A company lost 4 trillion yuan in market value in nine days. It’s the most valuable company in the world.
That company is Nvidia. At the start of the month, its market cap had just climbed to $5.76 trillion. By Friday, it was down to $5.16 trillion. Its biggest drawdown over those nine days exceeded 12%. I compared the figures reported by Bloomberg and Xinhua, and they match.
This isn’t just about one company. The Philadelphia Semiconductor Index fell for eight consecutive trading days, its longest losing streak since the financial crisis. Over the same period, U.S. semiconductor stocks lost nearly $2 trillion in total market value. The Nasdaq 100 closed down 5.5% this week, its biggest weekly drop since the tariff shock last spring.
The hardest hit was CoreWeave. This AI infrastructure company fell 20% over the first three trading days of the week, including a 15% drop in a single day. At its core, its business model is to borrow money to build computing capacity, then rent it out to pay off the debt. As long as interest rates don’t come down, the whole story remains under strain.
On the 16th, Jensen Huang said in Berlin, “This will all pass. AI will transform every industry and every country, and this is only the beginning.” When he said that, his company had just lost 4 trillion yuan in market value in nine days.
People in crypto shouldn’t treat this news as mere entertainment. AI trades have led global risk assets over the past two years, supported by three things: leverage, low interest rates, and a big enough story. Crypto has all three. The valuation logic for computing-power tokens is cut from the same cloth as that of AI infrastructure companies: neither has cash flow to fall back on; both rely entirely on the story.
So when one side contracts, the other is bound to contract too. Risk appetite is one pool; trading on Binance doesn’t mean the laws of physics suddenly change.
My view is that the bottleneck for AI trades right now is financing. Jensen Huang is still telling the demand-side story, but the bond market has stopped listening to the financing-side one. If capital spending gets reprioritized, the first things to be cut will be computing-power stories with no cash flow to back them up. The impact on crypto will be more direct than most people think. The valuation anchor for AI-themed coins has never been on-chain; it’s the Nasdaq.
There’s one thing I’m not sure about: whether the Fed will cut rates again this year. The bond market changes its tune every day, and China and Japan are still reducing their Treasury holdings in August. When it comes to interest rates, I can’t give an answer that even I find convincing.
The chips are still being made, and the story is still being told. It’s just that the people footing the bill have started doing the math.
$BTC $TAO $RNDR
#英伟达 #AI泡沫 #黄仁勋 #比特币 #纳斯达克
That company is Nvidia. At the start of the month, its market cap had just climbed to $5.76 trillion. By Friday, it was down to $5.16 trillion. Its biggest drawdown over those nine days exceeded 12%. I compared the figures reported by Bloomberg and Xinhua, and they match.
This isn’t just about one company. The Philadelphia Semiconductor Index fell for eight consecutive trading days, its longest losing streak since the financial crisis. Over the same period, U.S. semiconductor stocks lost nearly $2 trillion in total market value. The Nasdaq 100 closed down 5.5% this week, its biggest weekly drop since the tariff shock last spring.
The hardest hit was CoreWeave. This AI infrastructure company fell 20% over the first three trading days of the week, including a 15% drop in a single day. At its core, its business model is to borrow money to build computing capacity, then rent it out to pay off the debt. As long as interest rates don’t come down, the whole story remains under strain.
On the 16th, Jensen Huang said in Berlin, “This will all pass. AI will transform every industry and every country, and this is only the beginning.” When he said that, his company had just lost 4 trillion yuan in market value in nine days.
People in crypto shouldn’t treat this news as mere entertainment. AI trades have led global risk assets over the past two years, supported by three things: leverage, low interest rates, and a big enough story. Crypto has all three. The valuation logic for computing-power tokens is cut from the same cloth as that of AI infrastructure companies: neither has cash flow to fall back on; both rely entirely on the story.
So when one side contracts, the other is bound to contract too. Risk appetite is one pool; trading on Binance doesn’t mean the laws of physics suddenly change.
My view is that the bottleneck for AI trades right now is financing. Jensen Huang is still telling the demand-side story, but the bond market has stopped listening to the financing-side one. If capital spending gets reprioritized, the first things to be cut will be computing-power stories with no cash flow to back them up. The impact on crypto will be more direct than most people think. The valuation anchor for AI-themed coins has never been on-chain; it’s the Nasdaq.
There’s one thing I’m not sure about: whether the Fed will cut rates again this year. The bond market changes its tune every day, and China and Japan are still reducing their Treasury holdings in August. When it comes to interest rates, I can’t give an answer that even I find convincing.
The chips are still being made, and the story is still being told. It’s just that the people footing the bill have started doing the math.
$BTC $TAO $RNDR
#英伟达 #AI泡沫 #黄仁勋 #比特币 #纳斯达克