The count from halving to peak works three times out of three, and it’s still not a law

How well does it work, exactly?

That’s 21 days between the shortest and the longest, on a wait of almost eighteen months. Put another way, if someone had written in May 2024 that “the peak comes around October 2025”, they would have been right to within three weeks. This regularity is precise, and the people who mock it have usually never sat down to count it.

I start here because this argument only works if it takes on the strongest case for the other side. The count works, so the real question is a different one: what does a count that works three times entitle you to do?

What can a sample of three carry?

And the three points aren’t even comparable, which is the bigger problem. In 2016 this market was worth a few billion and was made of enthusiasts. In 2020 there were the first funds and a pandemic flooding the world with liquidity. In 2024 there are exchange-listed products, companies with the coin on their balance sheet and regulators writing the rules. Calling three situations that different “the same experiment” is the part that doesn’t hold.

Then there’s the third variable, the most boring explanation and also the most likely. Global liquidity has a cycle of its own, and in all three cases it moved in a way that fits what the price did. If a third thing drives both, the coincidence between halving and peak is true and useless at the same time, like ice cream and sunstroke, which go together without one causing the other.

What the halving does, and what gets pinned on it

Everything else is inferred, and nobody has ever proved it: that the cut causes the rise, that the rise lasts about eighteen months, that the fall starts twelve or eighteen months after the peak, that the sequence repeats. Those are four different claims. The first is debatable, and the other three describe what happened and get passed off as rules for what will happen.

And one piece has changed, and almost nobody has updated for it: block producers no longer sell the way they used to. The flows from their wallets show them cashing in differently from past cycles: they fund themselves on the markets, hold, or sell ahead of time. If the only plausible causal mechanism has changed, the theory resting on it at least deserves a second look.

The halving reduces the issuance of new coins by fifty percent, and therefore the amount whoever produces blocks has to sell: those are checkable facts. That it causes a rise, that the rise lasts eighteen months, that the fall arrives a year later and that the sequence repeats are, by contrast, unproven inferences.

And the altcoins, which were somewhere else in the meantime

The interesting part is the sequence, more than the level. In classic bear markets bitcoin fell first and the others followed it more violently, but over the same stretch. This time the others had been falling for months while bitcoin held up: two markets on two different clocks, more than one market moving at two speeds.

The simplest explanation is that the money from the listed products and corporate treasuries went into bitcoin alone, while the rest were left relying on retail money, which pulled out. If that’s right, the single cycle is neither over nor alive. There simply isn’t a single cycle anymore, so asking where we are in it becomes a question with no subject.

What do I watch instead of counting the months?

Those limits need saying, because people selling onchain certainties never say them. A wallet isn’t a person. A transfer between two addresses owned by the same entity reads as a sale even when it isn’t, and over a single week that noise weighs as much as the signal. It works on big numbers and long horizons, as a category, and as a Tuesday horoscope it’s useless.

And above all of it sits the thing nobody wants to hear: none of these measures predicts anything. They tell you what state the market is in now, which is already far more than a countdown does. Knowing where you are and knowing where you’re going are two different things. Telling them apart is the whole difference between a method and a faith, and on faith this market is doing splendidly.

Put bluntly: the sum works three times out of three, and three times is exactly the number of observations that convinces everyone and proves nothing.