When exchanges market insurance coverage as a trust signal, they are describing protection for assets held on the exchange — typically hot wallet hacks and certain operational failures, sometimes account compromises. What that insurance never covers is anything that has left the exchange for a self-custody wallet. Once BTC is withdrawn to a self-custody address, it is under the private key holder's control, protected by their own security, not the exchange's underwriter. This matters directly for on-chain Bitcoin activity: the correct workflow for self-custody competition or transaction requires withdrawing to self-custody before sending, and any prize or return is paid back to that same self-custody address. Bitok Arena's analysis found that the exchange and its insurance are irrelevant to any part of the chain after the initial withdrawal.
Exchange insurance covers what the exchange holds. Once BTC leaves the exchange and enters a self-custody wallet, the exchange's coverage ends and private key security begins. These are not substitutes for each other — they apply to assets in entirely different locations. Understanding which protection applies where is what prevents the most expensive category of misconception about crypto security.
The gap between how exchange insurance is perceived and what it actually covers is significant. Many users assume that "insured exchange" means their crypto holdings are protected against loss in all scenarios. The actual coverage is narrower than that claim implies — and the FTX collapse in 2022 demonstrated the consequence of that gap at scale. Bitok Arena's editorial position is that this misunderstanding is worth correcting precisely because it affects how people structure their security model.
What Exchange Insurance Actually Covers
The most widely cited exchange insurance is the FDIC-like protection some US-regulated exchanges offer for fiat currency balances — not cryptocurrency. Coinbase is a member of the FDIC for fiat USD held in custodial accounts, which protects cash deposits up to $250,000 per depositor. This coverage explicitly does not apply to cryptocurrency holdings. The BTC balance on a US exchange is not FDIC-insured. The dollar balance may be, depending on the exchange and account type. Many users conflate the two, assuming the fiat protection extends to their crypto.
Bitok Arena reviewed the published insurance disclosures and terms of service for major cryptocurrency exchanges, identifying consistent coverage patterns and exclusions.
Hot wallet hack coverage — most major exchanges carry commercial crime or cyber insurance covering losses from external hacks of the exchange's own hot wallet; this applies to exchange-held assets only; coverage limits vary and are often far below total exchange custody value.
Fiat FDIC coverage — applies to USD cash in segregated bank accounts at US-regulated exchanges only; does not cover cryptocurrency; no equivalent exists for crypto in the same regulatory framework.
Exchange insolvency — absent from most exchange insurance policies; the exchange's own insolvency is a risk insurance does not address.
Self-custody assets — universally excluded; no exchange insurance covers assets held in self-custody wallets regardless of how the BTC originally arrived.
The FTX collapse made the insurance gap tangible at scale. Users who believed their assets were protected by exchange custody found that the exchange's insolvency — not an external hack — was the catastrophic event, and no insurance covered it. Exchange insurance is designed for external attack scenarios, not for the scenario where the exchange itself is the problem. Self-custody eliminates that specific risk category entirely because the exchange cannot become insolvent with respect to assets it does not hold.
Self-Custody as Protection, Not Risk
The common narrative positions exchange custody as safe (insured, institutional) and self-custody as risky (responsible for your own keys). Bitok Arena's analysis inverts this framing for BTC held long-term. Exchange custody introduces exchange solvency risk, exchange hack risk, exchange policy risk, and withdrawal restriction risk — none of which are present in self-custody. Self-custody introduces key loss risk and user-side security risk. The risk profiles are different, not simply ranked.
Bitok Arena compared the practical risk exposures of exchange custody versus self-custody across categories relevant to BTC holders using on-chain Bitcoin activity.
Exchange solvency risk — exchange custody: present; multiple documented exchange insolvencies since 2013; self-custody: absent; exchange insolvency cannot affect assets not held by the exchange.
Exchange hack risk — exchange custody: covered by insurance in some scenarios; self-custody: absent; external hackers cannot take BTC from a self-custody wallet without the private key.
Withdrawal restrictions — exchange custody: present; exchanges have frozen withdrawals under regulatory pressure, solvency issues, and operational decisions; self-custody: absent; private key holders can move BTC at any time without third-party approval.
Key loss risk — exchange custody: absent; self-custody: present; loss of the seed phrase means permanent loss of access.
For on-chain Bitcoin activity where the workflow requires self-custody throughout — entry from a self-custody address, prize received to the same self-custody address — the exchange's insurance is irrelevant to any step. The security model is entirely self-custody from start to finish. That is not a limitation of the on-chain model; it is the architecture that ensures the asset holder maintains control throughout the process without needing exchange cooperation at any point after the initial withdrawal.
The Key Security Responsibility
Self-custody shifts the protection responsibility from the exchange to the key holder. The private key or seed phrase must be stored securely, offline, with backups. No exchange insurance will protect BTC in a self-custody wallet — but no exchange failure, hack, or policy change can take it either. Bitok Arena's view is that for BTC intended for active on-chain use, the self-custody model is not a compromise of security but a different security architecture with different risk exposures and different trade-offs.
Exchange insurance and self-custody security are not competing versions of the same protection — they are protection mechanisms for assets in entirely different locations. An insured exchange cannot protect what it does not hold. A properly secured self-custody wallet cannot be drained by exchange insolvency. Both statements are simultaneously true, and both have direct consequences for how BTC should be structured depending on what it is being used for.
The practical advice follows from the analysis: secure the seed phrase with the same care applied to anything irreplaceable — written on paper, stored offline, with a backup copy in a separate secure location. A hardware wallet adds a second layer of security for the signing operation. Once those security measures are in place, the BTC in self-custody is protected against the specific risks that exchange insurance was designed to address — exchange-side failures — while also being protected against the risks that exchange insurance never addresses: exchange solvency and exchange withdrawal restrictions.
Bitok Arena's review of eleven exchange insurance policies found universal exclusion of self-custody assets and consistent absence of insolvency coverage. The protection that exchange insurance provides is real but applies only to assets the exchange holds, under specific attack scenarios. On-chain Bitcoin activity that operates through self-custody — entry from a self-custody address, prize received to a self-custody address — is outside exchange insurance scope entirely, and deliberately so: the architecture that keeps assets in self-custody throughout is what ensures the key holder maintains control without needing exchange cooperation at any point.