The market's favorite summary is a supply chart. Roughly $180 billion to $230 billion in stablecoin circulation. Two issuers controlling more than 80 percent of the float. Settlement rails that run 24/7 and cost a fraction of correspondent banking. Fine. Now run the counterfactual that actually matters: if one of those two dominant issuers lost its remaining U.S. banking relationships tomorrow, how far would the supply chart revert within ninety days?
That question is not theoretical. In the spring of 2023, Silvergate chose voluntary liquidation; New York regulators closed Signature. The roster of crypto-friendly banks collapsed to a handful of names, and the industry absorbed the shock, then moved on to the next price narrative. But the institutional disposition has shifted since then. Treasury teams, payment giants, and licensed issuers now demand stablecoin products that compliance officers can approve without a briefing dossier. The new consensus is clean and forceful: stablecoins will not scale through DeFi loops; they scale when banking infrastructure defines the issuance layer. That consensus is spreading through boardrooms and regulatory drafts. My job is to audit it. Is this consensus the whole truth, or only the most legible version of it?
Think of the market structure as an hourglass. The middle is crowded: a concentrated issuer group with massive balance sheets. The upper bulb, federally insured and correspondent-capable banks, is narrow. The lower bulb, payments, remittances, treasury automation, and DeFi, is expanding at a rate limited by how fast the upper and middle segments connect. Since 2020, the center has been overbuilt while both flanks stayed thin. Issuers grew enormous because digital distribution has no geographic constraint; the upstream banking layer, meanwhile, actively shrank. Remove Silvergate and Signature from the picture in 2023, and the result is structural: supply abundant, yet trusted entry points connecting that supply to bank-settled dollars dangerously few.
Regulators responded faster than the sector expected. In Europe, MiCA now requires issuers to hold an electronic-money license and to respect one-to-one reserve rules. In Washington, successive legislative drafts, with the Senate's GENIUS Act debates being the most visible, have converged on the same formula: if you issue dollar stablecoins at scale, you must be chartered, supervised, and audited. The intellectual shift is complete. The policy conversation no longer asks whether stablecoins are financial assets or software tokens. It asks who may issue them, with what charter, and through which supervised pipes. For most market participants, that question has one answer: the bank.
In my own audits of stablecoin issuers after the 2023 banking crisis, the recurring gap was never cryptographic. It was never in the smart contract. Every token is auditable until the audit trail reaches a bank statement; at that moment, the audit trail becomes a relationship, not a protocol. This is the real meaning of reliable plumbing. It is what analysts are trying to communicate, vaguely, when they declare that stablecoin scale requires regulated infrastructure. The unresolved issue is whether that regulated infrastructure must be a bank's balance sheet, or whether it can be installed as external supervised rails.
The first step in a genuinely rigorous analysis is to clarify what the bank thesis actually prices. Most observers focus on market share and transaction volume; they ignore a simpler financial fact: the stablecoin issuer's core earning asset is reserve yield. A dollar stablecoin is fundamentally a money-market position with a payment interface. The issuer invests reserve assets in short-term Treasury paper and captures overnight income on the entire circulating float. This explains the industry's succession of zero-fee strategies; fees were never the point. Yield is the lie; liquidity is the truth. Token liquidity on decentralized exchanges matters far less than the liquidity of banking relationships, custodial access, and regulatory approval that backs the float.
That reframing splits the current market into two distinct trajectories. One dominant issuer spent years purchasing the regulated route: banking partners, institutional custody, public attestations, a policy posture designed for integration with core banking. The other dominant issuer built its scale through distribution and an infrastructure that avoided deep integration with the U.S. banking core, relying instead on global treasury assets and offshore presence. Under a bank-led framework, both issuers still exist. Their strategic trajectories diverge on the marginal cost of regulatory acceptance. Auditing the code, not the charisma: the code of both tokens is trivial, but the structure around them is the real product. That discrepancy is what an analyst should be measuring.
The thesis, however, contains an unproven middle step. Almost everyone repeating 'no banks, no scale' collapses two distinct propositions. Proposition one: institutions will not deploy large balances into stablecoins without supervised, auditable infrastructure. That is observably true and has been confirmed by every stalled pilot and compliance review in the last two years. Proposition two: therefore, banks must issue or run stablecoin networks. That does not follow. Regulated non-banks have already issued stablecoins at meaningful scale, through U.S. state trust charters and through European electronic-money licensing structures. MiCA explicitly creates a route for a well-capitalized non-bank to be the licensed issuer; the GENIUS Act debates have likewise considered paths beyond the full bank-charter model. The market narrative is skipping an entire category.
The new insight hiding in plain sight is architectural: what the market needs is not necessarily banks-as-issuers, but a supervised settlement membrane between crypto rails and bank rails. A stablecoin is, at its core, a promise to pay fiat; the token is only the envelope. That membrane can be operated by a chartered non-bank, a clearing utility, or a bank. Its essential properties will be similar regardless of owner: a transparent legal identity, a supervisory window, and a credible accounting claim that token liabilities match reserve assets. Everything before the membrane is cryptographic verification; everything after is institutional trust. In previous market cycles, I found the best risk-adjusted edge by auditing token mechanisms before narratives matured. The same habit applies here: the code defines the token; the license defines the network.
The sharpest contrarian signal is the confusion between 'regulated' and 'bank-supervised.' MiCA's framework already proves licensed non-bank issuance is viable in a major jurisdiction. U.S. trust-chartered models have survived more than a decade of regulatory whiplash. For a product whose only binding constraint is trustworthy reserves, a licensed entity can be the solution; yet the narrative treats non-bank infrastructure as invisible until a major bank logo appears. Arbitrage exposes the cracks in consensus, and the purest arbitrage in the coming cycle may be the valuation gap between brand-name bank stablecoin pilots and licensed non-bank issuers with proven operational history.
The second blind spot is global. Emerging-market stablecoin adoption is not waiting for correspondent banks. Cross-border labor payments from the Gulf to South Asia, intra-African trade settlements, and Latin American currency hedges do not begin or end with a U.S. compliance officer. In those corridors, scale emerges from local distribution and currency stability, not from Basel or New York approval. The 'bank is the essential gateway' narrative is a Western institutional viewpoint. It mistakes the largest wallet for the entire market.
The third blind spot is the most uncomfortable one for the banking lobby: putting a bank in the middle does not eliminate counterparty risk; it relocates it. A bank-protected stablecoin inherits the bank's balance-sheet fragility, its deposit-run exposure, and its susceptibility to supervisory discretion. The crypto experiment was built partly as a response to that fragility. Replacing an offshore reserve with a fragile commercial bank layer is not automatically progress. If the next crisis arrives with a bank-managed stablecoin at its center, the sector will rediscover that regulated does not mean risk-free.
Here is the positioning playbook for a sideways market. Over the next twelve to twenty-four months, track the GENIUS Act's final text and whether non-bank licensing routes survive; track MiCA-authorized issuance volume; track the number of serious bank partnerships each issuer announces; and track whether bank-linked stablecoin projects expand total liquidity or simply reshuffle existing supply. Pivot not panic: the data reveals the path. The strongest signal will be the next banking shock, not the next exchange announcement. When it comes, ask only the question that matters: does the stablecoin have a bank that stays in the room? Narrative follows logic, never precedes it. The logic here says the value is in the structure, not the token. Position accordingly.