The short profits from CME can't save the long losses of Hyperliquid. Why has perfect hedging turned into a one-sided gamble?

Written by: Andjela Radmilac

Translated by: Saoirse, Foresight News

Imagine a scenario where a hedge fund is overall profitable on paper yet still faces forced liquidation by the exchange. Bitcoin prices drop, and while the fund profits from its short position on CME, it incurs massive losses on its long position on Hyperliquid. These two trades were supposed to hedge against risk, but Hyperliquid cannot withdraw the profits from the CME account to offset its position losses. The fund must supplement its collateral before the exchange starts forced liquidation.

Inter-platform transfers experience time delays; during market crashes, withdrawals may be further delayed or even blocked. Even if the total assets of the fund are sufficient to cover all positions, the earned funds being locked in other accounts could still lead to half of the hedged positions being lost.

Once this situation occurs, a hedge strategy originally designed to mitigate Bitcoin's unilateral upward and downward risks instantly transforms into a high-risk gamble on the direction of the prices.

In the high-risk institutional Bitcoin trading sector, even if the fund's overall portfolio is in profit, it still faces forced liquidation; because the exchange does not know about the losses incurred, nor will it consider the earnings made on other platforms.

Moreover, the higher the fund's capital utilization efficiency, the less idle funds it has available to resolve crises.

Bitcoin plummeting by 20% can wreck a perfect hedging strategy

Let’s look at another hypothetical case: the fund simultaneously holds two sets of opposite Bitcoin positions, going long on Bitcoin on Hyperliquid and shorting Bitcoin futures on CME, with the market value of both sides being $4.5 million.

If Bitcoin drops by 20%, assuming the underlying price of the contracts remains consistent, the shorts will profit by about $900,000, while the longs incur a commensurate loss. From an overall accounting perspective, price fluctuations in Bitcoin have nearly no impact on the fund's losses; the short profits hedge the long losses, which is the original intention of the strategy.

However, the exchanges only examine their own platform's positions separately and do not judge from the perspective of the overall portfolio.

When Hyperliquid detects losses on positions, it will require the fund to supplement sufficient collateral to maintain the positions. Although the shorts on CME have already profited, the earned funds are stored in an independent account, executing a different set of margin and settlement rules. The fund either needs to transfer the profits from elsewhere to supplement the margin, or it must close both sides before Hyperliquid conducts forced liquidation. If there are delays in withdrawals or restrictions on transfers, the profitable position might not be disposed of in time, and even if the overall assets are sufficient, a single account may still face liquidity issues.

Once Hyperliquid forcibly closes the long positions, the fund is left with only the short positions which no longer have hedging protection. If Bitcoin rebounds afterwards, the fund would incur losses, while the original intention of this strategy was to avoid directional volatility risk.

The CEO of trading technology service provider CoinRoutes, Ian Weisberger, validated the destructive power of this risk with examples from the chaotic liquidation of exchanges during the cryptocurrency market crash in October 2025. Many traders believed they had fully hedged, only for exchanges to unilaterally force liquidate one side of the positions, exposing the risks entirely.

The problem often lies not in misjudgments about trade directions; but in the margin required to maintain positions being held in accounts that are not accessible on the current exchange.

With $1 million in capital, astonishingly large trading positions can be leveraged

When funds use borrowing and derivatives to create large positions with a small amount of capital, the risks become further amplified. Weisberger explained to CryptoSlate: a hedge fund depositing $1 million in USDC can theoretically hold Bitcoin exposure worth $9 million.

With $1 million of the fund’s own capital, it borrows $2 million from lenders, resulting in a total of $3 million available. The fund allocates $1.5 million to both CME and Hyperliquid and then utilizes derivatives to establish $4.5 million positions on both platforms: the fund goes long $4.5 million worth of Bitcoin on Hyperliquid while simultaneously shorting $4.5 million through futures on CME. Thus, the total exposure reaches $9 million, funded by $1 million of its own capital + $2 million of external borrowing, plus leverage from the derivatives.

Funds like this are not betting on Bitcoin’s rise or fall. When Bitcoin rises by 10%, the long position profits approximately $450,000, with the losses on the short nearly equaling that (assuming contract prices match). The fund’s real return comes from futures basis, perpetual contract funding rates, and various small market price differences, relying on bilateral hedging to isolate most directional market risks.

However, in reality, hedging strategies do not always operate as intended.

The potential risks are myriad: futures and perpetual contracts may diverge in pricing, funding rates could surge suddenly; if the two $4.5 million positions incur a 1% price difference, the profit and loss gap could reach $45,000. Even if prices later converge, during this period, the fund must still hold enough collateral to weather the volatility. Even if positions are designed to offset one another, each exchange still independently executes its own margin requirements.

CME will not relax margin requirements because you are profitable on Hyperliquid; Hyperliquid will not automatically credit profits that have not yet been transferred from CME. While holding high collateral across all exchanges is certainly safer, the profit margins of spread strategies are quite slim, and a lot of capital tied up can quickly erode returns.

Another option is to optimize capital efficiency, but this brings new risks: the larger the exposure leveraged against unit capital, the greater the dependency on the timely availability of funds during a crisis.

The exchanges will not consider your profits elsewhere

For decades, traditional prime brokers have continuously handled cross-market financing, collateral, and trading for hedge funds. However, the cryptocurrency market is highly fragmented.

The fund trading Bitcoin futures on CME, trading perpetual contracts on Hyperliquid, and trading spot on other platforms, even if they all belong to the same trading strategy, must still prepare separate collateral pools for each exchange; the margin and settlement systems of each platform are completely independent.

Swiss institutional prime broker CRX Trade, leveraging the technology from CoinRoutes, is working to address this pain point. Professional traders can use Bitcoin, stablecoins, and tokenized assets as cross-platform collateral, covering cryptocurrency and traditional trading markets like Hyperliquid and CME. Funds do not need to deposit funds separately into each exchange; all positions and financing can be consolidated and managed in a single account.

Weisberger stated that the system will simultaneously calculate the fund's overall position size as well as the remaining unilateral risks in the portfolio. This is also why lenders are willing to provide financing to allow the fund to leverage a $9 million exposure; the client is actually only borrowing $2 million, not $9 million; long and short positions can theoretically offset one another.

CRX's risk engine monitors all portfolio positions, capable of actively reducing exposure before triggered forced liquidations occur at exchanges. One mode is called Delta-neutral liquidation, which attempts to pair close opposing positions together. For example, if a fund shorts a Tesla perpetual contract while simultaneously going long an equivalent amount of tokenized Tesla stock, the system will close both positions together, preventing the client from passively holding unwanted unilateral risks. Another mode directly reduces positions that cause the greatest unilateral risks.

The core goal of both operations is to prevent the exchange from solely closing the unprofitable side of positions, leaving the fund exposed to market fluctuations that it intended to avoid.

However, this collaborative risk management system has its limitations. The software cannot force orders in the event of an exchange outage, nor can it guarantee that there are always counter-parties available to transact at reasonable prices. During extreme market fluctuations and liquidity shortages, closing positions can still lead to losses. Moreover, exchanges still have the authority to liquidate under-margined positions.

CRX can identify pairs of hedged positions to avoid both being forcibly liquidated separately.

Even if the platform does not support Bitcoin as collateral, financing trades with Bitcoin is still possible

This model also supports funds using Bitcoin holdings to trade in markets that do not accept Bitcoin as collateral.

Weisberger gives an example: a client holds $1 million worth of Bitcoin and wants to trade CME futures. The client transfers Bitcoin to a cryptocurrency exchange and sells half, worth $500,000, then rebuilds a $500,000 Bitcoin risk exposure through futures or perpetual contracts. At this time, the client still holds $500,000 in spot Bitcoin along with $500,000 in derivatives exposure, maintaining basically the same sensitivity to Bitcoin price risk. Selling off the spot releases $500,000 in cash, and through CRX's USDC system, this cash can be allocated to CME as trading margin.

This does not involve the duplication of funds. It merely involves selling half of the spot Bitcoin and using contracts to offset the corresponding risk exposure. This derivative position has its own margin and financing costs; if the collateral is insufficient, it will also face liquidation.

Weisberger estimates that using Bitcoin as collateral directly for cash typically incurs financing costs of around 8%; while using derivatives to replace part of the spot exposure incurs only the futures basis or perpetual contract funding rate. Costs are expected to be lower, but funding rates will fluctuate, along with transaction fees and slippage costs.

The company has not provided a comprehensive actual cost comparison of the two options. Whichever option is chosen, the fund has activated existing capital but simultaneously added positions that require financing and margin; in times of market turmoil, it must still rely on counterparties to take up orders.

Transferring collateral out of exchange custody does not mean risk is completely eliminated

One way to mitigate the risk of exchange failure is not to keep all collateral stored within the exchange. When conditions allow, CRX adopts a third-party settlement model, whereby collateral is held by an independent custodial institution rather than being stored directly on the trading platform. The exchange is only responsible for matching orders, and the assets remain with the custodian while profits and losses are settled periodically.

Weisberger states that the settlement cycle is 8-24 hours. Thus, only the unsettled profit and loss portions will face the risk of withdrawal freezing by the exchange, not the client’s entire collateral. However, the protective effect depends on the protocols and settlement terms and cannot avoid issues related to exchange outages disrupting liquidations.

Simultaneously, this model introduces new participating institutions whose performance capabilities are crucial in times of crisis. CRX Trade, operated by RAS Capital, is a Swiss financial intermediary subject to the regulatory self-discipline organization VQF. It is neither a bank nor a brokerage and does not directly issue loans; financing comes from external lenders connected to the platform. Weisberger states that assets stored in isolated dedicated wallets and exchange sub-accounts legally belong to the clients.

However, once clients initiate borrowing, under an independent agreement, lenders will have a lien on the pool of collateral. The ownership of Bitcoin still belongs to the client, but if the client defaults, the lender has the legal right to liquidate the pledged collateral. The loan agreements will specify the borrowing limits, asset valuation methods, and the triggers for the lender to exercise liquidation rights.

Exchanges have their own margin rules and trading agreements, while custodial institutions have their own contracts for asset custody and access rights.

Consolidating all assets into a single account does not eliminate the various constraints of the above agreements; it merely facilitates more convenience among multiple parties.

CRX has not disclosed the complete texts of the custodial and lending agreements, therefore it cannot confirm the asset disposal details after the platform, custodian, or lending partner goes bankrupt. Weisberger claims that assets are stored in isolated wallets, but whether they can be smoothly recovered in the event of bankruptcy depends on contract clauses and applicable judicial jurisdiction laws.

Additionally, the company has not clearly stated whether collateral is permitted to be re-pledged. Moving assets out of the exchange does indeed evade a certain kind of risk, but this does not mean that your assets are accessible at any time when other institutions assert claims.

In a crisis moment when all parties want their funds back

Building large positions by relying on borrowed funds also harbors another hidden danger: lenders inevitably want to recoup the principal upon expiration.

Weisberger explains that the loans arranged by CRX usually have terms of 30-90 days, with leverage limits, collateral conversion rates, and pledge ratios agreed upon from the start of the lending. Upon loan expiration, lenders have the right to refuse to extend loans; regardless of whether the fund’s trading strategy is profitable, it must return the funds or seek other financing sources.

During the loan period, the margin standards of exchanges and the funding rates of derivatives may change, which are not bound by the initial lending terms. The fund may find itself in a dilemma: one one hand the exchange demands additional collateral, while on the other, it must repay or refinance the same batch of assets. Shared collateral increases capital efficiency but cannot override the contracts of lenders and the rules of exchanges. During turbulent market times, multiple exchanges may simultaneously pursue funds, while the costs of transfers and liquidations can soar.

This is the price that institutional Bitcoin trading pays for pursuing capital efficiency. If capital can be effectively coordinated with positions and collateral, there is no need to disperse large amounts of capital across various exchanges. This mechanism can reduce unnecessary forced liquidations, release usable capital, and lower the operational costs of hedging strategies.

But it also allows funds to establish larger positions without putting in additional own capital. The larger the position, the more reliant it becomes on the proper functioning of lenders, exchanges, and custodial institutions in times of crisis.

When every penny of the fund's capital is fully utilized, the backup funds available to handle sudden situations will diminish. Shared collateral can alleviate issues like "overall profitability but forced liquidation due to funds stuck in other accounts," but it cannot fundamentally eliminate dependency on financing, market liquidity, and the stability of exchange services.

To judge the effectiveness of capital efficiency, one should not look at stable market phases, assessing how large an exposure $1 million can leverage; rather, it should focus on how many effective positions the fund can safely retain amidst a crisis environment where everyone wants to withdraw their funds.