelerates the decline, which triggers more redemptions. This is a textbook bank run, recreated without a central bank.
This is not a theoretical scenario. The industry has already lived through two dry runs of uninsured depositor behavior in the space of thirty months. In 2022, Celsius and BlockFi froze withdrawals not because their code failed, but because their redeemable liabilities exceeded the liquidity of their asset books. In 2023, First Republic Bank collapsed in forty-eight hours for the same reason that would topple a stablecoin if its custody bank marked its reserve assets to market in real time: the maturity structure of the asset side did not match the redemption availability of the liability side. The lesson is identical at every scale. A calm balance sheet is not a liquid balance sheet.
There is a second structural issue. When the crypto market generates a meaningful share of its aggregate yield from Treasuries, the industry's correlation with the duration curve rises. In a zero-interest-rate environment, stablecoin reserve yields disappear and tokenized funds have nothing to distribute. In a high-rate environment, the concentration of assets migrates to the largest issuers, who operate the reserve book at the lowest cost. Small players are squeezed twice: when rates are high, they cannot achieve the scale to compete; when rates fall, they lack the margins to survive.
The phrase 'shadow banking' is not a pejorative here. It is a technical classification. Traditional shadow banking includes money market funds, securities lending, and repo transactions that carry out bank-like maturity transformation without a bank charter. Every one of those products experienced a run in March 2020 when the Treasury market broke and the Federal Reserve had to intervene in a market it had previously not entered. The crypto industry has adopted the product structure of shadow banking while removing the backstop infrastructure that the traditional financial system eventually built through painful experience.
The statutory responses have been predictable and partial. The European Union's Markets in Crypto-Assets Regulation, now in application, requires stablecoin issuers to hold reserves, submit regular reports, and obtain electronic money licensing. The GENIUS Act in the United States has advanced with a framework for payment stablecoins that mandates one-to-one reserves and bankruptcy-remote isolation. Both frameworks treat the stablecoin issuer as a type of depository institution. Neither creates a public backstop for redemption during systemic stress. Regulators are writing safety rules for a vehicle that lacks the one mechanism that made traditional deposits crisis-resistant: the central bank.
A stablecoin issuer under a perfect regulatory framework is still an institution whose liabilities are instant claims and whose assets are marketable securities subject to price dislocations. The regulation reduces the probability of a fraudulent reserve report. It does not eliminate the probability of a liquidity crisis. The next major stablecoin event will not be a failure of disclosure. It will be a failure of market depth. No attestation mechanism solves that by itself. Regulation can mitigate it only by limiting leverage, not by guaranteeing par value.
The institutions that survive the next cycle will treat the audit trail as a live operational system, not a quarterly ritual. The verification stack needs to move from signed PDFs to continuously updated, tamper-evident proofs of reserve. The token standards need to converge on compliance in code rather than compliance in the back office. The custody chains need to diversify. Any single point of failure in the settlement layer becomes a single point of failure for the entire market if the issuance base is large enough.
I have spent enough years inside this industry to understand that the public ledger is a witness, not a regulator. It records the transaction whether the underlying asset is properly reserved or not. The transaction itself is not a margin of safety. Code is law only if the audit trail is unbroken. The next systemic event will not begin with a failed smart contract. It will begin with a delayed redemption. The instructions are simple: monitor the month-end attestations, the custody concentration, the redemption-to-issuance ratio, and the time between a redemption request and its settlement. The ledger will show you the event after it has happened. The first line of defense is the institution that watches the liquidity, not the narrative.
Traditional banks earn their charter by accepting the burden of supervision in exchange for confidence. The crypto equivalent has not yet established which institution receives the burden or how it will be enforced. In the interim, the price of confidence is paid by the users who expose themselves to the spread between accounting truth and liquidity reality.