Bitcoin for Family Offices: A Governance Decision

A family office does not have a Bitcoin allocation problem. It has a governance problem wearing an allocation problem's clothes. The size of the position is the easy part and the last part. Who holds the keys, what happens when the patriarch dies, how a trustee in George signs off on a transfer without ever seeing the seed: those are the questions that decide whether the holding survives the second generation. Settle the governance and the number almost chooses itself.

Why the traditional reserve playbook is under strain

A family office exists to do one thing across generations: protect purchasing power and hand it on intact. The tools it has always used for that job are quietly failing the brief.

Cash is the obvious casualty. The rand has lost more than 70% of its value against the dollar in twenty years, a slow bleed I set out in more detail in what the rand's slide means for Bitcoin. Dollar and euro balances erode more slowly but they still erode, because every major central bank has spent the last two decades expanding its money supply. Fixed income, once the dependable real-return anchor of a conservative portfolio, now swings between negative real rates and genuine default risk across the emerging markets a South African family is most exposed to.

Equities and property are supposed to carry the diversification load. They do, until the week you actually need them to. In a real liquidity shock the correlations between those asset classes climb toward one, and the spread that looked comfortable on a normal Tuesday collapses precisely when it is meant to protect you.

None of this makes the traditional playbook wrong. It makes it incomplete. A family thinking in decades wants at least one asset whose value does not depend on a committee's mood, and that is a narrow shortlist. The point of a reserve was always to hold something the government of the day could not quietly help itself to, and that discipline has atrophied in an age of easy debt and central bank intervention.

Reserve strategy is measured in decades. Speculation is measured in charts.

What makes Bitcoin different from other reserve assets

Gold has done this job for centuries. It is scarce, durable and sits outside any government's direct control, which is exactly why every serious reserve conversation starts there. Bitcoin keeps those properties and adds a few gold cannot offer.

Start with the supply. There will only ever be 21 million bitcoin, and that cap is not a promise printed on a certificate. It is a rule enforced by every full node running the software, rejected automatically if anyone tries to change it. In traditional asset theory every major reserve instrument is elastic: fiat is printed, bonds are reissued to refinance old debt, shares are diluted and even gold sees supply creep upward as mining technology improves. Bitcoin is the only major asset with a supply that answers to nothing. For a sovereign wealth fund, an endowment or a family office holding capital that must survive several political cycles, that single property changes the arithmetic.

Then there is control. A bearer asset held in self-custody cannot be frozen by a bank tired of your industry, cannot be caught in a custodian's insolvency and cannot be haircut in somebody else's restructuring. As James Caw puts it in The Strategic Reserve, it is not held for you, it is held by you. That is a different kind of ownership from a share register or a deposit slip.

It also moves. Bitcoin settles to anyone, anywhere, without asking permission of an intermediary. In a country with exchange controls that is a property worth understanding carefully rather than casually, because the legal reporting obligations around moving value offshore still apply in full. Understanding the plumbing is not the same as ignoring the rules.

Sizing the position without pretending there is a magic number

Clients want a number. I understand the instinct and I resist handing one over, because the honest answer is a range with your own circumstances plugged into it. What the research supports is that the interesting range is small.

Fidelity's digital assets team ran Bitcoin through a standard 60/40 portfolio and found the sharpest improvement in risk-adjusted returns came from moving between a 1% and a 3% allocation, with the maximum drawdown rising far less than the eye-watering volatility would suggest, because Bitcoin's low correlation and disciplined annual rebalancing keep its weight contained. Their broader framework points at 2% to 5% as the band with outsized upside in an optimistic adoption scenario, while capping the damage at under 1% of annual income if Bitcoin went to zero entirely. That last clause carries more weight than the upside. A family office is allowed to be wrong about Bitcoin, provided being wrong is survivable, and single-digit sizing keeps it survivable.

The instinct is global, not just a Fidelity spreadsheet. A recent survey of Japanese institutions found that of those planning to allocate to digital assets, sixty percent expected to put between 2% and 5% of the portfolio to work. Different continent, same shape of answer.

Say the family carries R80 million in investable assets. A 3% position is R2.4 million, an amount large enough to matter to the estate in twenty years and small enough that a brutal drawdown does not threaten a single school fee or capital call. That is the frame I work in, and I go through the mechanics properly in how I size a Bitcoin position for a family office. The number is downstream of the governance, never the other way around.

Custody and governance for multigenerational holdings

This is the part most family offices underestimate, and it is the part that actually decides the outcome. Bitcoin is controlled by private keys. Whoever holds the keys holds the Bitcoin. Lose them and the coins are inaccessible forever. Have them stolen and the coins are simply gone. There is no fraud department to call.

A single hardware wallet with a single keyholder is fine for a hobbyist. It is entirely wrong for a family office.

The position has to survive the death, incapacity, emigration or plain absence of the person who set it up, and it has to resist both theft and coercion at the same time. Those requirements pull against each other, which is exactly why multisig exists. A 2-of-3 arrangement means three keys exist but only two are needed to move anything, so no single key can act alone and no single lost key is fatal. The keys sit with different people in different places: a family member, a regulated fiduciary, a third leg in geographically separate cold storage. Custody stops being a password and becomes a governance layer, one that scales from a single family up to a sovereign treasury without changing the underlying logic.

The Vault I run is built on exactly that principle. You hold two keys on your own devices, a Trezor and a Ledger, while I hold a Coldcard recovery key as the regulated third leg through CAEP Asset Managers (FSP 33933). No one keyholder can move a coin alone. The family keeps direct bearer-asset ownership with majority control, and the governance lives inside a licensed framework rather than in somebody's memory. I do five of these setups a month, for families and entities both, and each one begins with the custody design before any Bitcoin is bought.

Inheritance is where the design earns its keep. A court can confirm your heirs. It cannot decrypt a private key. So the estate plan has to name the keys explicitly: who holds which one, where they physically sit, and what event triggers access. A well built multisig allows a spouse and an adviser to recover funds if the primary keyholder dies, while a third key stays sealed until a death certificate or a legal timeline unlocks it. The point I make to every family is that inheritance here is infrastructure, not paperwork. Heirs who inherit a vault they do not understand inherit a locked box, so the education happens while the founder is alive, and I walk through that whole failure mode in passing Bitcoin down across generations in South Africa.

Compliance, FICA and the right legal home

Bitcoin held by a South African family office has to be documented properly, and the documentation is not optional. Crypto asset service providers have been accountable institutions under the FIC Act since December 2022, so verifying identity and source of funds is a legal duty rather than a courtesy. Onboarding a trust or a company simply takes longer than an individual, which is a reason to start early, not a reason to avoid it.

Where the coins should actually live is the harder question. A trust, a company and natural persons each tax Bitcoin differently and each solves a different succession problem, and choosing badly is expensive to unwind because moving appreciated Bitcoin later is itself a taxable event. I weigh those trade-offs at length in holding Bitcoin in a company, trust or personal name, and the short version is that the structure decision belongs before the growth, which for Bitcoin arrives sooner than most families expect.

Whatever wrapper you land on, the deed or the shareholders agreement has to contemplate Bitcoin by name. Who may instruct a transaction, what quorum is required, how keys are managed when a trustee resigns or a director emigrates: these belong in the document itself. A trustee who buys Bitcoin the deed never authorised answers for it personally.

There is a habit worth naming here, because I see it defeat otherwise careful families. They treat Bitcoin as an add-on to an existing structure rather than an asset the structure was designed around, and then discover at the worst possible moment that the trust deed is silent, the executor has never heard of a seed phrase and the third key sits with someone who has since emigrated. Every one of those failures was cheap to prevent and ruinous to fix. The families who design for continuity from the first coin pass on more than capital: they pass on control, optionality and a structure their heirs can actually operate.

These are not hard problems. They are simply problems that reward being solved first, and a family office is built for exactly that kind of patience. If you want the fuller institutional picture for the year ahead, I set it out in the family office view of Bitcoin in 2026.

Get the governance right and the allocation is almost an afterthought. Get it wrong and the size of the position only decides how much your heirs lose. If your family is ready to have the structure conversation before the number conversation, book a family office structure session and we will start where the real difference lies.