A payments company discovers that a seven-figure stablecoin balance has been swept from its hot wallet to an unknown address. Forensic tracing identifies the first exchange where the funds landed. The legal team files for a freezing order within forty-eight hours – and the court asks two questions before granting relief: what undertaking in damages will the claimant offer if the order proves wrongly obtained, and what security will the claimant post for the defendant's costs if the claim ultimately fails? These two obligations – the cross-undertaking in damages and security for costs – can determine whether interim relief issues at all, yet they receive little advance planning in the rush to stop a withdrawal.
Cross-undertakings and security for costs are the price of speed in asset recovery. A claimant seeking a freezing order or proprietary injunction before trial must compensate the respondent if the order was unwarranted, and must demonstrate that it can meet an adverse costs award if it loses. For digital-asset cases those obligations carry three complications: volatile asset values make damages hard to quantify at the application stage, the global movement of crypto requires relief across multiple forums with different security practices, and the victim company may lack liquid reserves after the theft itself. This analysis examines how cross-undertakings operate in crypto recovery, the courts' approach to security when the underlying asset is a digital token, and the funding structures that allow a business to pursue relief without tying up capital it no longer has.
The cross-undertaking in damages – what it is and when it bites
A cross-undertaking in damages is the claimant's promise to compensate the respondent for loss caused by the interim order if the claimant's underlying claim fails or the order should not have been granted. The undertaking is given to the court, not to the respondent, but it creates an enforceable obligation: if the order later proves wrongly made, the respondent may apply to enforce the undertaking and recover provable loss. In asset-recovery litigation the undertaking serves two functions. It protects a defendant who turns out to have acted lawfully from the commercial harm of having assets frozen during the dispute. And it acts as a brake on unmeritorious applications: a claimant unwilling to back its case with money will hesitate before seeking ex parte relief.
The undertaking is a condition of most interim injunctions in common-law forums. In England and Wales it is standard for a worldwide freezing order; the court will not grant the order without it unless the claimant can show exceptional circumstances. The same expectation applies in the DIFC Courts, in Hong Kong and in Singapore. The BVI, Cayman and other offshore forums follow the English practice. A party that refuses the undertaking or cannot satisfy the court of its ability to meet it will not obtain interim relief, regardless of the strength of the underlying claim.
For a theft victim the undertaking creates immediate risk. If the court later finds that the frozen respondent was not the thief – perhaps the funds moved through an innocent intermediary or the forensic trail was misread – the claimant must pay damages for the period the respondent's assets were locked. Those damages can include lost trading profits, financing costs on blocked funds, reputational harm and the defendant's own legal fees incurred resisting the order. Where the frozen asset is a volatile token, the potential liability multiplies: if the price rises sharply while the order is in force and the respondent can show it would have sold at the peak, the cross-undertaking covers that opportunity loss.
Quantifying the cross-undertaking when the asset is volatile
Damages under a cross-undertaking are assessed on normal contractual principles: the respondent must prove that the order caused a loss it would not otherwise have suffered, and that loss must be reasonably foreseeable. For frozen fiat currency the calculation is straightforward – lost interest, perhaps financing charges. For frozen cryptocurrency or tokenised assets the assessment becomes speculative. A respondent might argue that it would have traded the frozen tokens, or staked them, or used them as collateral in a DeFi protocol. Each path produces a different counterfactual profit, and the claimant's liability under the undertaking turns on which scenario the court accepts.
The English courts have begun to address valuation of crypto in the damages context. Where a freezing order locks a portfolio of digital assets, the court will not award damages based on the highest intraday price during the freeze period unless the respondent can show a realistic plan to have sold at that moment. Damages are more commonly assessed at a reasonable sale point – the closing price on the day the order lifts, or an average over a short window. In one recent matter a respondent claimed that a frozen holding of a Layer-1 token would have been staked for yield during the six-month freeze; the court accepted the principle but required evidence of the respondent's usual staking practice and reduced the award to reflect the risk that validators might have been slashed.
For the claimant structuring the undertaking at the application stage, the risk is that the court imposes a cap or requires the claimant to fortify the undertaking with security. Fortification means posting a bond, a payment into court or a charge over assets to back the promise. Whether fortification is ordered depends on the claimant's financial position and the likely scale of loss if the order proves wrong. A well-capitalised corporate claimant will often give an unfortified undertaking and the court will accept its balance sheet as sufficient comfort. A claimant that has just lost its operating funds in the same theft may be required to post cash or obtain a bond before the freezing order issues. The alternative is to ask the court for an undertaking limited to a stated sum – but if the cap is too low, the court may refuse the order altogether on the basis that the respondent's risk is not adequately covered.
Security for costs – when it is ordered and how much
Security for costs is a separate mechanism. It requires the claimant to post funds or a bond to cover the defendant's legal costs if the claimant loses at trial. Unlike the cross-undertaking, which runs only to interim relief, security for costs applies to the entire claim. The court's power to order security is discretionary and turns on a set of gating conditions. In most common-law forums the defendant must show either that the claimant is outside the jurisdiction and may be unable to meet a costs order, or that the claimant is impecunious and likely to be unable to pay if it loses. Once a condition is met the court weighs a series of factors – the strength of the claim, the claimant's reason for lack of funds, whether security would stifle the litigation – and decides whether to order security, and if so in what amount.
The jurisdictional gateway is straightforward: a claimant company incorporated in a non-Convention state and with no assets in the forum jurisdiction will usually be ordered to provide security unless there is a reciprocal enforcement treaty. A crypto business incorporated in the BVI, suing in England, can expect a security application from the defendant at the first case-management conference. The impecuniosity gateway is more fact-sensitive. A claimant that was solvent before the theft occurred but now has no funds because those funds were stolen is, on its face, impecunious. The question is whether that impecuniosity should disentitle the victim from access to the court. The modern English approach is to look at whether the lack of funds is connected to the defendant's conduct. If the defendant is alleged to have stolen the claimant's entire treasury, the court will be reluctant to shut down the claim by imposing security the claimant cannot afford. But if the claimant's financial distress arises from unrelated business failure, and the theft is only one element of a wider insolvency, the court is more likely to order security.
The quantum of security is set by reference to the defendant's likely costs to the next procedural stage or to trial. In a multi-defendant crypto-recovery claim with contested jurisdiction, disclosure disputes and tracing evidence, defendant costs can reach six figures sterling before trial even begins. The court may order security in tranches – an initial tranche to cover costs to disclosure, a second to trial – so the claimant is not forced to post the entire sum at once. Where the claimant cannot raise the security the court may stay the proceedings until funds are available, or in extreme cases dismiss the claim. That outcome is rare in a genuine theft case with a clear evidential trail, but it remains a risk the claimant must price from the outset.
The cross-border reality – multiple undertakings and inconsistent security
Digital assets move across borders and through multiple intermediaries in minutes. A recovery strategy often requires parallel freezing orders in several forums: an English worldwide freezing order to bind the defendant globally, a recognition application in the DIFC to freeze the exchange account in Dubai, perhaps an ancillary order in Singapore or Hong Kong to lock funds held by a local custodian. Each forum will impose its own cross-undertaking and may make its own security-for-costs order. The claimant thus faces cumulative exposure: three undertakings, three sets of security, and the possibility that a defendant in one jurisdiction successfully enforces an undertaking for loss that overlaps with a claim in another.
The English worldwide freezing order does not eliminate the need for local recognition. An exchange based in Dubai will not honour an English order without DIFC recognition, and the DIFC Courts will require their own undertaking as part of the recognition process. That undertaking may be narrower in scope – covering only loss arising from the DIFC-stage freeze – but it is a separate obligation. If the defendant later argues that the DIFC order caused additional reputational harm in the UAE market, the claimant's English undertaking will not cover it; the defendant must enforce the DIFC undertaking. The practical risk for the claimant is that costs multiply and the total security demanded across all forums becomes unaffordable. In that situation the recovery strategy must be staged: obtain the English order with undertaking and security, use forensic evidence to narrow the target, then seek recognition only where funds are confirmed to sit. Filing for recognition everywhere in hope of a hit will exhaust both the claimant's capital and the court's patience.
Who funds the undertaking when the victim has no reserves
A business that has lost its operating capital in a hack has little room to post security or fortify a cross-undertaking. The options are to raise fresh equity, which dilutes existing shareholders and takes time; to find a litigation funder willing to indemnify the undertaking, which is possible but expensive; or to ask the court to accept a limited undertaking or to waive security on impecuniosity grounds, which works only where the defendant's own conduct caused the loss. Each path has trade-offs.
Litigation funding in the crypto-recovery context has evolved rapidly. A handful of specialist funders now offer packages that include both the costs of the litigation and the cross-undertaking indemnity. The funder pays the claimant's legal fees, covers any damages under the undertaking if the claim fails, and in return takes a percentage of recovered assets – typically in the range of thirty to fifty percent depending on risk. For a company with no other funding route that return may be acceptable: half a recovery is better than none. The challenge is that most funders require a minimum recovery value and a strong merits opinion before committing. A case with a clear on-chain trail, identified exchange accounts and a solvent defendant is fundable. A case that depends on unproven novel arguments about crypto property or that targets anonymous wallet holders with no visible assets will not attract funding at any price.
Where funding is not available the claimant may apply for a reduced cross-undertaking or a waiver of security. The English approach to reducing the undertaking in impecuniosity cases is set out in a line of Commercial Court decisions: the court can accept a limited undertaking if the impecuniosity was caused by the defendant's alleged wrongdoing, the claim has strong prospects, and requiring a full undertaking would effectively deny the claimant access to justice. The court will not automatically grant a nil undertaking – even a victim must offer some protection – but may accept an undertaking capped at a sum the claimant can realistically meet, backed by directors' personal guarantees or a charge over remaining assets. The same principles apply to security for costs: a claimant stripped of funds by the defendant's fraud will usually not be required to provide security, but if it has other creditors or if insolvency looms the court may still order security to protect the defendant from a hollow costs award.
Strategic decisions at the application stage
The cross-undertaking and security questions must be addressed in the initial freezing application, not left for later. If the claimant drafts its application and supporting evidence without considering what undertaking it can afford, the court may grant an order in principle but then refuse to issue it because the claimant cannot meet the undertaking. That wastes time and tips off the defendant. The better practice is to model the undertaking exposure before filing. How much might the respondent lose if the order is later discharged? If the frozen asset is a token with high daily volatility, model a worst-case scenario – the price doubles and the respondent claims lost trading profit. If that potential liability is larger than the claimant's available capital, the structure must change: either seek funding, or ask for a limited undertaking and argue impecuniosity, or narrow the relief sought to reduce the respondent's potential loss.
One approach in multi-respondent cases is to phase the relief. Obtain a narrow order against the first exchange where funds landed, with a correspondingly smaller undertaking. Use that order to secure disclosure of the next destination. Once the onward trail is confirmed, apply for wider relief with a fuller evidential record that supports a stronger case for extended freezing. The phased strategy has two advantages. The initial undertaking is capped at the value frozen at one exchange, making it easier to fund or fortify. And if the disclosure reveals that the funds moved to a regulated custodian willing to hold them pending trial, the need for coercive relief may disappear, removing the undertaking risk altogether.
Another tactical question is whether to ask for an injunction against the token issuer. If the stolen asset is USDT or USDC, the claimant can apply to freeze the tokens at contract level rather than merely restraining the holder. The issuer's participation changes the security calculus: the issuer is unlikely to apply for security for costs because it is neutral, and the cross-undertaking runs only to loss caused to the respondent by the inability to transact in those specific tokens. The downside is that the issuer will require a clear legal basis – a court order from a recognised forum, usually with law-enforcement involvement – before freezing on-chain. The claimant cannot simply write to Tether and expect compliance; it needs an order, and obtaining that order will itself require an undertaking.
How defendants enforce an undertaking and what it costs
If the court later discharges the freezing order or finds that it should not have been granted, the defendant may apply to enforce the cross-undertaking. Enforcement is a two-stage process. First the court must find that the order was wrongly made – either because the underlying claim has failed, or because material non-disclosure at the ex parte stage rendered the order voidable, or because supervening events show the order was never justified. Second the defendant must prove its loss and show that the loss was caused by the order. The burden is on the defendant to quantify what it lost and to demonstrate that it would have acted differently but for the restraint.
In practice enforcement is expensive. The defendant must run a mini-trial on loss, often involving expert evidence on market conditions, trading strategies and counterfactual asset values. For a frozen cryptocurrency portfolio the defendant may call an expert to reconstruct what the portfolio would have looked like had the defendant been free to trade. The claimant can challenge every assumption in that reconstruction. The costs of the enforcement application can themselves reach five or six figures, and the court will usually order the unsuccessful party to pay those costs on the standard basis. For a claimant that has already lost the underlying claim and now faces a substantial undertaking award, the costs of defending the enforcement application add a further layer of exposure.
The risk of enforcement is one reason why insurers and funders price crypto-recovery mandates at a premium. The funder must reserve not only for the possibility of losing at trial, but also for the possibility of an undertaking claim that could exceed the original value of the frozen assets. That risk is highest where the frozen asset is a low-cap token or a new protocol token with high volatility. In one recent matter a claimant obtained a freezing order over a portfolio of DeFi governance tokens; the price of one token increased tenfold during the litigation, and when the claim was struck out on jurisdictional grounds the defendant's undertaking claim valued the foregone profit at a sum greater than the claimant's entire remaining balance sheet. The case settled, but it illustrates the tail risk that sits behind every interim order.
Mistakes that increase undertaking exposure
Three mistakes recur in the cross-undertaking context and each can be fatal to the relief or ruinously expensive to correct. The first is failing to disclose adverse facts at the ex parte application. An applicant for a freezing order must make full and frank disclosure of every fact that might bear on the court's decision to grant relief. That includes facts that weaken the claim, facts that suggest the respondent may not be the thief, and facts that bear on the applicant's own financial position. If the applicant conceals or forgets to mention that its own security was inadequate, or that a director may have contributed to the loss, the court will discharge the order for material non-disclosure. The cross-undertaking then crystallises: the applicant must pay damages even though the respondent may well have been guilty, because the order should never have issued in the first place.
The second mistake is over-freezing. If the evidence shows that $500,000 was stolen but the freezing order restrains $2 million of the respondent's assets, the respondent will apply to vary the order and may seek damages or costs in relation to the excess freeze. The court expects the freezing order to be proportionate to the claim, with a modest margin for costs and interest. An order that freezes everything the respondent owns when the alleged loss is a fraction of that sum invites a challenge and increases the undertaking risk. In crypto cases the risk is compounded by price movements: if the stolen amount was stated in dollar terms at the date of filing but the order restrains a token balance, a subsequent price rise may mean the order now covers far more value than the original loss. The claimant should build in a mechanism to release excess value or convert the restraint to a dollar cap.
The third mistake is treating the undertaking as a formality. Some claimants see the undertaking as a procedural checkbox and assume the risk of enforcement is theoretical. For large corporates with ample reserves that may be true – the cost of defending an undertaking claim is a line item. For a startup that lost ninety percent of its treasury in the hack, the undertaking is an existential risk. If the court later finds for the defendant, and the defendant successfully claims lost trading profit on a volatile token, the undertaking liability could drive the claimant into insolvency. The decision to seek interim relief must therefore be taken with a clear view of worst-case exposure and a plan to fund it if it materialises. In our practice we model the undertaking as part of the funding proposal before any application is filed; a recovery strategy that does not account for the cost of being wrong is not a strategy.
Cross-undertakings in disclosure and proprietary orders
Freezing orders are not the only relief that triggers an undertaking. A Norwich Pharmacal order – used to compel an exchange or intermediary to disclose the identity and transaction records of a user – also requires a cross-undertaking, though typically a lighter one because the target of the order is a neutral third party and the risk of loss is lower. A proprietary injunction over specific tokens, argued on the basis that the tokens are traceable proceeds of the claimant's property, will require an undertaking because the order restrains the defendant from dealing with an asset the defendant may ultimately be found to own. Bankers Trust disclosure orders follow the same principle. Each form of interim relief has an associated undertaking tailored to the nature of the order and the risk it creates.
For a cross-border recovery the claimant may be giving undertakings at multiple stages: one for the initial worldwide freezing order, another for a Norwich Pharmacal disclosure order against the first exchange, a third for recognition of the freezing order in the forum where a second exchange is located, and a fourth for a proprietary injunction over traced assets. The cumulative exposure is rarely discussed in the heat of filing but it governs whether the strategy is affordable. Before the first application the claimant should map the full relief sought, estimate the undertaking for each element, and confirm that funding – whether from balance sheet, equity or a funder – covers the aggregate risk.
Forum choice and its effect on security practice
Not all forums treat cross-undertakings and security alike. The English courts impose them routinely and enforce them with rigour. The DIFC Courts follow the English model but with a smaller local bar and less developed case law on crypto-specific damages, enforcement may be quicker and sometimes cheaper. Hong Kong applies similar principles but with a slightly more cautious approach to freezing relief; security is often required where the claimant is offshore. Singapore has tightened its practice around cryptocurrency injunctions after several high-profile cases and now expects clear evidence of tracing and a realistic undertaking before granting wide relief. The BVI and Cayman courts are procedurally aligned with England but may be more willing to accept a limited undertaking if the claimant's connection to the jurisdiction is strong and the defendant's alleged conduct is egregious.
For a claimant weighing where to file first, the choice of forum affects not only which substantive law applies but also how much security must be posted and whether the undertaking can be capped or funded through third-party means. A claimant incorporated in a jurisdiction with a reciprocal enforcement treaty may avoid a security-for-costs order if it sues in a treaty partner state. A claimant with no treaty may face an automatic security demand. In multi-forum strategies the usual pattern is to obtain the primary order in the forum with the strongest connection to the parties or the assets, then rely on recognition procedures elsewhere. That concentrates the undertaking risk in the primary forum and avoids duplicating security across several courts. The cross-border structuring decision must be taken at the outset, before any application, because filing in the wrong sequence can trigger security orders that could have been avoided.
Addressing the concern – is it worth applying at all?
A common assumption is that the cost and risk of a cross-undertaking make freezing relief unaffordable for any claimant that is not a major institution. The reality is more nuanced. Where the stolen asset is identifiable, the evidential trail is clear, and the defendant is within reach of a recognised forum, the prospect of recovering a significant portion of the loss justifies the undertaking risk. The risk is manageable if properly structured: a limited undertaking backed by insurance or funding, phased relief that expands only as evidence strengthens, and a realistic assessment of volatility at the modelling stage. What makes a recovery case unaffordable is not the existence of the undertaking but the failure to plan for it, leaving the claimant unable to issue the order even after the court grants permission in principle.
The courts expect claimants to take undertakings seriously and to come prepared. An applicant that treats the undertaking as a mere procedural formality will lose credibility with the judge and may find the order discharged at the first return hearing. An applicant that models the risk, secures the funding, and explains candidly to the court how the undertaking will be met is far more likely to obtain robust relief and to keep it in place through to trial. In our disputes practice we treat the undertaking as a first-order strategic question, not an afterthought; a recovery plan without a funding plan is incomplete.
Related at OBOLUS
- Disputes & Asset Recovery for Digital Asset Businesses – end-to-end representation from forensic tracing through recognition and enforcement.
- Insurance and Litigation Funding for Crypto Recovery – how to finance interim relief when operating capital is gone.
- Crypto Fraud Asset Recovery – the Disputes Angle – the procedural path from discovery to freezing relief across borders.
FAQ
Can stolen crypto actually be recovered?
Yes, where the movement is traced on-chain to an identifiable exchange account or custodian. Courts in England, the DIFC, Hong Kong and Singapore routinely grant freezing orders and disclosure orders over digital assets. The success rate depends on the speed of the application – filing within hours or days, not weeks – the quality of the forensic evidence, and whether the destination exchange or custodian is within a jurisdiction that will recognise the order. Stablecoins issued by Tether and Circle can be frozen at contract level if the issuer receives an appropriate court order, typically with law-enforcement involvement. Recovery becomes harder if the funds are moved to an unhosted wallet, mixed through a tumbler, or cashed out before the order issues, but even in those cases partial recovery is often achievable.
How fast must I act after a digital-asset theft?
The recovery window is measured in hours after the theft is discovered. On-chain assets move instantly and exchanges process withdrawals continuously. If you discover a misappropriation, instruct forensic tracing immediately and prepare the freezing application in parallel. The application can be filed within twenty-four to forty-eight hours if the facts and evidence are marshalled quickly. Delay of even a few days can mean the difference between freezing the funds at the first exchange and watching them disperse to a dozen wallets across jurisdictions. Where the stolen amount is material and the claimant has no prior relationship with recovery counsel, the first call should happen within hours of detection, not days.
Can a court freeze assets held on an exchange?
Yes, through a combination of a freezing order directed at the account holder and, where necessary, a separate order directed at the exchange as a third party. An English worldwide freezing order restrains the defendant from dissipating any assets, including cryptocurrency held in an exchange account, and requires the defendant to disclose those holdings. The exchange itself is not bound by a purely domestic order unless it has a presence in the jurisdiction or unless the order is recognised by the court in the exchange's home jurisdiction. In practice most major exchanges will comply with a recognised order once it is served through proper channels, particularly where the exchange operates in a regime with strong AML expectations. Obtaining that compliance often requires local counsel in the exchange's jurisdiction to secure recognition, and the claimant's cross-undertaking and security obligations apply at each stage.
OBOLUS is an independent digital-asset law boutique acting only for businesses. We advise exchanges, custodians, token issuers and funds on licensing across 70+ jurisdictions, on disputes and on-chain asset recovery across 25+ forums, and on the tax, banking and compliance that sit around them. Digital assets are the whole of our practice. In a recent recovery matter a payments company traced misappropriated stablecoins through two exchanges; we secured a disclosure order in a leading common-law forum and the funds were frozen before withdrawal. We move for freezing relief and exchange disclosure while the trail is live. To discuss your situation, contact info@oboluslaw.com.
By Glen Sorensen, Disputes & Recovery Analyst – specialist in cross-border asset recovery, freezing orders and on-chain enforcement for digital-asset businesses.
This publication is general information about the law and does not constitute legal advice. It is not a substitute for advice tailored to your circumstances. OBOLUS accepts no liability for action taken or not taken on the basis of this material. For advice on your situation, contact info@oboluslaw.com.