Could Bitcoin Reach $1 Million by 2033? How Reliable Is the Power Law Model?
According to Gate Market Data, Bitcoin was trading at $77,793.3 on November 4, 2026. Its market cap is about $1.57 trillion, down 38.31% from the all-time high of $126,080.0 recorded on October 6, 2025. Over the past 30 days, it has fallen 7.98%, while over the past 90 days it still shows a 21.65% increase. Over the past year, it has recorded a 29.45% decline. Market sentiment is marked as "neutral," with a total supply of 20.0161 million BTC. These figures sketch an asset caught between a deep drawdown and long-term divergence. It is precisely within this divergence that a mathematical relationship discussed for more than a decade has returned to the spotlight: the Bitcoin power law. The forecast it offers is that the Bitcoin price will approach $1 million around the end of 2033.

Source: Gate Market Data
The power law is not an invention of the crypto industry. It describes one quantity scaling with another by a fixed exponent. On a double-log chart, it appears as an almost straight line. Similar relationships date back to 17th-century astronomical research by Johannes Kepler. The key figure who applied this framework to Bitcoin price is the physicist Giovanni Santostasi. After researching gravitational waves for years, he first noticed in a 2014 Reddit post that when Bitcoin prices are plotted on a logarithmic scale, the trend lines appear unnaturally straight. In June 2026, that observation received an official academic endorsement: the paper Bitcoin Power Law, co-authored by Santostasi and Stephen Perennod, was published online in Nonlinear Science, a journal under Elsevier, and passed independent peer review. The paper uses 5,696 daily closing Bitcoin prices from July 2010 to February 2026. It reports that a single power-law curve can explain about 96% of the variation in long-term logarithmic prices, with an exponent of approximately 5.69. Across different settings, Santostasi himself has given exponent ranges of 5.8 to 5.9 and said the correlation coefficient reaches 0.97.

What keeps this number on people’s minds isn’t its precision—it’s its simplicity. Bitcoin’s "age"—its time in existence measured in days or years—is raised to a power close to 6, then multiplied by a constant. Together, they outline a trend line that spans more than a decade of extreme volatility. The model’s economic story is similarly straightforward: as network participants join at an accelerating but diminishing rate, network value compounds as well. The price trend, in this view, is the mapping of adoption dynamics—not just pure speculative noise.
96% explanatory power doesn’t mean 96% predictive power
The "beauty" of a power law when describing history must be strictly separated from its real ability to predict the future. The 96% explanatory power applies to in-sample fitting: the model draws a line that matches extremely well to the 5,696 data points that have already occurred. But that’s entirely different from the claim that "there’s a 96% probability the future price will fall within the model’s predicted range." A model can fit the past exceptionally well while placing no meaningful constraints on future direction.
The power-law model’s real stress test comes at the cycle peak level. Historically, the degree of deviation between Bitcoin bull-market tops and the power-law trend has narrowed systematically. In 2013, the peak was about 11 times the trend value. By 2025, that multiple had shrunk to about 1.2 times. Supporters interpret this as evidence that Bitcoin is maturing and volatility naturally decays. Critics argue the shrinking peak multiple can also mean the model is systematically overestimating the future growth slope—the trend line itself may be "too high," not that prices are "too deviant." This dispute cannot be settled by historical fitting quality alone. The model must be tested in the next cycle outside the sample.
The methodological criticisms go even further. A methodological critique published in April 2026 points to identifiable issues with part of the derivations in the Santostasi and Perennod paper. The curve derivation is, according to the critique, "post-rationalization" done after measuring the exponent, not an independent ex-ante prediction. In the Bayesian stability analysis, the rolling estimates of autocorrelation are treated as independent samples, which may artificially compress posterior uncertainty. The paper also omits a Granger causality test between price and the number of addresses, so the causal direction of "network adoption driving price" is not established statistically. These criticisms do not necessarily overturn the value of the power law as a descriptive tool, but they pull it back from being a "predictive law" to a "descriptive benchmark." It tells you where the price might land if the last 15 years of growth structure continues in some form. It does not tell you why that structure must continue.
What sits between the current price and the model target
Comparing today’s Bitcoin price with the model target of $1 million in 2033 implies an annualized path that isn’t extreme. The issue isn’t arithmetic. It’s the structural conditions.
The first condition comes from changes at the supply margin. After the April 2024 halving, Bitcoin’s annual new issuance is about 164,250 BTC—roughly 0.82% of circulating supply. After the next halving in 2028, annual issuance will drop further to around 82,125 BTC, or about 0.41%. The halving supply shock shrinks in both absolute and relative terms. That means if the power law’s "scarcity-to-adoption" compounding mechanism is to continue working, the force from marginal demand must take over as the dominant driver as marginal supply contraction weakens.
The second condition comes from a shift in demand structure. In Q3 2026, U.S. spot Bitcoin ETFs increased net holdings by 79,800 BTC, worth about $6.2 billion. That is close to double the newly mined supply of 41,300 BTC in the same period. ETFs have replaced corporate reserves as the main marginal buyer in the market. Broader institutional holdings—ETPs and publicly listed companies’ balance sheets combined—have already exceeded 2.7 million BTC, more than 16 times miners’ annual incremental production. The implication is two-sided: it does provide a channel for long-term capital inflows that earlier cycles lacked, making an "adoption-driven" narrative more credible mechanistically. But it also exposes Bitcoin’s price rhythm more to external financial variables such as macro liquidity, portfolio rebalancing, and subscription/redemption behavior in regulated products. The power-law model assumes a smooth trend endogenously determined by network age, while real prices are increasingly being shaped by exogenous financial factors.
The third condition concerns the power law’s own falsifiability. Santostasi explains deviations from the trend line as "oscillations" around a deeper trajectory. He argues the power law should be understood as a highway with parallel "lanes," where the price runs within the parallel bands on either side of the center path, and that this does not violate the model. At the descriptive level, that framing is flexible. But it also creates an epistemic difficulty: if deviations can always be reclassified as "running within the lane," then what scenario would actually count as falsifying the model? Any framework that cannot be ruled out by any observable price path must necessarily have limited predictive value. For readers using the power law as a long-term valuation reference, a more honest way to use it is as a "conditional trend benchmark." When price is significantly below the trend line, it suggests the historical growth structure may be getting repriced. When price is significantly above the trend line, it suggests accumulated pressure for mean reversion. It is not a standalone price target you can act on.
Conclusion
The appeal of the Bitcoin power law lies in how it uses ultra-minimal math to capture a real regularity in Bitcoin’s 15-year price history: growth is happening, but the rate of growth is slowing. The projection to $1 million in 2033 is an extrapolation of this rule under unchanged assumptions—not a promise about the future. The current price in November 2026, the 38.31% drawdown from the historical high, and the post-2028 halving leading to an annual issuance share sliding to about 0.41% together paint a reality that is more complex than any single curve. Supply-side mechanical scarcity is weakening, demand-side institutionalization is deepening, and the "network age" variable on which the power law relies is now competing with macro-financial conditions to define the price’s rhythm. A 130-year-old mathematical law can offer a powerful starting point for long-term thinking, but it cannot replace ongoing testing of adoption structure, capital flows, and the model’s falsifiability.
FAQ
Q1: What is the core formula behind the Bitcoin power-law model?
The power law expresses Bitcoin price as proportional to a power function of its existence time (measured in days or years), with the fitted exponent concentrated in the 5.7 to 5.9 range. On a double-log chart, this appears as a line with a slope close to a constant. A peer-reviewed paper published in 2026 by Santostasi and Perennod is based on 5,696 daily price data points from July 2010 to February 2026, and reports that this curve explains about 96% of the variation in long-term logarithmic prices.
Q2: How is the $1 million prediction for 2033 derived?
The number is the power-law trend line extrapolated along the time axis. Using the current fitted parameters, when Bitcoin age reaches around 2033, the trend line implies a price level close to $1 million. It’s important to be clear: this is a model output under the condition that the historical growth structure continues—not a probability-style price guarantee. Santostasi himself also frames it as an estimate assuming continuous inflows of capital and institutions.
Q3: What is the biggest controversy around the power-law model?
The core controversy is the gap between in-sample fitting and out-of-sample prediction. Critics note that the narrowing of the peak deviation multiple can be explained by asset maturity as well as by overestimation of future growth by the model. A 2026 methodological critique also identified issues such as post-rationalization in the derivation and untested causal directions. Supporters, meanwhile, emphasize that the model’s passage into a peer-reviewed scientific discussion in 2026 is itself a sign that it has entered a falsifiable framework rather than remaining only a community-chart narrative.
Q4: How should the relationship between the current Bitcoin price and the power-law trend line be understood?
The current price is down 38.31% from the historical high in October 2025. Under the power-law framework, the trend line is not an exact "fair value" line. It’s a long-term benchmark that moves over time. Prices below the trend line can reflect cyclical pullbacks, tighter liquidity conditions, or a phase-by-phase repricing of the growth structure. Prices above the trend line suggest the accumulation of deviation mean-reversion pressure. The model itself does not provide a short-term direction call.
Q5: How should ordinary readers position the power-law model?
The most prudent framing is as a "long-term valuation context provider," not a "price target generator." It helps explain the structural feature of Bitcoin’s declining growth rate over time and offers a reference for assessing where the price sits relative to its historical growth trajectory. Any use that directly equates the power-law output with a future achievable price goes beyond what the model’s methodology can support.
Share

Bitcoin Dominance Rises to 60.7% as Altcoin Season Index Falls to 29: Has the Market Trend Reversed?

Gate Web3 BountyDrop: Participate in the ShopinX Airdrop and split $10,000 SPX in value

