How Much Bitcoin Should a South African Investor Actually Hold? A Framework for Sizing Your Position

The right amount of Bitcoin is not the amount that maximises your return. It is the amount you can watch fall by half without touching it. That distinction sounds like a technicality until you have sat with a client in George in early 2022, staring at a position down 60%, deciding in real time whether ten years of patience survives one bad quarter. This is the framework I actually use to size a position, built around the number you can hold rather than the number a spreadsheet prefers.

What follows is a way of thinking, not a recommendation to buy any particular amount. Your circumstances bend the answer, and no article knows your income, your existing portfolio or your temperament. Read this as the shape of the conversation I have with clients, then have the specific version with someone who can see your full picture.

Start with volatility tolerance, not return expectations

Most allocation frameworks begin with expected return, and for Bitcoin that is where they immediately go wrong. The return range is simply too wide to plan around. Bitcoin has produced both the best multi-year returns and the most brutal multi-year drawdowns of any major asset class in the same decade, and anyone who tells you with confidence what it will do over the next five years is guessing in a lab coat. Fidelity's own data shows Bitcoin was the top-performing asset in eleven of the last fifteen years, which sounds like a reason to load up until you remember the four years it was not, and how those years felt to live through.

So flip the starting question. Instead of asking what you hope to make, ask what decline you can sit through without your hands moving to the sell button.

Bitcoin has had several drawdowns past 50%, some past 70%, every one of them so far followed by a recovery to new highs. The investors who came out ahead are the ones who did nothing during those falls, which is far harder than it reads on a chart. The number that decides your sizing is the size at which a 60% fall still lets you sleep.

Put a rand figure on it, because abstractions never got anyone through a crash. Say you hold a R2 million portfolio and you are weighing a 10% allocation. That is R200,000 in Bitcoin. A 60% fall, which is an ordinary Tuesday in Bitcoin's history rather than a catastrophe, leaves you looking at R80,000. Sit with that screen for a moment before you answer the only question that counts. Would you hold, or would you quietly sell at the bottom and tell yourself you were being disciplined? If the honest answer is sell, the allocation is too big, whatever the model says it should be.

Size to your stomach, not your spreadsheet.

What the institutional range actually shows

The large allocators who have publicly added Bitcoin have mostly landed between 1% and 10% of relevant assets, and the detail underneath that range is more useful than the range itself. Fidelity's digital assets team ran the classic 60/40 portfolio through a decade of history and added Bitcoin at various weights. The most significant improvement in risk-adjusted returns, measured by the Sharpe and Sortino ratios, came from moving between roughly 1% and 3%.

A small tilt did most of the work. Adding more kept lifting the raw return but stopped improving the return per unit of risk at anything like the same rate, which is the quiet argument for restraint that the headline numbers tend to bury.

The drawdown finding is the one that surprises people. In a backtest running from the end of 2015 to late 2022, Fidelity took a baseline institutional portfolio and shifted 5% out of equities into Bitcoin. Over that window the compound annual growth rate rose from 6.26% to 16.72% and the ending balance came out 88% higher. The maximum drawdown of the whole portfolio moved from 19% to just 22%, despite Bitcoin itself suffering several 40 to 70% crashes inside that same period. Two things did that: Bitcoin's low correlation with the rest of the portfolio, and the plain discipline of rebalancing once a year, which quietly sells strength and keeps the weight from running away.

BlackRock has made the same point from the other direction, noting that even a 1 to 2% position can materially shift a portfolio's risk and return because of Bitcoin's skewed, asymmetric payoff. A recent Nomura survey of Japanese institutions found that 60% of those planning to add crypto intend to hold between 2 and 5%, which is roughly where Western allocators have settled too. None of these are prescriptions. They are a picture of how careful, well-resourced investors have answered the same question you are asking, and the picture clusters lower than most newcomers expect.

It is worth knowing what the aggressive end looks like so you can place yourself against it honestly. Fidelity ran Bitcoin's actual ten-year record through the Kelly Criterion, the formula that asks how much capital a favourable bet deserves. On raw historical numbers it spat out a position of 65%, which no sane person should act on. Feed it conservative forward assumptions instead, a 25% average annual return and a 50% drawdown, and it settles at around 10%. A separate optimisation for the maximum Sharpe ratio put the ideal Bitcoin weight at 9.4%, with nothing in bonds. I read those figures not as a target but as a ceiling: the maths tolerates far more Bitcoin than the average temperament can, which is exactly why temperament, not maths, sets the real limit. How that sits inside a full portfolio is a question I work through in Bitcoin's place in a diversified portfolio.

The South African rand context

South African investors carry a variable the offshore models never price in. Every rand-denominated asset you own, equities, bonds, cash, the house, is quietly exposed to the currency's long structural slide. An asset that merely holds its value in dollars will still rise in rand terms, because the rand keeps weakening underneath it. Bitcoin has a fixed global supply and answers to no decision made in Pretoria, so it behaves like any hard global asset and gains in rand terms as the currency gives ground.

That shifts the framing for anyone whose real fear is not a Bitcoin crash but a lifetime of exposure to one small emerging-market currency. For that investor a Bitcoin allocation is not only a bet on return. It is a currency hedge, sitting in the same drawer as offshore equity or dollar cash, and it can reasonably be weighted toward the higher end of the institutional range for that reason alone.

How much weight to give it depends on your read of where the rand goes over twenty years and how exposed the rest of your life already is to it. I compared the three routes side by side in Bitcoin versus offshore equity versus USD cash, and the honest conclusion was that most people want some of each rather than all of one.

The position you can hold beats the position you open

The costliest mistake I see is almost never getting the opening size slightly wrong. It is selling into a drawdown that would have healed if left alone. The correct size is therefore whatever you can carry for three to five years or longer without being forced to act, either by your bank balance or by your own nerves, which means you size for the ugly scenario rather than the brochure one. Assume it falls hard and stays down for a year or two, then ask whether the position at that point still leaves your emergency fund intact and your sleep undisturbed. If it does, the number is right. If picturing it makes you flinch, trim before you buy, not after it falls.

Here I lean on The Strategic Reserve, the book I wrote on holding Bitcoin as a long-horizon asset rather than a trade. Speculation is measured in price charts, reserve strategy in decades. The two demand different position sizes, because the trader can stomach a wipeout the reserve holder is trying to avoid.

Building the position gradually helps on both counts. Regular purchases at a fixed rand amount average your entry across whatever the price is doing and take the agony of timing off the table entirely, which is why I structure most clients into steady monthly accumulation from R1,000 a month rather than a single nervous lump. It also lets the size grow as your familiarity with the asset grows, so your conviction and your exposure rise together instead of your exposure sprinting ahead of your understanding. Where a lump sum is unavoidable, spreading the entry over several months does much the same job. I set out the full first-position process in your first R100,000 in Bitcoin, and the deeper mechanics of how a position moves against the rest of a portfolio in an honest calculation of Bitcoin portfolio risk.

Structure and custody are part of the size

Size does not exist on its own. Once a Bitcoin holding becomes a meaningful share of your wealth, three things that were irrelevant at R20,000 suddenly matter. The name on the asset, personal, company or trust, starts to drive your tax, your estate and how the coins pass to the next generation, and fixing that later is far more expensive than deciding it early. Getting the structure right before the position is built is cheap. Restructuring an appreciated holding is a tax event you inflict on yourself.

Custody scales the same way. A small position on an exchange carries a risk you might tolerate for a few thousand rand and would never accept for a few hundred thousand. I move clients toward guided self-custody from around R10,000 and toward the Vault for larger holdings, where no single key can move the coins alone and control stays with you. Very large or once-off positions can go over the counter at any size.

The size question and the how-you-hold-it question are really one question wearing two hats. Answering one without the other is how people end up with a serious asset and a flimsy plan. Families running this at scale should read Bitcoin position sizing for a South African family office, where the same logic meets multi-generational governance.

Decide the size you can hold, then wrap the right structure and custody around it before it grows teeth. Book a Bitcoin sizing conversation and I will work through your actual numbers, not a generic range borrowed from someone else's balance sheet.