Here’s a hypothetical situation: a hedge fund is making money, but one of its exchanges is about to liquidate its position anyway. Bitcoin has fallen, its short position on CME is profitable, and the matching long on Hyperliquid is bleeding cash. The two trades were designed to offset each other, but Hyperliquid can’t use profits sitting at CME to cover the losses on its own books. The fund has to find more collateral before the exchange closes the position for it.
Moving money between exchanges takes time, and during a downturn, withdrawals can slow down or stop altogether. The fund could have enough money to cover every position and still lose half its hedge because the profits are sitting in different accounts.
Once that happens, a strategy designed to avoid betting on Bitcoin’s direction can suddenly become a very large bet on where the price goes next.
In the high-stakes world of institutional Bitcoin trading, a fund can be profitable across its entire portfolio and still face forced liquidation because the exchange holding its losing position doesn’t know or care about the money it has made somewhere else.
And the more efficiently the fund uses its capital, the less money it may have sitting around to solve the problem.
A 20% Bitcoin crash can break a perfectly good hedge
Here’s another hypothetical situation: a fund holding two opposing Bitcoin positions. It’s long Bitcoin on Hyperliquid and short Bitcoin futures on CME, with both positions worth $4.5 million.
If Bitcoin falls 20%, the short position earns roughly $900,000 while the long loses approximately the same amount, assuming both contracts track the price equally. On paper, the fund hasn’t lost much from Bitcoin’s directional move. Its short has offset its long, which was the entire point of the trade.
But unfortunately, the exchanges don’t see it that way.
Hyperliquid sees a losing position and demands enough collateral to keep it open. CME sees a profitable short position, but those profits are in a different account, subject to different margin and settlement arrangements. The fund needs to transfer some of those profits or close both positions before Hyperliquid decides to liquidate the losing one. If withdrawals are delayed, transfers are frozen, or the profitable trade can’t be closed quickly enough, the fund can find itself short of money in one account despite having enough assets across the portfolio.
Once Hyperliquid liquidates the long, the fund is left holding a short position that no longer has an offsetting trade. Now it loses money if Bitcoin rebounds, having gone to considerable trouble to avoid betting on Bitcoin’s direction in the first place.
Ian Weisberger, CEO of trading technology provider CoinRoutes, pointed to the disorderly exchange liquidations during the October 2025 crypto crash as an example of how dangerous this can become. Traders who thought their portfolios were balanced could suddenly be left exposed because an individual exchange closed one position without accounting for the other.
The problem isn’t necessarily that the fund made a bad bet; it’s that the money needed to keep the bet alive was sitting somewhere the exchange couldn’t reach.
One million dollars goes a (surprisingly) long way
The problem becomes more complicated when funds use borrowing and derivatives to stretch relatively small amounts of capital into much larger positions. Weisberger explained to CryptoSlate how a hedge fund depositing $1 million in USDC could, in theory, end up controlling $9 million worth of Bitcoin positions.
The fund starts with $1 million of its own capital and borrows another $2 million from a lender, giving it $3 million to work with. It allocates $1.5 million to CME and $1.5 million to Hyperliquid, then uses derivatives to establish a $4.5 million position on each exchange. It can buy $4.5 million worth of Bitcoin exposure on Hyperliquid while selling $4.5 million through CME futures. That’s $9 million in total positions, financed with $1 million of the fund’s own money, $2 million borrowed from a lender, and additional leverage through derivatives.
The fund isn’t necessarily betting that Bitcoin will go up or down. If Bitcoin goes up 10%, the long makes roughly $450,000 while the short loses about the same amount, assuming both contracts track the price equally. Instead, the fund wants to collect the difference between futures prices, perpetual funding payments, or other small discrepancies, with its opposing positions keeping most of the directional exposure out of the trade.
The problem is that the hedge still has to work in practice.
Any one of a hundred different things could go wrong: futures and perps can move apart, funding payments can become expensive, and even a 1% discrepancy between two $4.5 million positions amounts to a $45,000 difference. Even if the prices eventually converge, the fund needs enough collateral to survive whatever happens in between. And although the positions are supposed to offset each other, the exchanges still make their own margin decisions.
CME won’t waive a collateral requirement because the fund has a profitable position on Hyperliquid, and Hyperliquid won’t automatically credit profits that haven’t been transferred from CME. Keeping large deposits at both exchanges would certainly help, but that can get expensive pretty fast when the entire business depends on making small amounts of money from differences between markets.
The alternative is to make the same capital work harder, which introduces another problem: the more exposure a fund can support with every dollar, the more dependent it becomes on being able to access that dollar when something goes wrong.
The exchange doesn’t care that your other trade is profitable
Traditional prime brokers have spent decades helping hedge funds manage financing, collateral, and trading across different markets, but crypto markets have always been much more fragmented.
Funds trading Bitcoin futures at CME, perpetual contracts at Hyperliquid, and spot Bitcoin on another exchange need to maintain separate pools of collateral even when all of those positions are essentially part of the same strategy. This is because every exchange has its own margin requirements and settlement processes.
CRX Trade, a Swiss institutional prime brokerage built on CoinRoutes technology, is now trying to coordinate those arrangements. It allows professional traders to manage Bitcoin, stablecoins, and tokenized assets as collateral across crypto exchanges and traditional markets, including Hyperliquid and CME. Instead of funding each exchange separately and hoping money can move quickly enough when something goes wrong, funds can manage their positions and financing through one account.
Weisberger said the system considers both the total size of a fund’s positions and how much directional risk remains when they’re assessed together. That’s also why a lender might agree to finance a fund controlling nine times its original capital in trading exposure. The client has borrowed $2 million rather than $9 million, and the long and short positions are supposed to offset each other.
CRX’s risk engine monitors positions across the portfolio and can begin reducing exposure before an individual exchange forces a liquidation. Under one approach, called delta-neutral liquidation, it attempts to close both sides of a hedge together. If a fund is short a Tesla perpetual and long an equivalent amount of tokenized Tesla shares, the system can unwind both positions as a pair instead of leaving the client with an unwanted bet on Tesla. Another method reduces whichever position contributes the most directional risk.
Both approaches are designed to avoid the situation where an exchange closes the losing half of a trade and leaves the fund exposed to a market move it was trying to hedge.
But there’s a limit to what coordinated risk management can accomplish. Software can’t force an exchange to process an order during an outage, and it can’t guarantee there will be someone willing to take the other side at a reasonable price. That’s why the fund can still lose money closing its positions, especially when markets are moving quickly and buyers disappear. And the exchanges still retain the right to liquidate positions that fail to meet their margin requirements.
CRX can recognize that two positions were meant to work together and try to keep them from being separated.
Bitcoin can finance trades even where it isn’t accepted as collateral
The same approach can also allow funds to use Bitcoin holdings to support trades in markets where Bitcoin itself isn’t accepted as collateral.
Weisberger explained this using an example of a client holding $1 million in Bitcoin that wants to trade CME futures. The client transfers the Bitcoin to a crypto exchange, sells $500,000 worth, and replaces that portion of its holdings with $500,000 in Bitcoin futures or perpetuals. The fund now owns $500,000 in Bitcoin and has another $500,000 in derivative exposure, so its sensitivity to Bitcoin’s price is approximately the same. The spot sale has freed up $500,000 in cash, which can move through CRX’s USDC infrastructure to support trading at CME.
The money isn’t being used twice here. Only half the original Bitcoin has been sold, and the fund has bought a contract to replace the exposure it gave up. That contract has its own margin requirements and financing costs, and the position can be liquidated if the fund can’t keep enough collateral behind it.
Weisberger estimated that borrowing cash directly against Bitcoin would typically cost around 8%, while replacing some spot exposure with derivatives means paying the relevant futures basis or perpetual funding rate instead. That could be cheaper, although funding payments can fluctuate, and the fund still has to account for fees and spreads.
The company didn’t provide a full comparison of actual costs under both arrangements. In either case, the fund found a way to put more of its existing capital to work. It also added another position that needs financing, margin, and someone willing to keep the trade open when markets become disorderly.
Keeping collateral away from exchanges doesn’t eliminate the risk
One way to reduce exposure to an exchange failure is to avoid keeping all the collateral at the exchange in the first place. CRX uses tri-party settlement where available, keeping collateral with a separate custodian instead of depositing it directly at the trading venue. The exchange processes the trades, but the assets stay with the custodian, and profit and loss is settled periodically.
Weisberger said those settlements can occur every eight or 24 hours. This can limit the amount directly exposed to an exchange withdrawal freeze to the unsettled profit and loss rather than the client’s entire collateral deposit. But the extent of that protection depends on the agreements and settlement arrangements, and it doesn’t prevent an exchange outage from interfering with trades that need to be closed.
It also introduces another institution whose obligations matter when something goes wrong. CRX Trade is operated by RAS Capital, a Swiss financial intermediary affiliated with VQF, a regulator-recognized self-regulatory organization. It isn’t a bank or securities firm and doesn’t provide loans itself. Financing comes from independent lenders using the platform. Weisberger said clients retain legal ownership of assets held in dedicated, segregated wallets and exchange subaccounts.
However, if a client borrows money, the lender receives a lien over the portfolio collateral under a separate agreement. The Bitcoin still belongs to the client, but the lender has a legally enforceable claim against the pledged collateral if the client fails to meet its obligations. The agreement determines how much the fund can borrow, how the assets are valued, and when the lender can exercise its rights.
Meanwhile, the exchanges have their own margin requirements and contracts with the trader, while the custodian operates under another agreement governing where assets are held and who can access them.
Bringing everything into one account doesn’t eliminate any of those relationships, just makes them easier to coordinate.
CRX didn’t provide the custody and lending agreements needed to establish exactly what would happen if the platform, a custodian, or one of its lending partners became insolvent. Weisberger said clients retained ownership through segregated wallets, but recovering assets in an insolvency would depend on the contracts and laws governing each relationship.
Another important question is whether collateral can be pledged onward, something the company’s responses didn’t establish. So while keeping collateral away from an exchange can reduce one type of risk, it doesn’t necessarily mean the assets will be immediately available when another institution demands payment.
Everyone still wants their money back
There’s another problem with building large positions on borrowed capital, which is that, eventually, the lender will want its money back.
Weisberger said loans arranged through CRX usually run for 30 to 90 days, with leverage limits, collateral weights, and loan-to-value requirements agreed when the client borrows. The lender can decline to renew the loan when it matures, leaving the fund to repay the money or find someone else willing to finance its positions. That can happen regardless of whether the fund’s trading strategy is profitable.
Exchange margin requirements and derivative funding costs can also move during the loan term, regardless of what the lender originally agreed to. So funds can then face demands for additional collateral from an exchange while also needing to repay or refinance money borrowed against the same portfolio. Shared collateral can make that portfolio more capital-efficient, but it can’t override the lender’s contract or an exchange’s rules. And during a market disruption, the fund may need cash at several exchanges at once, precisely when transfers become harder and closing positions gets more expensive.
That’s the trade-off behind making institutional Bitcoin trading more efficient. There’s no reason for a fund to keep unnecessarily large amounts of capital scattered across exchanges if it can coordinate its positions and collateral more effectively. Doing so can reduce unnecessary liquidations, free up capital, and make hedged strategies cheaper to operate.
However, it also allows funds to support larger positions without committing more of their own money. And the larger those positions become, the more important it is that lenders, exchanges, and custodians all do what they’re supposed to do at the same time.
The better a fund gets at putting every dollar to work, the less money it has sitting around for emergencies. Shared collateral can reduce the risk of a profitable hedge being liquidated because its money is trapped in the wrong account. It can’t eliminate the underlying dependence on financing, liquidity, and exchange access.
The real measure of that efficiency won’t be how much exposure $1 million can support when markets are calm, but how much of it the fund can safely keep open when everyone wants their money back.