Explainers
Crypto ETF risks, ranked by what matters
An honest inventory of crypto ETF risk — market, custody, liquidity, closure, tax and structural — with what each one would actually cost you and what you can do about it.
Risk pages in this space tend to fail in one of two directions. Either they list crypto's dramatic failures as if an ETF were exposed to them, or they wave at "volatility" and move on to a buy button. Neither is useful.
What follows is the actual risk inventory for a US-listed spot crypto ETP, ordered by how much it is likely to cost you, with the mitigations that exist and an honest note where none does.
Market risk comes first, and it is not close
Bitcoin has repeatedly drawn down more than half from a prior high. In the year to March 2026 it fell roughly 27%, and by 2 September 2026 it traded near $77,000 with ether around $2,400. The fund wrapper contributes nothing to softening that. It is a distribution mechanism, not a hedge.
This matters because the ETF's arrival in brokerage accounts placed a highly volatile asset next to index funds and bond funds on the same screen, in the same format, with the same order ticket. The interface is identical. The risk is not, and the presentation quietly encourages people to size a bitcoin position the way they would size an S&P 500 position.
The full inventory
| Risk | Likelihood | What it means | Mitigation |
|---|---|---|---|
| Market risk | Very high | Bitcoin has repeatedly fallen more than 50% from a high. It was down roughly 27% year to date as of March 2026. | Position sizing. Nothing else meaningfully helps. |
| Custodian failure | Low probability, high impact | Nine of twelve US spot Bitcoin funds involve one custodian. Segregation should protect assets; process and delay would not be instant. | Prefer funds with diversified or affiliated custody. Split across wrappers. |
| Fund closure | Moderate for small funds | A liquidating trust sells holdings and distributes cash, forcing a taxable event on its timetable. | Avoid sub-$50m funds in taxable accounts. |
| Spread and execution | Certain | You pay the bid-ask on entry and exit. On thin funds this can exceed the annual fee. | Limit orders, mid-session, larger funds. |
| Tracking difference | Certain, small | A spot fund should lag bitcoin by roughly its fee. Materially more signals an operational issue. | Compare annual fund return against bitcoin return. |
| Fee drag | Certain | 0.14% to 1.50% a year, paid by the trust selling bitcoin — so the coin behind each share shrinks. | Choose on total cost, not headline fee. |
| Regulatory change | Uncertain | Tax treatment of digital assets is still evolving; wash-sale extension has been proposed but not enacted as of August 2026. | Do not build a strategy that only works under current rules. |
| No access to the asset | Certain | You cannot withdraw, send, spend or stake the underlying, and cannot trade outside US market hours. | Hold some directly if any of that matters. |
Ordered roughly by expected cost to a typical holder. The four marked 'certain' are not tail risks — they are what holding the product costs, and they are the ones most often left out of risk discussions.
Custody and counterparty risk
Coinbase Custody appears in nine of the twelve US spot Bitcoin ETP arrangements. That is a real single point of failure across most of the regulated category, and it should be stated without either alarm or dismissal.
The protections are meaningful: assets are held in segregated accounts in the trust's name rather than on the custodian's balance sheet, under audited controls with defined standards of care. The limitation is equally real: in a genuine failure, "not available to creditors" becomes a legal process with delay and uncertainty attached, not an instant transfer.
Sponsors are diversifying — BlackRock added Anchorage Digital for IBIT, CoinShares runs Coinbase and BitGo in parallel, and Fidelity self-custodies FBTC entirely. If this risk concerns you, those arrangements are the practical response available. Detail in custody explained.
Liquidity and closure risk
The largest funds are fine. IBIT trades tens of millions of shares daily at a median spread near 0.02%. The problem lives in the long tail, where several products still hold well under a billion dollars — and in the altcoin categories, where some funds hold single-digit millions.
Three costs stack in a small fund. Wider spreads, paid twice. Persistent premiums and discounts, because arbitrage on a small basket is barely worth an authorised participant's attention. And closure risk — a liquidating trust sells its holdings and hands you cash, realising your gain or loss in that tax year whether or not you intended to sell. Losing control of your realisation timing is a genuine cost for a long-term holder.
One risk you can remove entirely
Custodian risk exists because someone else holds the keys. Buying bitcoin on a licensed exchange and withdrawing it to your own hardware removes that party — and introduces key management as your responsibility instead. CEX.IO is FinCEN-registered, licensed for money transmission across US states, and authorised in Gibraltar as a DLT provider.
Tax and regulatory risk
Two live issues are worth tracking. First, the grantor-trust structure means the trust sells bitcoin to pay its fee, and shareholders are treated as owning a pro-rata interest in the underlying — so those sales are small reportable disposals on a position you never traded.
Second, wash-sale treatment is genuinely unsettled. The wash-sale rule under section 1091 applies to securities, and crypto classified as property has historically sat outside it. But many broker-dealers report spot Bitcoin ETP transactions as securities on Form 1099-B and apply wash-sale adjustments, while a grantor-trust look-through analysis suggests the rule should not apply. Congress and Treasury have proposed extending wash-sale rules to digital assets — most recently in a bill introduced in June 2026 — but as of August 2026 nothing had been enacted. That is an unresolved question, not settled guidance, and we say so on the tax page.
Risks people invent that are not really there
- "The fund might not hold the bitcoin." Authorised participants have capital at risk verifying holdings daily through creation and redemption, on top of audited SEC filings. See arbitrage.
- "BlackRock could seize the bitcoin." The sponsor does not own the trust's assets and cannot appropriate them. Governance risk here is about fee changes and operational decisions, not theft.
- "An exchange hack would wipe out the fund." Trust assets sit in segregated cold storage at a qualified custodian, not in an exchange's hot wallets.
- "ETF selling will crash bitcoin." Redemptions transfer coins to authorised participants, who sell into a global market that has absorbed far larger flows. See flows and price.
What you can actually control
- Position size
The only real defence against the dominant risk. Decide the number before you open the order ticket, and size it against a 60% drawdown rather than your expectation.
- Account type
A retirement account eliminates the tax layer of several risks on this page. See Roth IRA and 401(k).
- Fund selection
Larger funds for liquidity and closure risk, diversified custody if that concerns you, lowest total cost if you are holding for years.
- Execution discipline
Limit orders, mid-session. This converts an uncertain cost into a known one at zero expense. See limit orders.
- Wrapper mix
Fund shares in tax-advantaged accounts, direct holdings where access and 24/7 trading matter. Neither wrapper is strictly better — see ETF vs owning crypto.
Our take from the desk
The risk that costs real people real money here is not custodian failure or regulatory reversal. It is position sizing, and specifically the way a familiar brokerage interface makes a 70% drawdown asset feel like a portfolio holding rather than a speculation. We would trade every other mitigation on this page for a habit of writing down the position size before opening the order ticket.