Sophronepsis · research · 5 August 2026
The halving is one of the most quoted facts about Bitcoin, and it is often told wrong. Here is exactly what it does, with the real size of the mechanism put in euros. Including the two places where our own internal fact sheet told it wrong until this week.
“Every four years Bitcoin cuts the issuance of new coins in half. Less supply and the same demand: the price goes up. It's pure maths.”
You'll find this sentence in every video, every thread and every brochure. It has two flaws you can see without a single piece of data: it isn't every four years, and if it were pure maths the price would go up the next day. It doesn't.
Let's take it step by step, no rush. By the end of this you'll understand the mechanism better than most of the people who repeat it.
Bitcoin has no bank writing down who owns what. It has a public ledger. Every ten minutes or so, a page of that ledger is closed with all the transfers from those ten minutes. That page is called a block.
Closing the page costs electricity — a lot of it. Whoever closes it gets paid for doing so, and gets paid in newly created bitcoins. That payment is called the reward, and it is the only way new bitcoins are born. There is no other.
The halving is this and nothing more: every 210,000 blocks, that reward is cut in half.
| When | Reward per block | Days since the previous halving | Years |
|---|---|---|---|
| 28 November 2012 | 25 BTC | — | — |
| 9 July 2016 | 12.5 BTC | 1.319 | 3,61 |
| 11 May 2020 | 6.25 BTC | 1.402 | 3,84 |
| 19–20 April 2024 | 3.125 BTC | 1.439 | 3,94 |
Look at the last column. None of them came at four years. Three years and seven months the first. Three years and eleven months the last.
The reason is that Bitcoin doesn't count days. It counts blocks.
Every 2,016 blocks — about two weeks — the network checks how long those blocks actually took. If they came out faster than ten minutes, the difficulty of the work goes up. If they came out slower, it goes down. It's a thermostat.
The famous “four years” come from one calculation and nothing else: 210,000 blocks × 10 minutes = 1,458 days, which is 3.995 years. That's the theoretical figure. Reality has given 1,319, 1,402 and 1,439.
Always below. And always for the same reason: the thermostat arrives two weeks late. If more computing power joins the network during that fortnight, blocks come out faster than ten minutes for the whole fifteen days, and those minutes gained are never given back. The 2016 halving came four and a half months ahead of the theoretical figure because in those years the network's power was growing at rates of +275 % and +475 % a year.
Strong growth in computing power can bring the halving forward by months. A moderate drop only delays it by days, because the thermostat corrects it within two weeks. Bitcoin's clock runs fast easily and runs slow only with great difficulty.
This is where the story falls apart. If the halving moves the price through scarcity, that scarcity has to be large. Let's measure it.
After the 2024 halving, each block pays 3.125 bitcoins. There are about 144 blocks a day. That's 450 new bitcoins every day. Before the halving it was 900.
| Item | Value |
|---|---|
| New bitcoins per day, before the 2024 halving | 900 BTC · 57.1 M$ |
| New bitcoins per day, after | 450 BTC · 28.6 M$ |
| The selling pressure the halving removed | 28.6 M$ a day |
| What is traded in bitcoin on a normal day | tens of billions of $ |
| A whole day's issuance, over all the bitcoin in existence | 0,0022 % |
Valued at 63,446 $ per bitcoin, the price on 2 August 2026.
New supply is on the order of 0.1 % of what moves each day. It isn't an earthquake of scarcity: it's a scratch.
Picture it this way. If a city sells a thousand flats a day and the council stops building one, house prices don't even notice. That's the halving against bitcoin's daily market.
If scarcity pushed the price up, it would go up the day after. It doesn't. It takes between twelve and eighteen months to reach the top, and along the way it usually falls. This looks illogical. It isn't. What's illogical is the explanation.
Down to the exact block. Anyone can count it with a calculator. If the market works at all, that is already in the price before it happens. There shouldn't be a surprise the next day, and there isn't.
Those who close the ledger's pages see their income cut in half overnight, while the electricity bill stays the same. The least efficient sell their bitcoin reserves to pay it, and some go bust: in 2022 a mining company that had once been worth more than 4 billion dollars collapsed, and two others came close.
And it's the big ones that go bust, not the small ones. Someone with a machine at home unplugs it and waits — they have no debt, so they don't go bust. The big ones had borrowed to buy equipment and build facilities, often putting up bitcoins as collateral. What brings them down isn't their size: it's the debt.
In other words: in the months after the halving, miners' selling pressure rises. It doesn't fall.
A scarcity shock would be instant. What you see is something else: twelve to eighteen months of rising prices, euphoria, and then a long year of falling. That fits a completely different mechanism — it rises, draws attention, new money comes in, it rises more, borrowed money shows up, the last buyer runs out, forced selling arrives and it falls in a cascade.
The halving does not cause the cycle. It's the date on the calendar around which the cycle organises itself, because it's the one thing everybody watches at the same time. And from that comes the consequence that orders everything else we publish: if what the cycles share is the date and not the mechanism, the only thing you can expect to repeat is the timing, not the price.
The internal fact sheet for this strategy, written on 23 July 2026, says word for word: “Less new supply entering the market, the same or greater demand: price goes up. This is not opinion.”
Well, it was opinion, and a small one at that. 28.6 million dollars a day against tens of billions doesn't support that “price goes up”. We had swallowed the industry's story without putting a number under it, which is exactly what we criticise when others do it. The fact sheet is corrected with this document.
For the depth of the three completed winters, one of our documents says −86.63 % / −82.59 % / −77.73 % and another says −84.7 % / −83.4 % / −76.6 %. They're the same three winters.
The difference comes from measuring with different price series and taking the low of one day or another. Here we publish the second set, because those are the ones we measured ourselves on daily closes and they can be repeated. The first were quoted, not measured. And the lesson in method is worth more than the numbers: a quoted figure and a measured figure are not the same thing, even when the same house writes them.
| Winter | Top | Bottom | Drop |
|---|---|---|---|
| 2013–15 | 1.163 $ | ≈178 $ | −84,7 % |
| 2017–18 | 19.497 $ | 3.237 $ | −83,4 % |
| 2021–22 | 67.567 $ | 15.787 $ | −76,6 % |
| 2025–? | 124.753 $ | 58,559 $ so far | −53.1 % for now |
And the rises, measured from the floor of each cycle to its peak:
| Cycle | Rise from floor to peak |
|---|---|
| First | +10.086 % |
| Second | +2.946 % |
| Third | +659 % |
The two tables tell the same story from both sides: Bitcoin is normalising. It rises less than before and falls less than before. It stops behaving like a niche bet and starts behaving like a large asset — with everything good and everything boring that brings.
And careful reading the first row. A +10,086 % in a market where a few thousand people took part and which was worth less than a mid-sized company is not a promise of anything. It's a piece of archaeology.
In bitcoin almost everything is opinion. What it's worth, whether it's a bubble, whether it goes up or down: everyone has their own and none of them can be checked. The halving is the exception. It has been written in the code since 2009, it doesn't depend on anyone's mood and it can be calculated years in advance.
That makes it the one thing on this market's calendar that isn't up for negotiation. And that's why it's worth knowing exactly what it does: because on top of a verifiable fact, a story has been built that isn't. When someone tells you the halving will push the price up, you now know two things you didn't know before: that the mechanism really exists, and that its size — put in money, next to the volume that moves each day — is much smaller than it sounds.
It doesn't tell you whether to buy. It gives you something better: a date nobody can move, and the real measure of what happens on it. With that, you can listen to the rest without being impressed.
It's the question you always have to ask, and the answer here is uncomfortable:
This describes what has happened, not a forecast of what will happen.
The next halving is block 1,050,000. On 2 August 2026 the chain was at 960,753, so 89,247 blocks remain.
We put it in writing, dated today: we expect the 2028 halving to arrive before 12 April 2028.
It's a small prediction, and on purpose: it can be checked, it doesn't depend on our opinion about anything, and if it fails you'll see it. It would fail if the network's computing power fell in a sustained way — something that has never happened in Bitcoin's whole life, but which in 2026 has started to hint at itself because data centres are moving over to artificial intelligence. We'll come back to it right here.
The underlying bias here isn't about cryptocurrencies. It's in the head, and we all have it.
When an explanation is simple, has numbers and fits what we already wanted to believe, we stop checking it.
“Less supply, price goes up” meets all three conditions. It's short, it sounds like maths and it confirms what you wanted to hear if you already hold bitcoin. That's why it's been repeated for ten years without almost anyone putting the 28.6 million next to the billions.
We didn't do it either, until this week. It's written above.
We're people who measure. We got it wrong twice inside this very document — our fact sheet repeated the scarcity story without putting a number on it, and two of our papers give different depths for the same three winters — and both are told above, in full detail.
We'd like you to check these numbers. Really. Block heights, halving dates and daily issuance are public, and anyone can redo the calculation in ten minutes. If you find a mistake, a period that doesn't fit or an interpretation we missed, write to us and we'll correct it right here, in public and with your name on it if you want it to appear.
We don't publish to be right. We publish to be right more often, and for that we need more eyes than ours. A method nobody can break isn't a method: it's a belief.
hola@sophronepsis.com — everything is read and answered.
If you want to keep pulling the thread
This document is the way in. The ones that follow measure specific things: how alike the cycles are in time, what happens when you take four famous bottom signals and measure them one by one, and what it really costs someone not to catch the bottom. All published, free and with no sign-up: sophronepsis.com/informes.html
And if what you want is to learn to decide this on your own, the six-day course is open, with no sign-up, at app.sophronepsis.com/empieza.
And if you don't want this to stay a snapshot. A report is a measurement from one day. Inside the platform the follow-up does not stop: measurements are redone when new data arrives, each strategy comes with its test and its published failures alongside, and the mentor answers whatever you ask about this at any hour. Subscribe and follow it from the inside: sophronepsis.com/mentoria.html