How much
The honest answer is somewhere between 0% and 5% for most retail investors who already hold a diversified equity-and-bond portfolio. The case for non-zero is that Bitcoin in particular has shown low correlation with equities over multi-year windows, has a fixed supply schedule, and is now accessible through regulated wrappers - so the diversification argument is real even if you remain sceptical about long-term valuation. The case for capping at 5% is that the asset class remains higher-volatility and higher-tail-risk than equities, and the diversification benefit saturates quickly above 5%.
Aggressive practitioners go higher; conservative ones stay at 0%. There is no universal correct answer, only the answer that fits how much sequence-of-returns risk you can tolerate in the next decade.
Spot ETFs vs direct holding
There are now three real ways to hold Bitcoin and Ether for a UK or EU retail investor. Direct on a regulated exchange (Coinbase, Binance, Bitpanda, Kraken). A spot ETF or ETP through a mainstream broker. A custodial wrapper inside a wealth platform (some private banks now offer this).
Each has trade-offs.
- Direct exchange. Lowest fee, highest control. You manage the seed phrase or trust the exchange's custody. Carries platform-credit risk (FTX is the warning shot). For amounts above a few thousand euros, self-custody hardware wallets become worth considering.
- Spot ETF / ETP. Familiar wrapper, fits inside an ISA, a Depot, an assurance-vie, a PEA-PME. Higher TER (0.20-0.95% typically) than holding direct. No self-custody overhead. Probably the right default for amounts you would otherwise hold in a brokerage account.
- Custodial wealth wrapper. Highest fee. Useful only if you are a private-bank client who already values that channel.
Where the wrapper interaction gets interesting
A spot Bitcoin ETP inside a UK ISA is a genuinely new construction. The ISA wrapper protects against UK capital gains tax on growth and against UK income tax on any distributions; the ETP gives you direct Bitcoin exposure. For UK higher-rate taxpayers, the after-tax case for a small ISA-wrapped Bitcoin position is meaningfully better than the same exposure in a GIA. The same logic applies in France via the CTO (assurance-vie can hold UC-listed crypto ETPs in some 2025+ contracts), in Germany via the Depot, and so on.
This is not a recommendation to do it. It is a flag that the choice of wrapper is now part of the crypto-allocation question, not separate from it.
Beyond Bitcoin and Ether
Most of the case above is for Bitcoin and Ether specifically - the two assets with the deepest liquidity, the clearest regulatory treatment, and the most institutional adoption. Smaller-cap tokens carry a different risk profile: higher idiosyncratic risk, weaker custody options, less coverage from regulated wrappers, and a wider range of failure modes. The 5% allocation argument above does not extend cleanly to a long tail of micro-cap tokens.
If you do hold tokens beyond BTC and ETH, treat them as a satellite to a satellite - a small percentage of an already-small allocation - and be honest about the higher tail risk.
Visibility across exchanges
If your crypto allocation crosses two or more venues - say Binance for spot, a hardware wallet for cold storage, and an ETP in your Depot - the operational pain of seeing all of it on one screen used to be high. Open-banking aggregators are starting to reach into the crypto-exchange surface as well as the bank/broker surface. Bank AI brings Binance, Bitpanda, Coinbase, Kraken and Crypto.com into the same daily-refreshed net-worth view as your bank accounts and brokers, with allocation visible at the asset-class level.
Wallet-level visibility (Ledger, Trezor) is harder because it is outside the PSD2 perimeter - those are typically still manual entries. The right answer for hardware-wallet holdings is to mark them in the app and update them periodically.