Three days of selling and a 3.7% decline in the NASDAQ 100 Index is all the correction we got! Given the magnitude and speed of the rally and the fact that the Iran war is not technically over, that seems remarkable.
That was it for the big May “selloff” – just 3.7% down.
Right now, as per the White House’s statements, it looks like there will be an official 60-day ceasefire with Iran, giving the warring sides time to hammer out a permanent deal. If hostilities do not reignite, bond yields should decline, and so will oil prices, which should be supportive of the stock market.
That does not mean the market will be straight up for stocks, though, as the rally has been fierce and some type of consolidation pullback should come this summer, it looks like it will start from higher levels.
Assuming the Iran war does not reignite, the big market event is the IPO of SpaceX, which is likely to take in $85 billion of fresh capital into the company and give it a market value of $1.75 trillion (at present estimates, although these amounts can change due to investor interest at the time of the IPO).
It could become a major market event, like an intermediate-term top, as $85 billion in sales of other securities to get into the IPO will move the stock market. I would not be surprised if the underwriters move the intake to $100 billion, but we have to wait for details on the offering, which is likely to come on June 12. In a best-case scenario, the rally continues for the next three weeks and could be substantial.
As expected, Elon Musk lost his lawsuit against OpenAI last week – on statute of limitations grounds, not the merits of the case – in what I have referred to as a clear-cut case of “sour grapes,” as there are emails to Sam Altman strongly arguing to take over OpenAI and make it a for-profit subsidiary of Tesla.
This is hilarious: We asked OpenAI’s chatbot to get the details on those emails that have been released by OpenAI in the public domain. Here is Elon Musk in his own words to OpenAI:
Some of the most cited excerpts were:
“Tesla is the only path that could even hope to hold a candle to Google.”
And:
“OpenAI needs to go for-profit.”
According to the released correspondence, Musk argued that OpenAI could not compete with Google unless it raised vastly more capital and adopted a for-profit model. OpenAI says Musk proposed either merging OpenAI into Tesla or giving him majority equity and control.
One email chain from 2017 reportedly included Musk saying:
“Either go do something on your own or continue with OpenAI as a non-profit.”
OpenAI gave us a link to the full email release, not just those excerpts. Elon Musk may be brilliant with self-driving cars, giant rockets shooting huge amounts of satellites into orbit, satellite internet and his ambition to colonize the Moon and Mars, but when it comes to OpenAI, he has a “sour grapes” problem.
OpenAI itself is coming up for an IPO with a market value of around $1 trillion. Close behind is an OpenAI offshoot called Anthropic, with a private market value of about $800 billion. By the time those two IPOs debut, their market value could be over $2 trillion – the accomplishment of their CEOs, Sam Altman and Dario Amodei, who cannot stand each other, even though they used to work together for years. Earlier this year, everybody was holding hands at the end of an AI Summit in India with Indian Prime Minister Modi in the center, but those two refused to join hands (center, near the prime minister).
OpenAI’s CEO warned two years ago that we would have issues with memory and CPU shortages, because the more GPUs are deployed (the AI Chips that Nvidia is famous for), the more CPUs are used.
It used to be that when GPU clusters were smaller, the ratio in the market was that for every eight GPUs deployed, there was demand created for one extra CPU (8:1). Because of the demand for the management of huge GPU clusters, the demand for CPUs has exploded, and now the ratio is 1:1.
The huge surge in CPU and memory demand, as more and more GPUs are deployed, is causing shortages in both CPUs and memory chips. Computer memory makers have gone parabolic, but so have CPU markets, causing Intel, which a year ago was scraping multi-year lows, to go to an all-time high, with the U.S. government taking a 10% stake just before the surge began.
It is very simple, since the spending on data centers is accelerating, the rally in semiconductor stocks and the NASDAQ 100 Index is simply not over, even though when stocks and sectors go parabolic, they are prone to violent corrections.
Right now, in the newly developing, more benign geopolitical environment, the reason that makes sense to trigger a more meaningful correction is the SpaceX IPO, which may end up sucking $100 billion from the rest of the stock market, but until then, the rally may turn out to be explosive.