vor 9 Monaten
Summary Bitcoin is rated 'Strong Buy' due to robust network value, user growth, and an attractive current entry point. Quantitative models forecast BTC trading at $182,892–$215,000 by early 2027, with current prices below even the lowest model estimate of year-end price. BTC’s correlation with equities has increased post-ETF approval, but its risk profile and diversifying benefits remain compelling. Key risks include potential institutional centralization, proxy challenges from ETF adoption, and existential threats like cryptographic advances. Key Findings As it will be discussed in the article, Bitcoin’s (BTC-USD) value is mostly based on its network value and user growth in the long term. Although in recent years, there have been signs that Bitcoin is affected by systematic risk factors just like the overall equity market, its risk is still mostly unique and thus adds great diversifying value to a portfolio. In addition, the results of the various quantitative price forecasting tools show that Bitcoin stands to offer substantial gains to its investors. Recent Developments Since its peak of ~$125 thousand on October 7 th , Bitcoin’s price has dropped by more than 45% as of February 10 th . It can be argued that it was caused by overall shaky market dynamics, overextension of the previous rally, the discovery of Epstein’s ties to Bitcoin , and overall negative risk sentiment around the asset. On February 6 th, it had seen a low of $60 thousand, and as of February 10 th, it had slightly recovered to ~$68 thousand. Quantitative Models Metcalfe Network Value Metcalfe’s law has been used to value telecommunication networks in the past, and many, together with the CFA Institute, have proposed it as a method to value Bitcoin . Metcalfe’s law uses “users” or “participants” in the network as a base of the model. As there is no single number that represents the exact number of users of Bitcoin, I have decided to use the Active Addresses data from Glassdoor, spanning from 2010 all the way to January 2026. Author, using data from Glassnode as of February 10th As can be seen in the above chart that shows the log-log relationship between the Metcalfe model value and Bitcoin’s price, there is a strong positive relationship. The regression model results in an r^2 of 0.86 (correlation of ~0.92), with a coefficient of 0.9211 and a constant intercept of -15.72, with t-stats well over 100 each. Normally, the process stops here, but for a more comprehensive valuation and forecasting, I have taken the process several steps ahead. Author, using data from Glassnode as of February 10th The chart above shows the log-log relationship between time (days) and active wallets. Again, it can be seen that the relationship is very strong, with an r^2 of 0.917 (correlation of ~0.95), a coefficient of 1.71 (t-stat of 250), and a constant intercept of -0.5461 (t-stat of -10.368). This effect, called the Power Law Adoption path, will also be the theoretical foundation for the power law pricing model, which is simpler and yet very powerful, that we will discuss later. As the user base is now easily forecastable in the future, and Metcalfe’s law is based upon it, Bitcoin’s price can be forecasted. First, some adjustments must be made. Author, using data from Glassnode as of February 10th Although above I had discussed that Metcalfe’s network value and Bitcoin price were heavily correlated, it can be seen that the base model with only one coefficient breaks over time. This can be attributed to Bitcoin’s overall adoption as a reserve asset by firms such as MicroStrategy (Strategy) (MSTR), Tesla (TSLA), and even the country of El Salvador, started purchasing . In addition, on January 10 th of 2024, the SEC approved spot Bitcoin ETFs , which resulted in large-scale institutional purchases. These two events changed how Bitcoin acts and thus how the model should value it. I have added two dummy variables, one in 2020 and one in 2024, to change the slope in these two years to more accurately reflect the real value of the network. As it is present in many financial assets, in this model, autocorrelation is present. Basically, positive forecast errors tend to remain positive for some time and vice versa. In the charts below, the autocorrelation and partial autocorrelation figures can be seen showing the severe and strong autocorrelation that is present. As most of it is from the 1 st lag (correlation of 0.9571), an AR(1) process is an adequate addition to the model. Author, using data from Glassnode as of February 10th In the above chart, the model’s forecast (model anchor, the green line) and the actual price can be observed. Although the price and predicted value diverge from time to time, they follow the overall trend and converge from time to time, which is exactly how a similar model is expected to act. Author, using data from Glassnode as of February 10th Based on the user forecasting methodology discussed above and the model constructed step by step, these are the results. The thick red line that can be seen is the expected path of Bitcoin according to the model, and the red bands represent a 99.9% statistical confidence interval. The predicted price for January 1 st, 2027, is $182,892, with the range being $99,238-339,961, and $274,369 with the range of $148,247-512,184 for January 1 st, 2028. The ranges are quite ample due to the volatility of Bitcoin and the choice of a large confidence interval, but it is quite helpful at giving an overall idea of the future to position ourselves in. Power Law Bitbo Based on a similar theoretical basis of the idea that Bitcoin users grow logarithmically, the famous and widely used Power Law is a simple log-log regression of time on Bitcoin’s price. ( More information can be found on Bitbo ). According to this model, on December 31 st, 2026, Bitcoin is expected to trade at $199,379, with a range of $70,742-683,258, and $273,432, with a range of $97,017-896,620 on December 31, 2027. This model has even larger ranges, but it is still very powerful and elegant through its simplicity and gives a good idea of the future probable outcomes. Stock to Income Bitbo I prefer the stock-to-income model to the classic and well-known stock-to-flow due to its incorporation of mining costs and sale pressures. Both models incorporate the halving cycles very well and are inspired by models used in commodities such as gold. ( More information can be found on Bitbo ). According to this model, on January 1, 2027, Bitcoin is expected to trade at $215 thousand with a range of $109-430 thousand and $409 thousand with a range of $204-817 thousand on January 1, 2028. This model has a slightly tighter range, but Bitcoin has had a history of trading slightly below and very much above the range. Model Results According to all three models, Bitcoin is expected to trade at substantially higher prices, and in all three models, the current price of ~$68 thousand is below the lowest price range for the year-end estimates. Thus, according to the models, the current price is an attractive entry for Bitcoin. Macro Variables Fed Net Liquidity Author, using data from Glassnode & FRED as of February 10th Author, using data from Glassnode & FRED as of February 10th One of the macro variables most followed by Bitcoin investors recently has been the Fed's net liquidity, but the numbers show us that it shouldn’t be the case. The two have a slight positive relationship, which is not statistically significant, with an r^2 of 0.003 (correlation of ~0.05) and a beta coefficient of 0.42 with a t-stat of 1.655 (p-value of 0.098). Although an argument would be that the effect would be lagged due to the monetary transmission mechanism, even with lags, a relevant statistical inference cannot be made. Here are the lags with the highest statistical significance: Lag (Weeks) Correlation T-Stat P-Value Coefficient 57 -0.0883 -2.4243 0.0156 -0.5459 61 -0.0808 -2.2112 0.0273 -0.4907 3 0.0732 2.0788 0.0380 0.5298 It can be seen that none of them are highly significant, and the one with the highest significance even has a negative coefficient. Thus, further analysis of this variable has no real value. 10-Year Treasury Bond Yields Author, using data from Glassnode and FRED as of February 10th Author, using data from Glassnode and FRED as of February 10th Surprisingly enough, the situation is almost the same for the treasury yields, too. The relationship between the yields and price is just slightly negative, with an r^2 that is virtually zero and a coefficient of -0.0012 with a t-stat of -0.16 (p-value of 0.988), which shows that the relationship is practically nonexistent. But the lag analysis has resulted in an interesting finding. Lag (Weeks) Correlation T-Stat P-Value Coefficient 80 -0.1719 -4.6993 0.0000 -0.3054 82 -0.0837 -2.2591 0.0242 -0.1478 45 -0.0801 -2.2142 0.0271 -0.1579 The most significant and substantial lag in the finding is the 80-week lag, with a coefficient of -0.3054. Meaning that a one percent change in the yield (not a percentage point increase in the yield itself) in the yield would lead to a 0.35% decrease in the price of Bitcoin 80 weeks later. Although the correlation of -0.1719 isn’t huge, it's still quite significant. 80 weeks ago, on July 29 th, 2024, the US 10-year bonds traded at a yield of 3.794%, compared to the current 4.16%. This would correlate to an overall negative movement in Bitcoin. But the reality is that this relationship doesn't seem theoretically sensible, as although the monetary transmission mechanism is slow, it shouldn't be this slow. In any case, it's better to keep it in mind as a peculiarity that may be real but is suspect. S&P 500 Author, using data from Glassnode and FRED as of February 10th Bitcoin and the S&P 500 have a positive relationship, with an r^2 of 0.067 (correlation of ~0.25) and a beta coefficient of 0.95 with a t-stat of 13.620 (p-value ~0). Although the relationship isn’t that strong, it's way stronger than any of the other macro variables analyzed above, it's statistically significant, and the effects can also be quite substantial due to the coefficient value and the volatility of the equity market. There isn’t any substantial result from the lag analysis; thus, I have decided to omit it. Author, using data from Glassnode and FRED as of February 10th The most interesting aspect of the relationship is that it’s getting stronger through time, as can be seen in the above 4-week rolling correlation over time. Although the dates are not very visible on the chart itself, the correlation peaks tend to occur during market drops. This is a phenomenon that can be seen in many markets, where correlations tend to approach 1 in times of market distress. Thus, this brings us to the next variable, which will help us better understand this relationship. Market Implied Volatility ( VIX ) Author, using data from Glassnode and FRED as of February 10th Author, using data from Glassnode and FRED as of February 10th The regression has an r^2 of 0.016 (correlation of ~0.126), a beta coefficient of -0.0896 with a t-stat of -8.07 (p-value ~0). Although the relationship isn’t as strong as the one with the S&P 500, it gives us some important information. As can be seen, the correlation between Bitcoin and VIX has been decreasing over time, pointing towards a stronger negative relationship, which is the inverse of what we see with the S&P 500, but in both cases, the meaning is that the relationship is getting stronger. What we can understand from here is that the same risks that affect the S&P 500, such as future uncertainty, geopolitical risks, and so on, have a similar effect on Bitcoin. Now, this is a lot more different than Gold, which has a well-known role of increasing in price in times of distress. This is a well-known phenomenon already, but the overall picture is that Bitcoin has a different risk structure than Gold. Although the same risks that affect the S&P 500 do affect Bitcoin, it must be remembered that the r^2 is just 0.016 with VIX and 0.067 with the S&P 500, meaning that the unexplained variance percentage is still 98.4% and 93.3%, respectively. Thus, pointing us towards reality has historically been a great diversifier. Author, using data from Glassnode as of February 10th Author, using data from Glassnode as of February 10th While the institutional acceptance of Bitcoin seems to have increased its correlation with the equity markets, it also seems to have decreased its risk. As can be seen in the chart above, the rolling 30-day standard deviation has decreased over time. Standard financial theory tells us that this should also lead to lower returns, which is also what we observe in the figure below of the former one. Although in the long term a lower risk, lower return structure will probably occur, in the medium term, like the next 2-3 years, the attractive low correlation structure and still ample potential of adoption growth can be expected to bring in much more capital. Thus, also with the quantitative models in mind, Bitcoin is an attractive asset to buy right now. Risks Although quite improbable, if the ties between Epstein and Bitcoin turn out to be way deeper, and if the “conspiracy” that some have thrown out that he may be Satoshi himself does indeed come out to be true, it could be catastrophic for Bitcoin. There are also several more conspiracies about Satoshi being the CIA or the NSA, but again, they’re conspiracies. Although the model based on Metcalfe’s law and the power law model are based on the logarithmic growth of users, it's getting harder and harder to find a proxy for it. This is due to ETFs and third-party custody becoming increasingly more common. The current number of active addresses (~640 thousand) is way below its peak in April 2021 of 1.35 million. Although it's not too big a problem right now, this may lead to the valuation of Bitcoin getting harder. Bitcoin may struggle to “compete” against other cryptocurrencies, as many of them have more uses, and thus could result in Bitcoin not being the preferred store of value. Although this seems quite improbable due to the legacy of Bitcoin. Finally, the biggest risk of Bitcoin could result from its institutional adoption. As more and more institutions hold Bitcoin and even acquire the mining facilities and thus the voting rights of the network, Bitcoin may lose its decentralized nature. There is the risk that the supply of Bitcoin could technically be increased, or a fork of Bitcoin could be created that becomes the institutional favorite, which has a more easily manipulable supply. Again, due to institutional adoption, most prominently the financialization of Bitcoin, where ETF shares, futures contracts, swaps, and loans create many claims on the same Bitcoin, thus synthetically increasing its supply. This may lead to increased volatility in times of market distress, flash crashes, and even loss of faith in the mission of Bitcoin. Additionally, there is the risk that quantum computers or AI may be able to crack Bitcoin’s cryptography. Although it's not well understood how that may be the case, and if that is actually doable, how it would even work, and if something can be done about it, it is a real risk that should be kept in mind. Finally, any sale from Satoshi’s wallet may lead to market distress. Concluding Remarks Bitcoin has been one of the most interesting developments in the 21st century due to its enigmatic nature and its hard-to-understand nature. But to many people who live in tyrannical countries, Bitcoin has been a door to freedom. Apart from that, its acceptance from institutions and even governments has put Bitcoin on the track to becoming one of the reserve assets of the financial world. In this article, several valuation methodologies have been proposed and used to forecast its future performance, and several macro variables have been used to show how Bitcoin even acts and to understand its standing in the financial world. The conclusion is that Bitcoin offers a great return opportunity while also acting as a diversifying asset in a diversified portfolio. And for these reasons, I rate Bitcoin as a “Strong Buy.”