Bitcoin risk and volatility

The chart above shows the shape of the annual record. The figures below measure it. Every one is derived from the same curated datasets the rest of the site publishes — the calendar-year returns, the drawdown and recovery table, the highs and lows, and the daily candle series — and each is labelled with the window it covers. The method is stated with each figure so a reader can check the arithmetic rather than take it on trust.

The measured figures are loading from the curated datasets. The annual-return chart above and the drawdown record on the drawdowns page remain available while they load.

Realised volatility is backward-looking. It is computed from the actual price changes that have already occurred, usually as the standard deviation of daily returns over a window, annualised so that figures from different windows can be compared. A realised volatility of eighty per cent means that the asset's daily moves have been large enough that a year of them would spread outcomes across a range of roughly that width. Bitcoin's realised volatility has spent most of its history between fifty and one hundred per cent, against roughly fifteen to twenty per cent for a broad equity index.

Implied volatility is forward-looking. It is backed out of the prices of options, which are contracts that pay off if the price moves in a particular direction by a particular date. Because an option is worth more when the future is more uncertain, the market price of options reveals what participants collectively expect volatility to be. Implied volatility is therefore a forecast, and like any forecast it can be wrong. When it sits above realised volatility, options are expensive relative to what actually happened; when it sits below, they are cheap.

The gap between the two is itself informative. Bitcoin's options market has historically priced in more volatility than the asset subsequently delivered, which is consistent with a market that pays a premium for protection. That premium is not free money for the seller of options — it is compensation for taking on the risk that the forecast is wrong in the expensive direction. For a reader trying to understand the asset, the useful takeaway is that volatility is traded, priced and anticipated, not merely observed.

The first reason is size and liquidity. Bitcoin's market capitalisation is large in absolute terms but small relative to the pools of capital that can move in and out of it. When a modest allocation decision by a fund or a single large holder changes the balance of buyers and sellers, the price has to move further to find a new equilibrium than it would in a deeper market. Thin order books amplify every order.

The second is the absence of a cash-flow anchor. A share can be valued against the earnings it produces and a bond against the interest it pays, which gives those assets a reference point that limits how far price can drift from fundamentals. Bitcoin produces no cash flow. Its value rests on the expectation that someone else will want it later, and expectations of that kind are far more sensitive to news, sentiment and narrative than a discounted earnings stream is.

The third is the market's composition. Bitcoin trades around the clock across venues with different rules, different levels of oversight and different access for retail and institutional participants. Leverage is widely available, and forced liquidation of leveraged positions turns an ordinary price move into a cascade. Several of the sharpest single-day declines in the record were driven less by new information than by the mechanical unwinding of borrowed positions.

The fourth is regulatory and structural uncertainty. Bitcoin's legal treatment differs by jurisdiction and continues to change. An exchange closure, a policy announcement or a court decision can alter who is able to hold the asset and on what terms, and markets price that uncertainty as volatility. This is a risk that has no analogue in the equity market for a large, established company.

Bitcoin's volatility is not constant, and it has generally fallen as the asset has matured. In the early years, when the market was a handful of venues and a small number of participants, daily moves of twenty per cent were unremarkable and annualised volatility regularly exceeded one hundred per cent. As liquidity deepened and a wider set of holders emerged, the amplitude of the swings narrowed. The trend is not monotonic — volatility spikes in every crisis — but the peaks have been lower and the troughs shallower than in the preceding cycle.

Volatility also clusters within a cycle. It is highest during transitions: the sharp advance into a peak, the violent decline that follows, and the capitulation that ends a bear market. The middle of a long advance or a long decline is comparatively calm. This clustering is a general property of financial markets, not a Bitcoin peculiarity, but Bitcoin's amplitude makes it unusually visible. A reader looking at a chart of realised volatility will see bursts rather than a smooth line.

The practical consequence is that any single volatility figure is a statement about a window, not a permanent property. A number computed over the last thirty days and a number computed over the last three years can differ by a factor of two, and both can be correct. When a figure is quoted without its window and its vintage, it is not telling the reader very much.

Volatility is the risk that gets measured, but it is not the only risk a holder faces, and it is not the one most likely to cause a permanent loss. The risks below do not show up as price swings, and no amount of historical volatility data speaks to them.

Custody risk is the risk of losing access to the asset rather than losing value in it. Bitcoin has no issuer and no customer service line. A holder who controls their own keys is solely responsible for them: a lost seed phrase, a corrupted backup or a compromised device can destroy the holding irrecoverably, and there is no chargeback and no recovery process. A holder who leaves the asset with an exchange substitutes that risk for counterparty risk — the risk that the exchange fails, is hacked, or freezes withdrawals. Both failures have occurred repeatedly, and in each case the loss was total for the affected holders regardless of where the price went afterwards.

Regulatory risk is the risk that the rules governing the asset change in ways that affect a holder's ability to buy, sell, hold or move it. Different jurisdictions treat Bitcoin as property, as a commodity, as a security or as something else entirely, and those classifications carry different tax and reporting obligations. A change in treatment can make an otherwise sound holding expensive to liquidate or difficult to transfer. This risk is not hypothetical: several jurisdictions have restricted exchange access, and tax authorities in many countries now require detailed reporting of disposals.

Concentration risk is the risk that a holder has put more of their net worth into a single volatile asset than they can afford to see fall by eighty per cent. The drawdown record shows that declines of that size have happened. A position sized so that such a decline would force a sale — to meet a margin call, to cover living expenses, or to satisfy an emotional limit — converts a temporary drawdown into a permanent loss. Position sizing is the control that actually addresses this, and it is a decision made before the decline, not during it.

Technological and protocol risk is the risk of a flaw in the software, a failure of the mining network's security assumptions, or a change in the protocol that the holder did not anticipate. Bitcoin's design has proven remarkably robust over more than fifteen years, and its conservatism is deliberate. But robustness observed over a short history is not a guarantee of robustness over a long one, and a holder should understand that the asset's security rests on assumptions — about honest mining majority, about cryptographic hardness, about the incentives of participants — that are reasoned rather than proven.

None of these risks can be measured with a standard deviation, and none of them appear in the volatility figures quoted above. They are the reason a reader should treat volatility as one input to a decision rather than the whole of the risk picture.

A volatility number is only meaningful with three things attached: the window it was computed over, the frequency of the observations behind it, and the date it was calculated. A figure described as "Bitcoin's volatility" without those qualifiers is closer to a slogan than a measurement. Two analysts can quote eighty per cent and forty per cent for the same asset in the same week and both be right, because one used a thirty-day window during a crisis and the other a three-year window spanning a calm period.

It is also worth remembering what volatility does not measure. It treats upside and downside moves as equally risky, which is not how most holders experience them. It assumes returns are distributed in a particular way, which financial returns famously are not — extreme moves happen more often than a normal distribution predicts. And it says nothing about the direction of the next move. A high volatility figure tells a reader that the range of plausible outcomes is wide. It does not tell them which end of that range they will get.

The most useful habit is to pair a volatility figure with the drawdown record. Volatility describes the texture of the price path; drawdown describes the worst point on it. Together they give a reader a sense of both how bumpy the journey has been and how far it has fallen, which is a more complete picture than either measure alone.