The stock-to-flow (SF) ratio has historically applied to commodity markets. It is a measure of the quantity of that commodity currently held (the stock), divided by the amount of it produced each year (the flow), as a means to effectively measure the scarcity of something. The commodity with the highest SF ratio is gold. In contrast to industrial assets that use up existing inventories quickly, gold inventories remain large compared to the relatively small increase in new supply each year.

So why is this relevant to bitcoin?

The relative price stability that comes with a large inventory provides gold with a monetary element that is not apparent in lower SF ratio assets. As a result, the store of value aspect of gold outstrips its industrial demand. Most other commodities have a ratio of near or less than 1 (their stock is similar to or less than the yearly production). None are anywhere near that of gold.

Bitcoin is another example of an asset that has a high SF ratio, albeit in a different form. Bitcoin is the first scarce digital asset we have ever seen. The SF ratio of bitcoin is increasing over time as the block rewards reduce every halving cycle (210,000 blocks). This reduction in new supply inflation continues until the mining rewards end, which is estimated to be by 2140, at which point the supply becomes fixed. Satoshi defined bitcoin’s issuance from the beginning, so we can easily plot its SF ratio over time.

Measuring scarcity

To paint a picture of the pace of this change, the current SF ratio for bitcoin is approximately 27. This figure derives from blockchain data showing total bitcoins currently in circulation of 17.93m / current new annual supply of 657,000. It has already exceeded that of silver and nearly trebled since 2014. The next halving is expected by April or May of 2020, after which we’ll have mined 84% of all bitcoins, and its ratio will be closer to 56. Gold stands at around 62 (185k tons / 3k tons per year), so bitcoin is almost guaranteed to surpass the SF ratio of gold within the next 4-year cycle, and beyond the projected 2024 halving will be double that of gold. At that point, it would become the first asset to have a stock-to-flow ratio of over 100, arguably forming the best monetary store of value in history. From there, we launch into uncharted waters in terms of scarcity.

What does this mean for market value?

By overlaying the price action of bitcoin to date, we can see a correlation as it oscillates around the SF ratio, producing a model to measure the value of bitcoin over time. Work by analysts such as Plan₿ shows that the current price of bitcoin is near the value predicted by the model and indicates the potential of at least $55,000 after the 2020 halving (a market value of $1 trillion). If you follow the model over time, and this is where it becomes mind-boggling, it predicts a potential price of at least $400k after the projected 2024 halving and $3m after 2028. Therefore, somewhere between 2024 and 2028, bitcoin’s market value could exceed that of the US dollar, and the relevance of USD as a unit of account also gets called into question.

In statistical terms, while many factors influence price, the R2 (the coefficient of determination that measures how well the model replicates observed outcomes) is around 95%. The likelihood of the relationship between SF and value just happening by chance; therefore, is near-zero. All the other influencing factors only make up the remaining 5% – suggesting that they have a more significant impact on time than the cyclical model price. While previous price highs have deviated by multiples from the corresponding model value, they always oscillate back and continue the correlation (reverting to the mean that is the SF value).

As such models become increasingly understood, the question is to what extent the SF increase, brought about by halving events, become front-run in anticipation of the impact of the restriction of new supply. The last two halvings have seen price highs lagging those events. Increased understanding and expectations could, therefore, speed this cycle up, to the extent that the next peak could potentially even come in before the halving. Some would also argue that increased belief in this model could cause front-running of future halvings too, not just the next one. Added to that is by how much the model price is overshot. The last three price peaks have exceeded the model by 13, 10, and 3 times the model value, respectively. If this trend continued, we could expect the next price peak to overshoot to around at least $61k, though past performance is no guarantee of future results and it could perform better or worse than these levels. The evidence so far suggests that the extent to which price peaks overshoot this model declines over time, something which would correlate with general market maturity and liquidity.

Realistic expectations

No model is perfect, and anything can happen. The SF ratio for bitcoin shows continued goodness of fit, however. The fact that the unrelated gold and silver markets are so in line with the bitcoin model values for SF provides further evidence of its validity. The model has, thus far, proved a statistically significant relationship between SF and market value. Its correlation has a near-zero probability of occurring purely by chance, and there is an indication of a power-law relationship (every halving, the SF doubles, and market value increases 10x).

The ultimate test will come after the next halving. If we don’t see the projected level of $55,000 by at least the end of 2021, the model will likely break down. If we do, it will only fuel further confidence in the model for the years that follow.

It should be noted that the figures discussed in this article are from the initial model by @100trillionUSD. It is the most conservative model, with rounded parameters based on data up to the end of 2018. Since then, the parameters have been refined, nearly a year of data has been added, and data archeology from the earliest years improved. This leads to $60,000 – $90,000 levels, rather than the initial $55,000. There is also a third model, however, that is only based on pre-November 2012 data, before the first halving. This model successfully predicted the approximate 10x jump following the 2012 and 2016 halvings, therefore performing really well out of the sample. This third model would predict a $100,000 level beyond the next halving, and by the end of 2021 for the model to remain valid.

For all the noise, theories and analysis surrounding bitcoin, ultimately, it’s about supply and demand. As its scarcity increases with continued restriction of new supply, even stable demand would lead to an increase in price. Unless proven otherwise, the stock-to-flow ratio represents one of the best methods by which we can model this balance to date.

Disclaimer: Investing or trading in bitcoin involves significant risk. It is important to research and carefully assess any such investments. All the information presented in this article does not constitute financial advice or recommendations of any kind.

No Comments