S2F model: The prediction that Bitcoin will reach $100,000 in May 2022 was "disproven". Will the magic continue?

Wu Shuo Author | Liu Quankai

Editor of this issue | Colin Wu

On March 22, 2019, Twitter analyst PlanB published an article titled "Modeling Bitcoin Prices with Scarcity". More than two years after the release of the S2F model, Bitcoin prices have been tracking the model's predicted prices with a high degree of accuracy.

(Image source: Twitter @dan_pantera)

The S2F model predicted that the price of Bitcoin would reach $62,968 on April 15, and the price of Bitcoin exceeded the predicted value on April 13. According to Binance data, Bitcoin hit a new high of $64,854 on April 14. Previously, the time between Bitcoin and the price predicted by the S2F Model was no more than 15 days.

Although investors occasionally raise doubts about the S2F Model, believing that predicting the market is impossible, the high accuracy demonstrated by the model is truly amazing.

However, this "law" came to an end in May. The S2F Model predicted that the price of Bitcoin would reach $74,474 around May 15. In fact, according to Binance data, the closing price of Bitcoin on May 15 was only $46,762.99. Since reaching its peak on April 14, Bitcoin has been falling all the way, and is currently working hard to return to the 40,000 mark, which is almost double the price predicted by the model on June 15.

This article takes a non-technical perspective and tries to use popular language to present the operating principles and limitations of the S2F Model as well as some of my own thoughts.

Stock to Flow Ration in Economics

In economics, SF is often used to measure the scarcity of goods, from which two formulas are derived:

① S2F (SF/StFR): Stock-to-Flow = stock/flow = commodity inventory/commodity circulation

② 1/SF = commodity supply growth rate

And the resulting associated definitions:

① StFR average value, showing the average height of the annual growth rate of supply. The larger the average value, the more suitable it is to serve as a currency.

② StFR range shows the expectations achieved by extreme outliers over a long period of time. The smaller the range, the more suitable it is for currency.

③ The median of StFR’s annual growth rate shows the average short-term volatility. The lower the median, the more suitable it is for currency.

In economics, in order to quantify the availability of money, the SF value of money is often compared with that of gold and silver to determine a reasonable value, as shown in the following table:

(Image source: Nowandfutures.com, Incrementum AG)

Taking gold as an example, its StFR value is about 67.1, which means it would take 67.1 years of gold mining to produce the current stock of gold in circulation. Silver's StFR is 74.3.

Plan B stated in the article "Modeling Bitcoin Prices with Scarcity" that the SF value of gold is 62 and the SF value of silver is 22. The high SF values ​​make them value stores. In history, both gold and silver have been circulated as currencies. In addition, it can be found that the SF value will change, and the absolute value of the SF value of silver has dropped significantly. In the article, Plan B also pointed out that BTC-SF is 25 (the total number of bitcoins issued as of the time of publication in 2019/the supply in 2018).

BTC: Stock-to-Flow Model

Once you understand the economic definition, the idea behind the BTC-S2F Model becomes clear - Bitcoin is considered comparable to value stores such as gold and silver.

Gold that circulates on the market generally needs to go through six links: exploration, mining, mineral processing, smelting, processing and sales. Among them, exploration, mining and mineral processing are time-consuming and costly. From the supply side, it is difficult to increase its supply on a large scale. Bitcoin is very similar. Its supply has an upper limit and a unique halving mechanism. Bitcoin production is halved every 4 years. By 2140, all 21 million bitcoins will have been mined. Mining also requires a lot of electricity and computing power. This makes Bitcoin a store of value comparable to gold because they are both relatively scarce.

As the scarcity of Bitcoin continues to increase, its price will also rise. This is the core principle of the BTC-S2F model created by Plan B.

Based on this principle, Plan B collected the monthly price data of Bitcoin in the past ten years from 2019, and used linear regression to fit the extracted data (a straight line) with the natural logarithm of the SF value as the independent variable and the Bitcoin price as the dependent variable, and found that the SF value has a positive correlation with the market price. Then, the fitting adjustment was combined with the Bitcoin halving cycle to form the S2F model that we are all familiar with, and this model was used as a prediction of Bitcoin prices.

The BTC-S2F Model, which is based on the BTC-SF value as the prediction result of Bitcoin price, is shown in the following figure (April 27):

(Image source: lookintobitcoin.com)

The price of Bitcoin will rise around the fitted S2F price line, either above it or below it. Plan B once said that if Bitcoin does not reach $100,000 per coin before 2022, the model will be invalid. But at least before May, the results predicted by the BTC-S2F Model are still quite accurate.

question

After understanding the basic principles of the model, I believe that most investors will have this question: Can a simple supply and demand relationship really serve as a reason for investment decisions? Because in the cognition of most people, the return of any asset is uncertain.

What investors complain about most is that the model ignores market factors, such as changes in mining power, investor sentiment, policy factors, national policy and regulatory supervision, etc. These factors are generally difficult to add to the model as quantitative factors, and the rise or fall of prices is often inseparable from the changes in these factors.

So let’s take a look at what happened in May that caused the Bitcoin price to gradually deviate from the price predicted by the S2F Model. The following figure (June 1) shows:

(Image source: lookintobitcoin.com)

On May 12, Musk announced that Tesla would no longer accept Bitcoin payments based on global environmental considerations, and the price of Bitcoin fell by more than 12% that day.

On the evening of May 18, the China Payment and Clearing Association and other organizations jointly issued a notice, clarifying that financial institutions and payment institutions are not allowed to conduct business related to virtual currencies. On May 19, the price of Bitcoin fell by more than 14%;

On the evening of May 21, the minutes of the 51st meeting of the Financial Committee mentioned the crackdown on Bitcoin mining and trading. In the follow-up, the prices have not improved due to negative factors such as the closure or relocation of mining farms in relevant regions and the closure of related businesses by exchanges.

In addition, the potential for the Biden administration to introduce an interest rate hike policy continues to ferment, domestic policies remain unclear, and factors such as excessive leverage, market panic and funding differences have caused the downturn to continue to spread in May.

It can be seen that once at a certain node, the S2F Model cannot fully take these factors into account due to excessive interference from external extreme news, major policies and internal factors such as high leverage and high speculative sentiment, and the predicted price will be distorted.

Some thoughts based on the S2F model

The SF value is introduced in economics to measure the scarcity of goods, while the BTC: S2F Model uses SF as a "reference" for prices. Looking closely, this seems to assert that the price of a commodity is directly derived from its gradually insufficient or scarce supply growth rate. A well-known theory is that in Marxist economics, the price of a commodity is determined by its value, and the price fluctuates around the value under the influence of supply and demand; the value of a commodity is determined by the socially necessary labor time to produce the commodity (this can explain the mining cost of Bitcoin, which is reflected in part of the price). From a theoretical perspective alone, there are certain shortcomings in using SF in reverse, because even for gold and silver, there is no evidence that their prices are directly determined by supply.

But in any case, as of May, the BTC-S2F Model has indeed demonstrated amazing accuracy. Although the actual price has deviated greatly from the predicted price, judging from past prediction trends, after deviation, the Bitcoin price will eventually return to the established direction predicted by the S2F Model. Based on Plan B's imagination, there is still half a year until 2022, so will history repeat itself?