Investment scenarios explained

A scenario has three parts, and a statement that is missing any of them is not a scenario. The first is a set of inputs: a starting price, a horizon, a rate of change or a schedule of purchases. The second is a rule for turning those inputs into an outcome — the arithmetic that connects the beginning to the end. The third is the outcome itself, expressed as a range rather than a point. A scenario that names only the outcome, without the inputs and the rule, is a prediction wearing a scenario's clothes.

What a scenario is not is a claim about probability. Saying that a given set of assumptions would produce a particular result says nothing about how likely those assumptions are to hold. This is the most common confusion in the genre, and it is worth stating flatly: the arithmetic of a scenario is exact, and the likelihood of its inputs is unknown. A scenario built on a twenty per cent annual growth rate is not a twenty per cent forecast; it is a statement that twenty per cent, compounded, produces a particular number. Whether twenty per cent is a reasonable assumption is a separate question, and the record can inform it without settling it.

The second thing a scenario is not is a description of the path. Two assumptions can produce the same endpoint by very different routes, and the route is what a holder actually experiences. An asset that rises steadily to a given level and one that triples, halves and then recovers to the same level have identical endpoints and completely different histories. The drawdown record exists precisely because the endpoint conceals the path, and any scenario that reports only a final value is hiding the part that matters most to the person holding the position.

A point estimate is a range with the uncertainty deleted. When someone says Bitcoin will be worth a particular figure in a particular year, they have collapsed a distribution of possible outcomes into its most convenient member. The collapse is not dishonest in itself — a single number is easier to hold in mind — but it discards exactly the information a reader needs in order to judge the claim. A range keeps that information. It says not only where the middle of the distribution sits but how wide the distribution is, and the width is the part that determines whether the position is survivable.

The historical record makes the case better than any argument. Across the sixteen calendar years from 2010 to 2025, Bitcoin's annual changes run from a gain of several thousand per cent in 2013 to a loss of more than seventy per cent in 2018. Those are not the tails of a distribution that clusters near its mean; they are the distribution. A reader who is shown only the average of those years has been told something arithmetically true and practically useless, because no year in the record actually delivered the average.

A range also forces the assumptions into the open. To state a range you must say what varies across it — the growth rate, the horizon, the entry date, the contribution schedule — and naming the variable is most of the analytical work. A point estimate lets the author choose the most flattering assumption silently. A range makes the choice visible, and a visible choice can be argued with. That is the entire value of the exercise: not that the range is right, but that it shows the reader what the answer depends on.

The single most powerful variable in any scenario is the length of the holding period, and it works in a direction that surprises people who are used to thinking about risk in the abstract. Over a short horizon, the range of outcomes is dominated by noise: a single month of Bitcoin's price history can contain a move of thirty per cent in either direction, and the direction is close to a coin toss. Over a long horizon, the noise partially cancels and the range narrows relative to the total return — but it does not narrow to a point, and the endpoints of the range remain far apart in absolute terms.

The record shows both effects at once. A one-year holding period drawn from the calendar-year series has produced results from a several-thousand-per-cent gain to a loss of more than seventy per cent. The longest holding window the site publishes — eight years — has never produced a loss, but the difference between the best and worst eight-year windows is still measured in multiples rather than percentages. The horizon compresses the range; it does not eliminate it. Anyone who reads a long-horizon scenario as a guarantee has mistaken a narrower distribution for a certain outcome.

There is a second, subtler effect. A long horizon gives the asset more time to recover from a bad entry, but it also gives it more time to experience a bad episode. The market-cycle record shows that the completed cycles have run for years at a time, with recoveries measured from the peak rather than the trough. A horizon that is long enough to span a full cycle is long enough to contain one, and the holder experiences the whole of it. The horizon is therefore not simply a dial that reduces risk; it is a choice about which risks to accept.

If the horizon is the most powerful variable, the entry date is the most consequential one, and it is the variable that scenario framing is best at exposing. The table below sets out the opening and closing prices for the eight most recent completed calendar years. Read it as a description of the terrain rather than as a result: the spread between the strongest and weakest entry years in this window is wider than the total return of most equity indices over the same period.

The record is the only honest input a scenario has, and it is worth being precise about what it contains. Bitcoin has traded for sixteen calendar years. In that time it has produced five completed cycles, each ending in a decline, and the correction that followed the October 2025 peak — a fall of roughly thirty-six per cent — is still open in the dataset. The deepest decline in the record is the 2011 episode, when the price fell from $32 to $2 between June 2011 and November 2011 — a fall of more than ninety per cent.

The most recent peak in the series was set on 6 October 2025 at $126,200, and the year closed at $87,508.83 with a market capitalisation of $1.75T. Those two figures, taken together, are the clearest illustration of why a scenario must be a range. The same year contained an all-time high and a double-digit annual loss, and a reader who had been shown only one of them would have formed a completely different picture of the asset.

The spread between good and bad starting points is not a property of Bitcoin's early years alone. It persists in the modern record. An investor who entered at the opening price of 2025 — a five-figure sum — was buying at a level the market had never sustained, and the year that followed closed lower. An investor who entered at the opening price of 2010 was buying at a level the market never revisited. Both are in the published record, and neither is an outlier to be discarded. They are the range.

The distinction is grammatical before it is statistical. A forecast is stated in the future tense and asserts what will happen: the price will reach a level, the cycle will turn at a date, the asset will be worth a figure. A scenario is stated in the conditional and asserts what would follow: if the price grew at this rate, the position would be worth that. The first claims knowledge of the world. The second claims only the validity of an inference, and the inference can be checked immediately.

That difference has a practical consequence for how each should be read. A forecast can only be evaluated after the fact, which means it can be wrong for years without anyone noticing and cannot be falsified in advance. A scenario can be evaluated the moment it is stated, because its arithmetic is either right or wrong and its assumptions are either stated or hidden. This is why the editorial standards of this site rule out prediction and permit analysis: the first cannot be held to account, and the second can.

There is a third category that is neither, and it is the one most often mistaken for a scenario. A target price is a forecast with the reasoning removed. It names an outcome, omits the inputs and the rule, and derives its authority from the confidence of its author rather than from anything checkable. A scenario that has been stripped of its assumptions is a target price, and a target price is not a scenario. The test is simple: if you cannot say what would have to be true for the number to hold, you have been given a forecast.

Four questions will separate a scenario from a sales pitch, and they can be asked of any claim about Bitcoin's future in a few seconds. What are the inputs, and are they stated? What is the rule that turns them into an outcome, and is the arithmetic shown? What is the range, and what varies across it? And what would have to be true for the assumptions to hold? A claim that answers all four is a scenario and can be argued with on its merits. A claim that answers none of them is a forecast, and its confidence is not evidence.

The most useful habit is to ask what the same scenario would produce with a different start date. If the answer changes materially — and on Bitcoin's record it almost always does — then the start date is doing the work, and the scenario is a description of a window rather than a property of the asset. That is not a reason to discard it. It is a reason to read it as what it is: one path through a range, with the range left out.

The record on this site is published in full so that the test can be run rather than asserted. The yearly returns give the calendar-year spread, the historical data page carries the highs, lows and year-end capitalisation series, and the data sources page records where every figure comes from and how old it is. A reader who wants to run the arithmetic on their own assumptions will find the tools for that on ApexCalc.net, linked from the pages here.