Not So Bad – or Is It?
The first jolt, seventeen days after the question
Addendum to: Is It ’29 Again?
Addendum of 8 November 2026: Forty to One
On 20 October, in “Is It ’29 Again?”, we asked who is calling small savers into the market and who can hold out in a crash. Seventeen days later came the first jolt. And the day after, the herald arrived too, advising people to hold on.
I. The Thursday
In the night to 4 November it became clear that Trump’s Republicans had clearly lost the congressional elections. Trump conceded, prices rose, and on Thursday the Dax reached a new record high. On the same Thursday, 5 November, the American technology index Nasdaq 100 fell by 6.4 per cent. In a single day about 2.3 trillion dollars of market value disappeared. Nvidia lost just over 9 per cent, more than 400 billion dollars.
The reasons were in plain sight. Qualcomm had reported weak figures. Ever since Alphabet’s results at the end of October, investors have doubted whether the billions spent on data centres will pay off quickly enough. Inflation in the USA is above three per cent, and many expect an interest rate rise in December. That hits above all the companies that build their data centres on credit, among them Coreweave, which we described in October.
II. The voters as negative feedback
The trigger, however, was the election. In Texas the Democrat Gina Hinojosa defeated Governor Greg Abbott. She has promised to halt the construction of new data centres and to strip the large operators of subsidies and tax breaks. In Georgia the new governor wants to pause construction for now in order to study the effects on electricity bills. In Ohio only those who pay their own electricity costs are to be allowed to build.
The stock market reads this as a risk. It can also be read differently. In “Negative Feedback Returns” we described how voters steer back when a policy goes too far. Here they steer against an expansion whose electricity bill they pay and whose profits arise elsewhere. That is not an accident of democracy. It is its job.
III. The loop
Why a single election night can move so much is shown by a figure from the Bank for International Settlements, the central bank of central banks. From 2021 to 2025, 55.2 per cent of all investment in AI firms came from other AI firms. Nvidia gives money to firms that use it to buy chips from Nvidia. Each books revenue and value with the other. As long as all of them grow, it looks like strength. If one of them falters, they all wobble.
On top of that comes a mechanism every control engineer knows. Many funds work with fixed loss limits. If the price falls by a certain percentage, they must sell. Their selling pushes the price down further and forces the next one to sell. That is positive feedback. It amplifies what it is meant to brake.
For completeness we repeat what we disclosed in October: one of the two of us is a language model made by the company Anthropic. The F.A.S. names Anthropic as a further example of such circular financing and mentions the planned stock market listing. We cannot judge that independently, and so we leave it aside.
IV. The herald on Friday
On 6 November the F.A.Z. published a text from the Frankfurter Allgemeine Sonntagszeitung. It is called “Three reasons for the sell-off – and what gives hope now”. The three reasons are essentially described correctly. The central banks’ figure, however, it gives as “almost half”. It is more than half.
Then comes the hope. “In the long run AI will pay off on the stock market”, it says, and the evidence offered is Microsoft and Amazon, the winners of the internet bubble. Anyone with a savings plan should “carry on as usual”.
What the text does not say: after the bubble burst, Amazon lost more than ninety per cent of its value. Microsoft needed almost seventeen years to regain its 1999 high, the Nasdaq index fifteen. The many companies that vanished entirely at the time do not appear in the text. One only ever sees the survivors.
And then the same text contains a sentence that cancels everything else: “Money you depend on in the short term should never be invested in shares.” That is correct. But the small saver does not know whether he will need his money in five years. The lost job, the illness, the broken roof, nobody sees them coming. That was the question of our October text, and the herald does not answer it. He does not even ask it.
V. So it is bad after all?
One bad day is not a crash. The Dax lost far less than the Nasdaq, and nobody knows whether prices will rise again next week. We make no forecast now either.
But Thursday showed how the system is built. The firms finance one another, the funds sell by rule, and one election night in Texas is enough to move two trillion dollars. This is the system into which savers, even the smallest, are meant to put their money. When it bangs, the same newspapers advise them to stay calm.
What is bad is not Thursday. What is bad is that nobody asks who can sit out the next one.
Further reading: Is It ’29 Again? · Negative Feedback Returns · What Remained of the Show