Home » The price is crashing. Has the S2F model failed?

The price is crashing. Has the S2F model failed?

by Thomas

Over the weekend, the prices of cryptocurrencies rattled down significantly. This not only pokes holes in bitcoiners’ wallets – but could also bury a model the scene likes to invoke. Or is it?

In bitcoin journalism, montages often begin not by preparing for what will happen this week – but by processing the weekend.

And this, the weekend now past, was a busy one on the bitcoin and crypto exchanges of the world. The Bitcoin price, which was still at a good 50,000 euros on Friday, plunged to below 40,000 euros for a short time and is now around 43,000 euros.

The bitcoin price in euros over the last 7 days according to Bitcoin.de

The bitcoin price in euros over the last 7 days according to Bitcoin.de


As usual, all other cryptocurrencies suffered along with Bitcoin, in some cases even more strongly. The total market capitalisation of all cryptocurrencies slumped from just under €2.4 trillion to less than €2 trillion, but is now just above that again.

Market capitalisation of all cryptocurrencies according to Coinmarketcap.com

Market capitalisation of all cryptocurrencies according to Coinmarketcap.com


The crypto markets reached their current all-time high of more than 2.6 trillion on 10 November. From there, it has gone down continuously ever since, which even shorter phases of recovery have not changed. Has the bear market now appeared after all, with a significant delay? Is the phase now beginning that many of us already know too well – a few years of falling or stagnating prices?

Maybe, maybe not. But instead of unwinding the pros and cons to this question, today we turn our attention to a statistical model that has significantly shaped the bitcoin scene and its expectations of future price in recent years: the stock-to-flow model (S2F).

A model to predict the price of Bitcoin

The S2F model actually comes from the commodities sector. It simply shows the ratio of the currently existing quantity of a commodity (stock) and the annually created quantity (flow). The stock-to-flow ratio is calculated by dividing the existing quantity by the annual new creation. It shows how many years one would have to produce in order to generate as many units as there are today. The more years, the “harder”, i.e. scarcer, a commodity is.

S2F became known in the Bitcoin scene through the economist Saifadean Ammous and his questionable book “The Bitcoin Standard”, which was published in 2018. The section on S2F is one of the strongest parts of the book. Ammous explains by ratio why gold, and not silver (or copper, platinum, plastic), has been the most stable money in history:

The higher the S2F value, the lower the “supply-side price elasticity”: it is true that gold miners ramp up capacity when the gold price rises, and they will also increase output – but this is largely irrelavant in relation to the existing quantity of gold. This is because gold has a much higher S2F value than silver or copper, making it less dependent on market cycles.

Bitcoin, Ammous mentioned, will have twice the S2F value of gold as early as 2025, and it will keep going up from there. If we apply the S2F formula to Bitcoin, we end up with something like this:

18892931÷(6.25×144×365) = 57.51

This puts Bitcoin not far from the S2F value of gold, which is usually estimated to be around 66. At the latest with the next halving, which will reduce the production of new Bitcoins per block to 3.125, Bitcoin will have a higher S2F value than gold. And as the creation of new Bitcoins continues to flatten, Bitcoin’s S2F ratio will continue to tilt towards infinity over time.

Less than a year after Saifedean Ammous brought the S2F ratio into the Bitcoin universe, Bitcoin trader PlanB formulated a formula to link the S2F ratio to the Bitcoin price. His formula, if you’re interested, was this:

ln(market value) = 3.3 * ln(SF)+14.6

His “model” is considered the most accurate modelling of the historical bitcoin price. Some have described its accuracy as “uncanny”. While the price has always temporarily deviated from the model, it has always, sooner or later, swung back to the line.


The S2F model seemed perfectly suited to predicting the future bitcoin price. As such, it became extremely well known – just about every Bitcoiner knows it, even Landesbank Bayern studied it – and PlanB became one of the most famous Bitcoiners, with more than a million followers on Twitter.

When model and reality fall apart

You cannot underestimate the power of such models. They can become powerful self-fulfilling prophecies – but they can also, when they fall apart, take an entire market with them. Nothing counts more on the stock market than faith.

Now, however, PlanB admits that his model no longer worked for the first time. His tweet sounds a bit like resignation – almost:

“No model is perfect, but this is a strong aberration and the first in 10 years. An outlier / black swan? I will give the floor model another month.”

The background to this is a forecast made by PlanB in June. After Elon Musk bemoaned Bitcoin’s energy consumption and China banned mining, the price had fallen sharply and the model that had accompanied the rally until then, almost congruently, lost touch with the price.

PlanB admitted the weakness but blamed it on fundamental and temporary reasons, predicting, “My worst-case scenario for 2021 is August ☻ 47k, September ☻43k, October ☻63k, November ☻98K, December ☻135K.”

As recently as the end of October, PlanB could cheer that its forecasts were working like clockwork. Bitcoin did move at the lower end of what was possible with the model. But the price still remained in the zone.

In November, even the most pessimistic interpretation of the model obviously failed, and not just by a narrow margin, but quite significantly: at around $49,000 today, the bitcoin price is just half of PlanB’s – pessimistic – forecast.

Does this mean the S2F model is dead? Does the modelling of the Bitcoin price by the concept turn out to be pseudo-scientific humbug? As a fairy tale made of numbers that had the primary purpose of spreading good cheer on the markets?

(k)a self-fulfilling prophecy

You could dismiss the discussion about the S2F model as noise in the market.

All the fundamentals are still true: Bitcoin is still the only truly scarce money there is; the fiat money of the world remains in a pitiful state; Bitcoin continues to spread around the world, both as a means of payment and a financial product. And so on. The S2F model is also fully in tact in itself, and continues to describe an important core characteristic of Bitcoin – only its relationship to price is ticked off.

But one should not underestimate the power of models in this. Models can hyponotise markets; they have the power to become self-fulfilling prophecies if only enough people believe in them. Or even if people no longer believe in them.

PlanB now wants to wait another month, as he announced, before admitting the failure of his model. However, he should not be misunderstood: He still thinks the S2F model is unscratched. “[The] S2F model not affected and still on course to $100K. Be on the lookout for trolls confusing Floor and S2F.” However, it is not entirely clear on what basis PlanB made his predictions if not on the S2F model, what values and theories underlie the Floor model, and why he continues to stick to S2F even though the bitcoin price remained well below the predictions.

Even earlier this year, PlanB announced he would abandon his model if the price did not reach six figures by the end of the year. Now, however, he says the price just has to somehow average 100,000, whether that happens this year or next.

A model is not invalid, one could say, when the numbers suggest it is – but only when the people behind the model admit it.

Related Posts

Leave a Comment