Bitcoin outlook, 2026-08-25

There is something reliably funny about Bitcoin: every few weeks a new model arrives to predict the price, with an accuracy said to be far better than everything before it. What makes it interesting is that the people publishing these are usually not amateurs. Last week it was Samson Mow, with OMEGA60 and a headline number attached: $1 million per bitcoin in February 2031. The number is not the interesting part. The machinery that produces it is, so it is worth opening the hood.

The model starts from an observation that is simply correct. When you compute a conventional compound annual growth rate, the answer depends enormously on where you put the endpoints. The study's own table makes the case better than any critic could: measured from 2014 to 2022, Bitcoin compounds at 39.9 percent a year. Measured from 2018 to 2022, it compounds at 3.6 percent. Same asset, same history, two completely different answers, and the only thing that changed was which years you chose to stand on. Anyone quoting a long-run Bitcoin growth rate without showing their window is telling you very little.

OMEGA60's proposed fix is to replace the compound rate with the median of annual returns. Across 2014 to 2022 that median is about 59.7 percent, rounded down to 60, which is where the model gets its name. Extend the window through 2025 and it rises to roughly 77 percent, so 60 is presented as the conservative choice. The trouble is that a median is not a compound growth rate, and the two cannot be swapped.

Take an asset that returns 100 percent, then 100 percent, then loses 75 percent. The median of those three years is a gain of 100 percent. An investor who held through all three is back precisely where they started, with nothing. The median tells you what a typical year looked like. It tells you nothing about what an investor actually ends up holding after several of them, because compounding cares about the order and the depth of the bad years, and the median is built to ignore exactly that. Bitcoin's history is unusually full of the deep drawdowns a median discards. Using that figure as the engine of a multi-year projection is the model's central move, and it is the one that does not hold.

The genuinely interesting idea is the Terminus. The argument is that beyond some level, measuring Bitcoin in dollars becomes progressively less meaningful, because a rising number would reflect two things at once: Bitcoin's own adoption, and the weakening of the unit doing the measuring. The example reached for is gold under the Weimar Republic, where the metal had not suddenly changed its nature; the mark was collapsing underneath it. That is a real question, and it is more thoughtful than most of what surrounds price forecasting.

The threshold, though, is arbitrary. The Terminus is defined as the point where Bitcoin's market capitalisation reaches half of gold's, which on the study's assumption of $31.5 trillion for gold works out to roughly $785,000 a coin at today's supply, reached in August 2030. Nothing establishes that the dollar becomes a less meaningful unit of account at that specific figure rather than at half or twice it. And there is a tension the model never resolves: it argues that dollar measurement starts losing its meaning around $785,000, and then announces a precise target of $1 million a few months later. If the ruler stops working at the Terminus, the number past it cannot be read with more precision, not less.

One more thing raises an eyebrow. The study notes that Stock-to-Flow broke down after 2021, describing its correlation as having failed. It then places Stock-to-Flow's own May 2028 date in a convergence table alongside its February 2031, presenting a five-year window across several models as mutual corroboration. You cannot impeach a witness and call them to the stand on the same page. Convergence between models that share the same assumption, that the past return distribution is the best guide to the future, is not independent confirmation; it is the same bet counted several times.

Underneath all of it sits the thing no model solves. Bitcoin's supply is close to perfectly predictable, which is exactly why it is the part everyone models. Demand is not predictable at all, and no formula tells you in advance how many people will want to own this, or at what price, or what will change their minds. Every projection of this kind is an extrapolation of past demand wearing the clothes of a supply argument.

None of which makes OMEGA60 worthless, and that is not the point of reading it closely. A debatable model examined carefully still leaves you with something: here it is the question of which unit you are measuring in, which is real and which outlives whatever number gets stapled to it. That question is worth more than the target, and it would survive even if the target never arrives. Meanwhile the market did what it does: Bitcoin traded above $80,000 on Tuesday after a high of $81,265, a fourth day of gains. Our three horizon views are unchanged, and a price model published by someone else is not a forecast input for us.

This is general market commentary, not investment advice or a recommendation to buy or sell any asset.