What Was the Prime Trust Bankruptcy?
Brian Cubellis | Chief Strategy Officer
The Prime Trust bankruptcy was the collapse of a US crypto custody and settlement company that held bitcoin and cash on behalf of other businesses, and then could not return it. Operating under Prime Core Technologies, the firm lost access to a set of its own wallets in December 2021, quietly used customer funds to cover withdrawals for roughly eighteen months, and by June 2023 owed customers far more than it held. It entered receivership in Nevada and then Chapter 11 bankruptcy, and in July 2025 a court ruled that the commingled client assets it had held were property of the bankruptcy estate rather than the customers. That ruling is the reason the estate is now suing firms that withdrew bitcoin before the collapse, and it is the case at the center of Onramp's research report The Architecture of Failure.
This guide explains what Prime Trust was, how the failure unfolded, the numbers regulators found, the contract clause that undid its customers, the ruling that changed everything, and why one custodian's collapse reaches far beyond its direct customers.
What Prime Trust was: infrastructure, not a consumer brand
Prime Trust was not a company most bitcoin holders interacted with directly, and that is the first thing to understand about it. It was infrastructure. Operating under Prime Core Technologies, it provided custody, settlement, and banking access to other crypto businesses, which in turn served the end customer.
That structure is the detail that matters, because it means Prime's failure was never going to stay contained to Prime. It sat one layer beneath firms that millions of people did business with. Several of those firms now appear as defendants in the bankruptcy estate's recovery actions, including Swan, Strike, Compass Mining, Fold, and Galaxy Digital.
The point is structural, not a question of size. When the layer other firms are built on fails, the consequences do not stop at that layer. They surface wherever someone was relying on it, often without knowing the name of the company actually holding the asset.
How the collapse happened
The failure did not begin with a hack or a market crash. It began with a lost set of keys and a decision to hide it.
In December 2021, Prime lost access to a set of legacy wallets. Rather than disclose the problem, management began using customer funds from omnibus accounts to purchase digital assets to satisfy withdrawals from the wallets it could no longer open. The Nevada regulator's petition describes the assets purchased in the plural, as digital currency rather than bitcoin alone, which means the practice was not limited to a single asset.
This continued for roughly eighteen months before regulators intervened. For a year and a half, in other words, customers were being paid with money that was not theirs, out of a shared pool, to paper over a hole the firm never disclosed.
The scale of that hole is best seen in the court record. An exhibit in the proceedings plots the difference between Prime's on-chain bitcoin balance and its internal ledger. For most of the company's existence the difference was negative, meaning the ledger recorded more bitcoin than the wallets actually held. At its worst, the shortfall reached approximately 1,400 bitcoin. In plain terms, the bitcoin that customers believed was held for them was not all there, and had not been for most of the firm's life.
What the regulators found: a balance sheet that did not exist
By the time the Nevada Financial Institutions Division petitioned for receivership in June 2023, the gap had become impossible to hide.
Prime owed customers roughly $85.7 million in fiat against approximately $2.9 million on hand, a hole of about $83 million. It owed a comparable amount in crypto against a near-matching balance. Separately, around $45 million of customer assets sat stranded in wallets it could not access.
The report's framing of this is worth repeating precisely: this was not a liquidity squeeze. It was a balance sheet that did not exist. A liquidity squeeze is a solvent business that cannot access cash quickly enough. The assets that were supposed to back customer claims were not merely hard to reach; for the most part, they were gone or unreachable, and had been misrepresented for months.
The contradiction in the fine print
If the operational story explains how Prime ran out of assets, the legal story explains why its customers, not Prime's owners, absorbed the loss. It comes down to a single self-contradicting clause.
Clause 1.8 of the custodial agreement stated that the account holder owns all Custodial Property, and that such property would not appear on Prime's balance sheet. The same clause then permitted Prime to take and hold title to Custodial Property in its own name. A separate end-user provision, quoted in the court's own opinion, went further, allowing Prime to use customer fiat at its own risk and to purchase other assets it could hold and register in its own name.
Two clauses, one page, in direct conflict. It said the account holder owns the property. It also let Prime take title to that same property in its own name. Both cannot be true once the custodian fails.
One detail here overturns a comfortable assumption many bitcoin holders carry: bitcoin's traceability did not save Prime's customers. One objecting customer argued that the unspent-transaction-output model made its holdings traceable on-chain and therefore identifiable. The court rejected the argument, finding the assets hopelessly commingled as a legal matter, such that even theoretically traceable bitcoin could not be attributed to individual customers. The chain may know whose coins are whose. The bankruptcy court decided it did not matter. Protocol-level transparency does not survive contact with custody-level commingling and a court's reading of it.
The July 2025 ruling that changed everything
The decisive moment in the Prime Trust bankruptcy was not the collapse itself. It was a court ruling two years later.
On 18 July 2025, Judge J. Kate Stickles issued the Prime Core distribution opinion in the Delaware Bankruptcy Court. The court held that the commingled cryptocurrency and fiat held by Prime at the time of filing were property of the bankruptcy estate, and could be distributed to creditors in dollars. It found that the customer agreements had not created a trust relationship, and that the extent of commingling made the assets untraceable, so that even bitcoin which is traceable on-chain in principle could not be attributed to individual customers in practice.
The significance of this is hard to overstate. Prime's customers were not owners recovering their own property. They were unsecured creditors of a failed company, standing in line with everyone else the company owed, and the assets that remained were the estate's to distribute.
This ruling is the precedent that enables the current wave of clawback actions. If the assets that stayed behind were estate property, then the assets that left before the filing were estate property that walked out early, which is exactly what bankruptcy law is designed to recover. Notably, the estate did not begin pursuing counterparties in earnest until it had this ruling in hand. It waited for the ground to be settled, then moved.
The clawback wave: pursuing the firms that got out first
With the ruling secured, the estate turned to the transfers that had left Prime in the months before it filed. This is the preference mechanism: under the US Bankruptcy Code, an estate can recover certain payments a debtor made shortly before filing where the recipient received more than it would have in a liquidation. The standard is structural, not moral. It does not require the recipient to have done anything wrong or even to have known the firm was in trouble. Its purpose is the equal treatment of creditors, so those who exited early do not keep a full recovery while everyone else shares the remains.
The largest of these actions seeks roughly $970 million from a single firm. It alleges that an insider tip was delivered through a disappearing-message channel in the days before a regulator meeting, and it reconstructs a precise five-day sequence involving the movement of 10,080 BTC. But it is not a one-off. The same estate, before the same judge, applying the same theory, has filed comparable actions against Strike, Compass Mining, Fold, and Galaxy Digital. The common thread is asset segregation, commingling, and preferential transfer in the months before the filing.
The point is not the fate of any single defendant. It is that a court has built a repeatable mechanism for treating custodied client assets as estate property, and is applying it methodically. The precise legal machinery behind these suits is the subject of a companion article, Can bitcoin be clawed back in a bankruptcy?
Why Prime Trust matters beyond its own customers
It would be easy to file the Prime Trust bankruptcy as a problem for the customers of one failed custodian. That reading is too narrow. The same structure, one party holding the keys while others hold contractual claims, runs through almost every way bitcoin is held on behalf of someone else.
Prime's direct customers were only the first layer. Bitcoin treasury companies, which hold bitcoin on their balance sheets and whose shares trade as a proxy for that bitcoin, generally do not hold their own keys; they use custodians. An investor in such a company holds a claim on the company, which holds a claim on a custodian, which holds the bitcoin. That is three layers, and the investor sees only the top one. Spot bitcoin exchange-traded funds sit in a similar position, and the market is heavily concentrated, with a single custodian estimated to underpin more than 80 percent of US spot bitcoin ETF assets.
In each case the pattern is identical. Someone else holds the keys. The holder owns a claim. And the strength of that claim, as Prime Trust has now demonstrated, depends entirely on what the custodian did and how a court later reads it. If a treasury company's or an ETF's custodian were to experience what Prime experienced, those holders would learn of it through a falling share price rather than any direct line to the asset. Whether a custodian can take title to your bitcoin, and who actually custodies the bitcoin behind ETFs, are therefore questions every holder is entitled to ask.
The structural lesson
Prime Trust failed for one root reason: a custody architecture that let a single entity hold the private keys, take title to client assets in its own name, and commingle those assets until they could no longer be traced. Every step of the collapse, from the hidden lost wallets to the estate ruling, flowed from that single point of failure.
The durable answer is architectural, because the problem is architectural. The report calls it multi-institution custody: a two-of-three arrangement in which three independent, regulated institutions each hold one key, and at least two must cooperate to move the asset. No single custodian can move it, because one key cannot sign a transaction. No single custodian can lose it, because the loss of one key still leaves a working two-of-three. No single custodian can take title to it or commingle it, because none of them ever holds it alone and each client's holdings sit in a segregated vault rather than an omnibus pool. This is the structure the second half of The Architecture of Failure develops in full, and it is the specific pattern Prime Trust lacked at every level.
The lesson is simple, and the industry has learned and forgotten it repeatedly since Mt. Gox. An account agreement describing ownership is not ownership. A regulated charter is not segregation. A qualified custodian is not necessarily one that holds your assets apart from its own. The protections that matter are the ones built into the architecture, not the ones written into the contract, because the contract is exactly what gets reinterpreted when the custodian fails.
If you are evaluating how your bitcoin is held and whether it could be exposed to a custodian's insolvency the way Prime Trust's customers were, schedule a consultation with Onramp to discuss Multi-Institution Custody and how removing the single point of failure keeps your bitcoin out of any one institution's bankruptcy estate. To open an account, sign up here.
Related Reading