FCA Crypto Stress Tests: Why Capital Alone Is Not Enough
Balance-sheet money does not guarantee customers can get crypto back safely. A bank outage, frozen custodian or blockchain queue can turn stress into a crisis. This article examines why crypto firms’ resilience depends on liquidity, legal rights and operational continuity—and what stress tests must reveal before confidence starts to break.
Frankfurt, Germany
Oct 5, 2026
A crypto firm can appear adequately capitalised and still struggle to return customers’ assets. Its bank may stop processing payments, a custodian may suspend transfers, or a blockchain’s withdrawal queue may lengthen just as confidence deteriorates. The balance sheet matters, but so does the firm’s ability to operate when several dependencies fail together.
That is the central prudential question raised by the framework described for the Financial Conduct Authority’s FG26/10 guidance on overall risk assessment: can a crypto business withstand a severe disruption while preserving a credible route for customers to recover their money and assets?
The issue sits within a broader regulatory project. In its June 2025 consultation, CP25/15 on a prudential regime for cryptoasset firms, the FCA proposed financial-resource requirements for stablecoin issuers and cryptoasset custodians. The underlying supervisory concern is familiar from conventional finance: minimum resources are a starting point, not a complete account of resilience.
For crypto businesses, however, the assessment must accommodate a different combination of financial, technological and legal dependencies. A stress test that measures losses without examining access to assets can miss the mechanism through which a firm actually fails.
A liquidity problem can begin outside the balance sheet
Stablecoin issuers illustrate the distinction particularly clearly. Reserve assets may be sufficient to cover outstanding tokens in ordinary conditions. That does not establish that an issuer can meet a sudden wave of redemptions.
The relevant questions are more demanding. How quickly can reserves be converted into cash? Which banks and custodians must remain available? What happens if redemptions accelerate outside normal banking hours? Would selling assets under pressure create losses that weaken confidence further?
Reserve quality, liquidity and concentration therefore need to be assessed together. An issuer dependent on one banking relationship has a different vulnerability from one with genuinely usable alternatives, even if their headline reserve figures are identical. An alternative account is of limited value if it cannot support the necessary payment volumes or settlement arrangements during a disruption.
The comparison with banking is useful, but incomplete. Stablecoin issuers should not assume access to the liquidity facilities available to eligible banks. Their analysis must reflect the resources and arrangements they can actually use, rather than support that might become available in a crisis.
Customer assets are not a financial buffer
For custodians, the key distinction is between the firm’s own resources and the assets it holds for customers. Customer assets cannot simply be treated as available funding for operating losses or an orderly closure.
A credible assessment would examine how a market shock interacts with withdrawal demand, wallet infrastructure, reconciliation systems and third-party service providers. A custodian might have enough money to continue paying staff but be unable to process transfers because a critical provider is unavailable. Conversely, functioning technology does not ensure continuity if the business lacks funding to retain security specialists or maintain essential contracts.
Legal arrangements matter as much as technical access. Segregation, contractual rights and the involvement of sub-custodians can determine whether assets can be identified, transferred or returned during distress. The ability to move tokens on a blockchain is not necessarily the same as the legal authority to do so.
Custodial staking adds another constraint: availability may depend on protocol rules. Withdrawal queues, validator failures and slashing can affect both the value of assets and the timing of their return. Stress assumptions should not treat staked assets as immediately accessible when the underlying network does not permit that.
Recovery triggers need decisions behind them
Stress testing is useful only if its results change how a firm prepares and acts. A threshold for deteriorating liquidity or rising withdrawal delays needs an owner, an escalation route and a response that management can execute.
That response might involve obtaining additional funding, reducing exposures, activating another provider or beginning a wind-down. Each option has limits. Funding can disappear precisely when it is needed; substitute providers may decline a distressed client; and transferring a customer book may require contractual, regulatory and technical work that cannot be completed overnight.
Boards should therefore distinguish between actions already within the firm’s control and those dependent on another institution’s agreement. A recovery plan built around an uncommitted investor or an untested transfer arrangement is an assumption, not an assured source of resilience.
Wind-down planning requires the same discipline. Returning customer assets can involve prolonged reconciliation, specialist staff, functioning systems and continued access to banks and infrastructure providers. Those costs may persist after revenue has fallen sharply. Financial resources need to support the closure process, not merely the decision to close.
The evidence matters more than the model’s complexity
Crypto markets offer limited historical evidence for some combinations of stress. That makes transparent assumptions particularly important. Firms need to explain why a scenario is severe enough, how exposures are measured and which uncertainties materially affect the result.
Sensitivity analysis can reveal dependence on assumptions about redemption speed, asset-sale discounts or outage duration. Reverse stress testing can work backwards from an inability to continue operating or return assets, identifying the combination of events that would produce that outcome.
These techniques do not predict the next crisis. Their value is in exposing weaknesses before management has to confront them under pressure.
The distinction between guidance and rules also remains important. Non-Handbook guidance does not, by itself, create a new binding requirement. Its significance lies in explaining the regulator’s approach to applicable rules; the legal obligation must be established from the relevant Handbook provisions and their commencement arrangements.
For firms and their banking counterparties, the broader lesson is straightforward. Capital, liquidity, operational continuity and customer-asset protection should not be assessed in isolation. A business is resilient only if those arrangements continue to work together when confidence falls—and if management can recognise when they no longer do.
Source note: The supplied account identifies FG26/10 as guidance published on 30 September 2026. That publication, its contents and the associated CRYPTOPRU 7 requirements could not be independently verified. The discussion above distinguishes established consultation context from analysis of the described framework; it does not confirm that the guidance or requirements are in force.