Friday, September 11, 2:30 p.m. in Belgium. In Washington, the statistics bureau publishes consumer prices for the month of August, and that is the only time of the week when I look at crypto. Not because I like it, because that is where things are decided.

The figure came out at +0.4% for the month, after +0.1% in July, and at 3.4% over one year. The underlying figure, the one that removes food and energy to see the underlying trend, rose by 0.3% for the month while analysts were expecting 0.2%. Three tenths of a percent too much, and everything changes sides.

Why does that affect bitcoin, you may ask. You have every right to ask, and the answer is simpler than it looks.

When inflation refuses to come back down, the American central bank no longer has much room: it keeps its rates high, or raises them. Now these rates are those of an investment that everyone can buy, guaranteed by a state, and which today yields 4.94% per year at ten years. The thirty-year rate has gone above 5.3%.

Put yourself in the place of someone who has money to invest. On one side, a Treasury bill that yields almost 5% without moving an inch, on the other, a currency that can lose several percent overnight. At some point, the second option is no longer worth the risk. This is not an opinion about bitcoin, it is arithmetic, and this week arithmetic won.

The markets, for their part, did their calculation in one morning. For the American central bank's meeting on Wednesday, September 16, a rate hike was put at around 72% before Friday's release, and above 80% afterward, according to CME FedWatch readings.

And it is not only American. On Thursday, September 10, the European Central Bank raised its rates for the second time this year, bringing its deposit rate to 2.50%. Same direction, same effect on everything that earns nothing by itself.

The figure comes out at 2:30 p.m., and everything else follows

The result, now, and it is in the figures rather than in the commentary. Over the week of September 7 to 11, exchange traded funds on bitcoin recorded a net outflow of 462.73 million dollars. The day-by-day detail tells the story better: Tuesday the 8th, 46.65 million out. Wednesday the 9th, 120.24 million. Thursday the 10th, 282.56 million, the worst session of the week. Friday the 11th, 13.29 million. Monday was a public holiday in the United States, the stock market was closed.

Put that against the previous week: 986.9 million dollars were entering the same funds. So we went from +986.9 to -462.73 million in seven days, a shift of 1.45 billion dollars. These are not individuals who emptied their pockets, they are managers moving balance sheet lines.

Bitcoin, in the meantime, remained below 77,000 dollars for four sessions in a row. Ether is hovering around 2,480 dollars. And this is where the story becomes more interesting than it looks, because the money leaving bitcoin is not leaving crypto.

Funds on ether took in 216.41 million dollars on Friday alone, and around 197 million over the week. This is their fourth week of consecutive inflows. To put it simply: money is moving from bitcoin to ether, and both remain in the same universe. This is not an escape, it is a change of seat.

The 462.73 million dollars of the week, in one image

You have to keep a cool head, and I’m going to try to do the same. One week of outflows does not make a trend, and an inflation figure has never said where a price would go the following month. What these figures do say, but, is measurable: listed bitcoin funds still manage $97.58 billion, and according to CoinDesk, they remained about $1 billion below their break-even point since the start of the year. In other words, in 2026, the money coming in and the money going out almost cancel each other out.

There’s one thing that bothers me in everything I’ve read on the subject this week: most articles explain the drop by the funds’ outflows. That’s mistaking the effect for the cause. The funds are flowing out because rates are rising, not the other way around. If you take only one thing away from this article, let it be this one, it will keep you from chasing explanations that aren’t explanations.

And for those wondering whether they should buy or sell right now: I’m not going to say, and it’s not an evasion. I don’t know where this price is going, nobody does, and an article that gives you a figure for tomorrow morning is lying to you. What I can do is tell you what happened and when.

There’s a more useful article than this one if you’re just starting out, the one I wrote last February about the difference between investing and trading. It isn’t about the week, it’s about the method, and it’s probably the best place to start.

The only piece of advice from this article you can apply tonight

What does this actually change for you?

If you hold bitcoin, nothing has moved on your key, in your wallet or on the network. The network produced its blocks every ten minutes as usual, without wondering what the fund managers were up to. What changed is the price that other people are willing to put up against it.

What also changes, and this is more useful day to day, is the schedule. The American appointment is at 2:30 p.m. Belgian time, and in the seconds that follow, it’s machines reacting to the raw figure and not to its context. The movement in the first minute is very often erased within the hour. The next date is Wednesday, September 16, with the decision of the American central bank. It’s in your calendar now.

And the real change of the week is nothing spectacular, it’s a bar going up. A risk-free investment pays 4.94% per year today. An asset that moves a lot therefore has to do significantly better than that to deserve its place, and this comparison wasn’t even made a few years ago, when rates were at zero. Every time the bar goes up, assets that generate nothing by themselves are the first to suffer from it, and that’s exactly what we’ve just seen.

One last point that has nothing to do with finance. Turn off price notifications at night. A price moving at 3 a.m. teaches you nothing, earns you nothing, and keeps you from sleeping. The market will still be there tomorrow, with its 462 million and its 5% rates.

Have a good week. And if someone can explain to me why there are ten articles about the 462 million that went out and none about the 986 million that had come in the week before, I’m all ears.