It is not often that prominent market analysts representing differing views voluntarily sit down to discuss bitcoin. However, this is exactly what analysts PlanB and Benjamin Cowen did, bringing together their opposing views.
PlanB, known for its Stock-to-Flow model, and Benjamin Cowen, arguing for a model of lengthening cycles and diminishing returns, discussed the future of bitcoin.
The prominent crypto market analysts exhaustively presented the most important arguments in favor of their BTC models. Moreover, they found many common threads, agreeing that more of their models are similar than different.
PlanB, Cowen and bitcoin
The two cryptocurrency and bitcoin market analysts began arguing for the models they presented in December 2020.
In a tweet, PlanB was critical of the hypothesis of lengthening cycles and diminishing returns. Cowen responded immediately, claiming that PlanB was criticizing a position that he himself was implicitly promoting.
Over time, the Twitter thread and individual YouTube contributions evolved into the idea of a collaborative discussion. This finally happened recently on Cowen’s channel.
PlanB has become one of the most popular long-term crypto market analysts. This is thanks to his Stock-to-Flow (S2F) model, which he created and announced in 2019.
With 25 years of experience in traditional finance, he says he lacked a way to model the value of bitcoin. So he took an idea from the world of commodities, and gold in particular. Moreover, he found the same idea in Saifedean Ammous’ famous book titled “The Bitcoin Standard.”
Benjamin Cowen holds a bachelor’s degree in mathematics with physics and a master’s and a doctorate in nuclear engineering.
Since the beginning of 2021, he has taken up cryptocurrency market analysis full-time. He runs his YouTube channel and the premium group “intothecryptoverse.” BeInCrypto recently ran an extensive interview with Cowen.
The basics of the S2F model
Created by PlanB, the famous S2F model for bitcoin is based on the ratio of BTC stock to its annual production (flow).
The relationship between these two quantities is used to determine the scarcity of an asset. The advantage of the model is that it can be used to compare BTC to other commodity classes such as gold, silver, diamonds, or real estate.
In the chart below, the X-axis represents scarcity on a scale of 1-100, while the Y-axis expresses market capitalization on a very large scale from $100,000 to $100 trillion. A logarithmic scale has been used to maintain the legibility of such a large range of orders of magnitude.
The colored dots and the orange lines connecting them correspond to the closing price of BTC at the end of the following months. The colors mark the time until the next halving. The chart also includes the silver (silver), diamond (white), gold (gold), and real estate (red) markets.
Discussing the fundamentals of the S2F model, PlanB said its main thesis in the context of the ongoing halving cycle is that bitcoin will reach the market capitalization of gold:
“Basic thesis of the S2F model is that when bitcoin’s scarcity jumped the last halving, from 25 to 50, it has the same order of magnitude as gold. The value should also grow towards the same market value as gold which is around 10T at the moment.”
PlanB adds that he has been monitoring the successive movements of BTC against the S2F model for the past 3 years. He admits that it has been an exciting journey.
According to him, the model can be used until bitcoin reaches real estate market capitalization. The analyst stresses that we don’t have any higher-order data. Therefore, it’s hard to speculate what might happen above this value.
Cycle peak after 2021?
Before he explains the details of his own model of bitcoin’s lengthening cycles, Cowen asks PlanB a question about the peak of the current cycle from an S2F perspective:
“The lengthening cycle model would predict that the market cycle peak in this cycle should be not in 2021. (…) Does the S2F model allow for this market cycle to peak beyond 2021?”
PlanB: “Yes, it does.”
The creator of the S2F model stresses that the main difference between his model and Cowen’s proposal does not come down to different predictions about the extreme valuations of BTC.
It is to average over the long term, rather than modeling a potential peak or bottom in the crypto market. PlanB says:
“One of the major differences between lengthening cycle and diminishing returns view and I should say the entire technical analysis view, and S2F model and S2FX model is that S2FX model is not at all interested or focused on peaks and bottoms. It’s a statistical view, and it’s looking at averages.”
The analyst then adds that all-time highs are interesting, but he doesn’t care about them in the S2FX model. The model is focused on the average price of BTC. Which is supposed to be maintained for a certain period of time.
Peaks and bottoms don’t matter
PlanB emphasized that peaks and bottoms don’t matter that much to large, long-term investors. A day or even a week near the ATH is not enough for such an entity to sell its entire holdings without moving the price.
The case is similar with absolute lows and price capitulations, which last for a short time and are usually based on a series of liquidations of long leveraged positions. The analyst concludes that “peaks and lows last too short to be meaningful to long-term investors.”
Cowen agrees, saying that the S2FX model is actually oriented towards the average price rather than trying to predict peaks or bottoms. He, therefore, questions his interviewer:
“When you look at this model it seems that it would allow for a lengthening cycle. But again that is not specifically what this model is concerned with. It aims to project the average price correctly. Is that correct?”
PlanB: Yeah, absolutely!”
PlanB, in a slightly humorous and self-critical tone, admitted that the S2F and S2FX models are inaccurate. However, he expressed hope that they are at least roughly true.
“It’s always a bit funny when people say the model is so accurate. In fact, the model is not very accurate (…), it is a very long-term model. I would rather be roughly right, than exactly wrong.”
$100,000 before Christmas
PlanB is quite open about the impossibility of predicting the exact peak of this bitcoin cycle. Still, he acknowledges that regardless of where the absolute top will occur, the average price of BTC for 4 years after the last halving should be $100,000.
“It wouldn’t be strange at all to have like $500 000 peak in the S2F model. To get a $100 000 average price in this cycle, so in 4 years after the halving, it should cross the 100 000 USD level somewhere.”
An interesting attempt to connect the assumption of breaking through $100,000 later in 2021 and the model of lengthening cycles is PlanB’s next comment. He argues that reaching this level in the coming months does not rule out a continuation of the bull market in the following year.
“I wouldn’t be surprised to see 100 000 USD before Christmas this year, and then an extension of the bull market towards 288 000 USD or maybe 500 000 USD in 2022.”
PlanB pointed out that he created the model in 2019 when the price was also strongly deviating from the white line of the model. It was a very similar deviation to the one we are currently seeing.
At that time, bitcoin’s price drop reached the $3000-$4000 range, and most analysts were predicting prices around $1000. The analyst pointed out that the S2F model’s price forecast of $55,000 was unimaginable at the time. Although the price fell well below the model, it later returned to it.
To illustrate the basis of his model of lengthening cycles, Cowen used a chart that PlanB posted on his Twitter profile.
The blue line displays the first, dark blue the second, and the red the current cycle of the BTC market. Looking at this simple chart, the suggestion is quite obvious that so far, each successive cycle has been longer (lengthening cycle) and provided lower yields (diminishing returns).
According to Cowen, if history is any indication, it can be argued that the current cycle will be longer than the previous two and that it will provide lower returns for investors.
This is not just the extrapolation of charts but also the fundamental reason for rising market capitalization. The larger the market cap of an asset, the more money it takes to increase its value.
“It just takes a lot more money to move the market. That’s why micro caps in altcoins can put in a lot higher gains because their market capitalizations are so much smaller.”
Declining ROI for bitcoin
Cowen’s additional argument for the hypothesis of lengthening cycles is the graph of returns (ROI, return on investment).
It takes into account one more period in bitcoin’s history before the first halving. According to the chart below, successive cycles have delivered ROIs of:
- 1st cycle: 700x from the bottom in 250 days.
- 2nd cycle: 500x from the bottom in 750 days.
- 3rd cycle: just over 100x from the bottom in 1100 days.
Cowen then points out that the bottom of the current cycle was $3,100. So at this point, the current cycle has delivered an ROI of around 20x in almost 1000 days. The analyst states:
“There is a higher chance than not that this cycle will ultimately extend further out than the last cycle. And I would say it would do that whether we want to measure from the bottom or from the halving.”
On this basis, Cowen provides his price prediction for the peak of the current cycle. In his view, a peak between $100,000 and $200,000 would be a more conservative expectation.
“I’m going to be cautious with the $100,000 – $200,000 range because this range would still technically be diminishing returns. In order for us to have diminishing returns as measured from the market bottom, we cannot go above $300,000. But as measured from the halving it means not going above 250 000 USD.”
Logarithmic regression and Stock-to-Flow
In an attempt to bring the two models closer together, Cowen presented a graph by PlanB that juxtaposes a standard S2F chart and a logarithmic regression line.
PlanB calls this line a time model. This graph shows another big difference between the models, which will be crucial around the next halving in 2024.
According to S2F, another spike in bitcoin’s scarcity will then take place. In the next 4 years, its consequence is to be an average price of BTC at the level of $1 million.
On the other hand, according to the logarithmic regression model, the price should stabilize in the $100,000 – $300,000 range during this time. If this happens, the next halving would not be that important for the valuation of bitcoin.
PlanB comments on these differences briefly:
“In the long run, if you’re a hodler, it doesn’t really matter, because both models expect a lot higher prices, and returns that cannot be made with other assets.”
The analyst explains that the above chart was his main model before he created S2FX. According to PlanB, there are several reasons why the model shown in this chart is valuable:
- Logarithmic regression does not tell you exactly when the price will rise. However, S2F gives an indication that increased movement is likely after halving.
- This model eliminates the time variable. PlanB prefers to use models without time that allow asset comparison (cross-asset).
- The model allows interpolation of data instead of extrapolation. Time models do not allow you to predict the future and compare prices with other assets. In a model without time, data from other markets can be used to create an STFX model.
- It emphasises the fundamental importance of halving dynamics.
Various colours of regression
PlanB asked Cowen about the mathematical formula by which his version of logarithmic regression is generated.
At the same time, he stresses that this formula is often unclear, and analysts have trouble properly justifying the chosen shape of the curve that the BTC price is supposed to follow.
Cowen explains the formula by which he calculates his curve. However, at the same time, he admits that it could be determined in many ways. At some point, an analyst makes an arbitrary choice that, long-term, can provide very different predictions.
To justify his choice, Cowen uses a model of two logarithmic bands, red and green. He emphasizes that:
“These two lines are completely different. I trust the green line a lot more than the red line.”
Next, he explains that the red band is based solely on three points, which marked the absolute peaks of previous cycles. The green band, on the other hand, is based on thousands of points. These are “non-bubble data” when nothing exciting was happening in the BTC market.
“The idea behind the logarithmic regression model is that the highest gains occurred earlier on in the lifecycle of an asset.”
Additionally, Cowen once again points out that this is closely related to the fundamental argument for increasing volume required to raise the asset price by the next order of magnitude.
In response, PlanB points out that his logarithmic regression (grey) matches all points throughout the history of the BTC price.
The future price of bitcoin
To conclude their discussion of the S2F and lengthening cycle models, both analysts admitted that they “realize that there are more similarities than differences.”
In his agnostic style, Cowen emphasizes that there is no single, infallible model that can accurately predict the price behavior of any asset.
“I don’t know if any of them are going to be exactly correct, and I don’t even know if there is a way to define them as getting exactly correct. I think they all provide value in some way. They can get people to see the market in a slightly different way. As long as those models are more or less holding up as time goes on, there are reasons to continue to use them.”
Looking for collective conclusions, PlanB, on the other hand, draws attention to the apparent similarities in both models to the bull market of 2013-2014:
“We both have overlapping views as we concluded earlier. There are differences in models. (…) But basically what we both say is that $64,000 is not the top, and we’ll go up from here before we go down. And it might look similar to 2013.”
PlanB said that in the next 6 months to 1.5 years would be very interesting to see which model proves to be more accurate in its prediction of the BTC price.
Cowen confirmed that this period is crucial but outlined a perspective where again, both models could be right. For this to happen, the average BTC price over the next 3-4 years should stabilize around $100,000.
Both analysts agree that the final conclusion about which of the two models will prove to be closer to the truth could only be provided after the next halving in 2024.