The standard account of the 2022 collapses is that a set of companies were run badly. They took on too much risk, made reckless bets, and got caught when the market turned. Replace the bad operators with good ones, the story goes, and the model works.
That account is incomplete. The operators were a problem. But underneath the operational failures was a conceptual one — a mistake about what Bitcoin was that the entire first generation made in common, and that no amount of operational competence would have corrected.
The first generation of Bitcoin firms failed because they treated Bitcoin as a tradeable asset. It is better understood as collateral. Almost everything that went wrong follows from that single misframe.
Two different questions
When you treat Bitcoin as a tradeable asset, the question you are asking is: which direction will the price go, and how do I capture that movement? The asset is something to hold, trade, lend out, leverage, and deploy in pursuit of return. Its value to the business is the return it can generate. Volatility is an opportunity to be harvested.
When you treat Bitcoin as collateral, the question is entirely different: will this asset be there, liquid and enforceable, at the moment a borrower cannot pay? The asset’s job is not to generate return. Its job is to protect principal under stress. Volatility is not an opportunity — it’s the condition the structure has to be built to survive.
These are not two strategies for the same business. They are two different businesses that happened to hold the same asset. The first generation built the first business while describing it as the second. They called themselves Bitcoin-backed lenders, but they operated as asset managers chasing yield, with Bitcoin as the inventory.
What the misframe produces
Once you’ve decided Bitcoin is a tradeable asset, every subsequent decision points the wrong way — and each one looks correct from inside the frame.
If the asset’s purpose is to generate return, then idle collateral is wasted inventory. So you lend it out, deploy it, put it to work. Inside the trading frame, leaving collateral segregated and untouched looks like leaving money on the table. Inside the collateral frame, it is the entire point — collateral that has been lent out is no longer there to protect anyone.
If the asset’s purpose is to generate return, then a higher yield is simply a better product. So you raise the promised rate to attract deposits, and you raise the risk required to pay it. Inside the trading frame, the yield is the offering. Inside the collateral frame, a yield above what borrower interest can honestly produce is a warning that the structure is being compromised to fund it.
If the asset’s purpose is to generate return, then custody is an operational detail — a place to keep the inventory between trades. So you bundle it with the lending business, or push it out to whichever counterparty offers the best terms. Inside the trading frame, this is efficiency. Inside the collateral frame, it is the destruction of the one property that made the collateral worth holding: that the lender could actually reach it.
The pattern is consistent. Every defect that surfaced in 2022 — commingling, rehypothecation, maturity mismatch, counterparty dependency, yield extraction — is what you get when you optimize Bitcoin for return instead of structuring it for protection. The operators weren’t ignoring the collateral principles. They never adopted them, because their frame told them the principles were inefficiencies.
Why competence couldn’t have saved them
This is why “better operators” is the wrong fix. A more disciplined version of a trading-framed firm is still a trading-framed firm. It will still treat segregated collateral as idle inventory, still see yield as the product, still regard custody as logistics. It might survive longer by managing those risks more carefully. But it is managing the wrong risks, because it has misidentified what business it is in.
The collateral frame isn’t a more cautious version of the trading frame. It is a different starting question, and the starting question determines everything downstream. A firm that begins by asking “will the collateral be there under stress?” builds segregation, conservative leverage, automatic liquidation, and reachable custody as foundations — not because it is more prudent, but because those are the requirements of the question it is actually answering. A firm that begins by asking “how do I generate return from this asset?” builds none of them by default, and bolts on risk management afterward as a constraint on the real business.
The failures of 2022 were not a verdict on Bitcoin. Bitcoin did exactly what it does — it stayed liquid, continuously priced, and verifiable throughout the stress. The asset performed. The firms failed because they had built businesses that depended on Bitcoin going up, while telling depositors they had built businesses that depended on Bitcoin being there. When the price stopped cooperating, the trading business underneath the collateral language was exposed, and there was nothing protective left.
The category was correct; the framing was not
There is a version of this history that concludes Bitcoin-backed lending is inherently unsound — that the failures proved the model doesn’t work. That conclusion makes the same mistake the first generation did, just in the opposite direction. It treats the trading businesses that failed as if they were the collateral model being tested. They weren’t. The collateral model was never built. It was named and then abandoned in favor of something that paid better until it didn’t.
The first generation proved something narrower and more useful: that Bitcoin’s properties do not enforce themselves. The asset makes a superior collateral structure possible — continuously verifiable, continuously liquid, directly transferable, free of issuer dependency. It does not make that structure automatic. Whether those properties become a resilient system or marketing language wrapped around a trading book depends entirely on which question the operator started from.
That is the opening the first generation left behind. Not a discredited model, but an unbuilt one — a frame nobody in the first wave actually committed to. The firms that come next are not competing against the failures. They are building the thing the failures only claimed to be.
The asset was always collateral. The first generation just kept trying to trade it.