Eighteen months after Silvergate Bank told the world it was winding itself down, its former chief executive finally said the quiet part out loud. The bank, Alan Lane suggested in public remarks, was not killed by bad credit or fraudulent accounting. It was killed by its regulators — squeezed, slowly and deliberately, until voluntary liquidation became the only exit that preserved any dignity at all.
Sit with how strange that sentence is. A publicly traded bank, one that filed audited financials with the SEC, that told shareholders in its own annual report that it had fallen short of its capital requirements and carried material weaknesses in internal controls, is now the subject of a claim that its real problem was political. Both things can be true. In banking, both things usually are. And that 'both' — not the accusation — is where the actual story lives. Not in the verdict, but in the mechanism.
Full disclosure before we go further: I have been skeptical of the Operation Choke Point 2.0 frame since early 2023, largely because it is built to be unfalsifiable. Any failure can be attributed to invisible pressure; any piece of counter-evidence can be dismissed as a regulator's talking point. That is a bad epistemic habit, and I try not to have it. But I have also sat in closed rooms in Abu Dhabi where regulators and founders negotiate precisely which banks will touch which dollars, and I have watched a single sentence in that room end a business model without ever appearing in a rulebook. Decoding the noise to find the signal means holding both truths at once.
To understand why Silvergate's epitaph matters, you have to understand what its operating system actually was. This was never a conventional lender. Founded in 1988 as a small La Jolla thrift, it pivoted hard into crypto around 2013 and rebuilt itself around one insight: the most valuable primitive in this industry was never the token. It was the settlement rail.
Its crown jewel was the Silvergate Exchange Network — a permissioned, always-on layer that let two approved counterparties move dollars against each other at two in the morning on a Sunday. No wire cut-offs, no correspondent bank holidays, no three-day settlement windows. For a market that never closes, SEN was the closest thing to a native dollar rail that crypto has ever had. Tracing the sharding roots of tomorrow's liquidity is something of a habit for me, and SEN now reads in retrospect as the first serious attempt to shard fiat settlement away from the batch-processing substrate the rest of finance still runs on.
I spent a good chunk of 2021 and 2022 doing something unglamorous: cataloguing every public reference to SEN counterparties I could find — exchange filings, conference panels, job postings, developer documentation — and building a map of who actually depended on that rail. The map was uncomfortable to look at. By the third quarter of 2022, roughly $11.9 billion in deposits sat with the bank, and an enormous share of the institutional crypto economy's dollar plumbing ran through one California charter.
Then November 2022 happened. Deposits tied to a single exchange group — just under a tenth of the deposit base — became radioactive overnight. By the end of the fourth quarter, total deposits had collapsed to $3.8 billion. A 68 percent decline in ninety days. To fund the exits, the bank sold available-for-sale securities into a rising-rate market and booked roughly $718 million in losses. The quarter's net loss to common shareholders approached $1 billion.
On March 8, 2023, Silvergate announced it would voluntarily liquidate. Four days later, Signature Bank was closed by its regulator. The Fed, the FDIC and California's DFPI issued a joint statement insisting that crypto exposure was not the cause. Very few people in this industry believed them, and I was not among them.
Here the narrative bifurcates, and this is where I want to slow down rather than speed up.
Lane's claim, stripped of rhetoric, is that Silvergate was solvent, that its business was fundamentally sound, and that supervisory pressure — risk-weighted asset treatment, examiner escalation, correspondent withdrawal, informal guidance — forced the bank into a death it did not deserve. If that is accurate, it is the cleanest piece of first-hand evidence yet assembled for the de-banking thesis.
The problem is the paper trail. Silvergate's own annual report acknowledged that the bank would fall below its required capital ratios and was no longer well capitalized. It disclosed material weaknesses in internal controls over financial reporting. It disclosed subpoenas and government investigations. A bank that writes those sentences into its own SEC filing is not an institution describing harassment. It is an institution describing an unravelling that runs along several axes at once.
So which is it? I think the question is malformed, and that is the first genuinely new thing I would ask a reader to take from this story. Regulatory pressure and structural fragility are not competing explanations. They are sequential stages of the same event.
Look at the mechanics. Almost nothing in American bank supervision happens through one dramatic order. It happens through a gradient. A supervisory letter raises the risk weight on a category of assets. The higher risk weight makes a line of business capital-inefficient. That inefficiency forces a conversation about concentration limits. The concentration limit forces the bank to slow onboarding. Meanwhile the correspondent relationships — the actual banks that let it clear dollars — begin to drift, because no compliance officer wants their name on a consent order. None of that is illegal. None of it is written down as a prohibition. All of it is lethal.
Here is the asymmetry that Lane's account skips past: the fragility came first, and the pressure chose the timing. A bank whose deposit base is a single industry, whose depositors are reflexive traders, and whose liabilities can flee forty-eight hours after a counterparty failure is not a stable institution being mistreated. It is an institution whose stability was always borrowed. When FTX imploded, roughly $8.1 billion walked out in a single quarter. No balance sheet survives that. Noticing this is not an insult to the bank.
This is where the narrative economics get interesting. Where capital flows, stories of value emerge — and where capital flees, stories of villainy emerge with equal reliability. The de-banking frame is attractive not because it is necessarily false, but because it is emotionally complete. It has a victim, a perpetrator, a motive and a moral. It converts a messy structural failure into a clean political crime. The architecture of belief built on code wants villains the way it wants whitepapers.
Now the honest counterweight, because there is one and I refuse to pretend otherwise. In 2024, following Coinbase's FOIA litigation, the FDIC released a batch of previously unseen 'pause letters' — correspondence in which the agency asked banks to hold off on expanding crypto-related activities pending further review. There were twenty-three of them. That is not a conspiracy theory. That is a filing cabinet. Custodia Bank spent years fighting for a master account and lost, twice. Signature's own board members said publicly that crypto was the reason for its closure, even as regulators insisted it was not. Listening to the digital tribe's hidden rhythm means noticing that the tribe is not always wrong.
So the accurate position — the one almost nobody seems willing to hold — is this: the soft choke is real, the machinery exists, and Silvergate was nonetheless a genuinely fragile bank whose fragility regulators did not create. Two true sentences that the industry insists on treating as an either/or.
There is a sharper way to say it. Silvergate was not a healthy bank murdered by politics, and it was not a reckless bank that simply failed. It was a single point of failure that the system tolerated for a decade because it was profitable, and stopped tolerating once it became politically expensive. Regulation did not cause the wound. It declined to treat the patient.
And what died was not really the bank. It was the rail. When SEN went dark, the industry lost the only near-native dollar settlement layer it had ever built at the banking layer. Notice what replaced it: nothing structural. Settlement migrated to non-bank intermediaries, to offshore banking relationships, to stablecoin rails carrying their own jurisdictional exposure. Liquidity is not just numbers, it is narrative — but it is also plumbing, and when plumbing disappears people forget it was ever there.
The bear-market lens matters more here than the political one, and this is the part I would underline for anyone reading now. We are watching the same structure reproduce itself at smaller scale. Protocols that looked adequately collateralized at higher prices are bleeding liquidity providers as marks fall, and the depositors — or LPs, or stakers, or whatever we are calling them this cycle — behave exactly the way Silvergate's depositors behaved. They leave together, and they leave fast. When people ask me which platforms are safe right now, I have started giving an answer they do not enjoy: safety is not a balance-sheet property, it is a liability-structure property. Show me where the money can run to and how quickly, and I will tell you whether the thing is survivable. I learned that lesson the hard way in 2020, when I tracked fifty Uniswap V2 liquidity providers and found four in five underwater while chasing yield. Concentration in a counterparty is not a footnote. It is the whole story.
One last layer, because it is my job to audit social capital and not just balance sheets. Chasing the archetype behind the avatar's mask — individually and institutionally — means asking who benefits from a story being true, and who benefits from it remaining unresolved. The de-banking narrative is worth real money to real actors in a US election cycle where crypto-aligned political committees deployed well north of $200 million. A grievance that produces legislative leverage is not the same thing as a grievance that produces an investigation. The Lane claim is far more valuable as a permanent argument than as a settled case — and that asymmetry, not the 10-K, is what should make you suspicious of how loudly it is being amplified.
The contrarian move here is not to defend the regulators. It is to point out that the entire argument is a distraction from a structural question nobody wants to ask.
Everyone is litigating whether the choke was real. Almost nobody is litigating why this industry routed its dollar settlement — its lifeblood — through exactly one charter for the better part of a decade. The de-banking narrative, even if it is entirely true, describes a system with a single bridge across a river. You can spend your energy being furious about who burned it, or you can build a second one. This industry has chosen fury for eighteen months and counting.
There is a strategic implication buried under the outrage. If the pressure was real and coordinated, then the correct conclusion is not 'regulate better.' It is 'become jurisdictionally shardable.' The answer to a choke point is sharding — distributing the function across charters, jurisdictions and instrument types until no single gatekeeper can shut the whole thing down. That is why tokenized treasuries, licensed stablecoin issuance and non-bank payment channels are where the next infrastructure cycle is actually being decided, while the industry keeps arguing about data availability layers that most rollups will never generate enough data to need. Capital is already routing around the choke. The commentary has not caught up.
The second blind spot is subtler and more damaging. This industry treats the 10-K as scripture when it supports a narrative and as regulatory propaganda when it does not. Silvergate told the market in writing that it was not well capitalized. That disclosure does not disappear because the CEO has a better story eighteen months later. If we want the de-banking thesis to have real force — the kind of force that survives a congressional hearing rather than a podcast — we have to be willing to read our own documents honestly.
So watch the record, not the recrimination. If a second crypto bank executive corroborates Lane under oath, or if further FOIA releases surface a supervisory letter with Silvergate's name on it, the frame shifts from grievance to evidence and the legislative consequences become real. If they do not, this settles into folklore — emotionally satisfying, analytically inert.
The question worth carrying forward is not whether the choke was real. It was real enough. The question is whether anyone in this industry is willing to build the rail that makes the question irrelevant.