Stock-to-flow: what the ratio measures, and where it stops working

The ratio was popularised for Bitcoin by a pseudonymous analyst writing as PlanB, in a 2019 Medium post titled "Modeling Bitcoin's Value with Scarcity". The argument ran in two steps. First, that scarcity, expressed as stock-to-flow, is what drives the value of monetary goods. Second, that because Bitcoin's stock-to-flow rises on a predictable schedule, its market value should rise on a correspondingly predictable path. A regression was fitted between the ratio and market value, and the fitted line was extended forward to produce price projections.

The projections were striking, and they were widely repeated. They were also presented by their author as a model rather than a certainty, and the original post is explicit that the relationship is a hypothesis. That distinction was often lost as the idea spread. A fitted line through historical data is a description of the past; it becomes a forecast only if you assume the relationship is stable and causal, and that assumption is exactly what was in question.

The model's appeal was understandable. Bitcoin's supply is genuinely fixed, its issuance genuinely declines, and the ratio genuinely climbs. A reader who accepts that scarcity drives value will find the rest of the argument natural. The problem is the first step, not the arithmetic that follows it.

The most direct criticism is that the model confuses a supply property with a price explanation. Stock-to-flow describes how fast the quantity of a thing grows. Price is set by demand as well as supply, and the model contains no term for demand at all. Two assets with identical stock-to-flow ratios can trade at wildly different values, and the ratio alone cannot distinguish them. Scarcity is a necessary condition for a high price in some goods, but it is not a sufficient one: there are scarce things that are worth very little.

A second criticism concerns the regression itself. Fitting a line through a handful of halving cycles gives very few independent observations, and the observations are not independent of one another: each price point is part of a continuous series, and the series is dominated by a single long uptrend. A regression on autocorrelated data can produce a strong apparent fit that carries no predictive information. Statisticians call the general problem spurious regression, and it is a well-documented hazard when trending series are regressed on one another.

A third criticism is empirical. The model's forward projections were published with dates attached, and the price did not follow them. The projections implied price levels that the market did not reach on the stated schedule, and in the period that followed, the relationship between the ratio and market value broke down rather than tightening. A model that fails its own out-of-sample test is not evidence for the mechanism it proposes, however elegant its in-sample fit.

There is a subtler point about what the ratio can and cannot say. Even if scarcity did drive value, the ratio would still be a poor forecasting instrument, because it changes slowly and predictably while price changes quickly and unpredictably. A variable that moves on a fixed schedule cannot explain a series that moves on no schedule at all. The ratio is a fact about issuance; the price is a fact about the market. Conflating them is the error the criticism identifies.

What survives is the descriptive use of the ratio. Bitcoin's issuance rate does decline, and the ratio does rise, and both facts are worth knowing. The ratio is a legitimate way to compare the supply growth of one asset with another, and it is a useful entry point to the issuance schedule. Readers who want the underlying numbers rather than the ratio will find them on the issuance and inflation page, and the comparison with gold's supply is treated on its own terms on the scarcity versus gold page.

What does not survive is the use of the ratio as a price model. The fitted line was a description of a period, not a law, and the period ended. The honest summary is that stock-to-flow is a sound measure of supply and an unsound predictor of price, and that the two claims were bundled together so tightly that rejecting one was often mistaken for rejecting the other.