AI chip bubble is now bigger than the Nasdaq during dot-com frenzy
The artificial intelligence rally has already reached historic proportions and is now passing some famous – or rather, infamous – milestones. The SOX semiconductor index has a peak price that’s 62% higher than its 200-day moving average – more than double the spread of the Dow Jones Industrial Average in the run-up to Black Monday in 1987, as well as the lead-in to Black Tuesday in 1929, according to a Thursday note from Bank of America strategist Michael Hartnett. The spread is closer to the Nasdaq ‘s margin of 55% ahead of the dot-com crash in 2000 when the commercial Internet was first taking off and companies with no clear path to profitability got valuations in the hundreds of millions of dollars. It’s even in the range of the 73% spread in the French CAC All Tradable index prior to the bursting of the Mississippi Bubble in 1720. In that episode, shares of the troubled French colonial Mississippi Company were permitted to be used as legal tender, leading to a doubling of the French money supply. “Exponential price action, market concentration, collapsing vol, stocks bossing bond yields higher, why melt-up everyone’s new base case … Here we go,” Hartnett mused on Thursday. AI stocks started going parabolic at the end of March – a highly unusual contour in price charts for securities. Share of chipmakers Micron , Advanced Micro Devices , SK Hynix, Marvell , and Intel , among others, all exhibit this trend. Some economists are very confident that all the investment in AI — which multiple Wall Street banks think will exceed $1 trillion next year — represents a bubble. “Having to amass more than a trillion dollars in cash to support the investment … has led to what everybody talks about as a bubble,” economist Ann Pettifor, director of the Policy Research in Macroeconomics organization, told CNBC. The AI buildout versus other historical booms Not all commentators are impressed by the scale of the AI buildout, despite the big capex numbers. The Financial Times’ Robin Wigglesworth called it “a tiny little gnat on the arse of an elephant compared to the railway boom,” of the 1860s, which had much greater bond issuance compared to total AI debt when adjusted for inflation and scaled to GDP. “There was about $5 billion, $6 billion worth of bonds issued, and that doesn’t sound like much, but if you scale it relative to the size of GDP at the time, because the U.S. was a small economy, that’s the equivalent of $10 trillion today,” the editor of the FT’s Alphaville blog said on the podcast “Unhedged” earlier this month, citing an analysis from JPMorgan. Still others are acknowledging the likelihood of a bubble without sounding too worried about it. “The railroads were a bubble and they transformed America. Electricity was a bubble, and it transformed America. The broadband build-out of the late-1990s was a bubble that transformed America,” author Derek Thompson wrote last year in a column referenced in a note by Oaktree Capital Management cofounder Howard Marks. “It’s unlikely that AI will be the first transformative technology that isn’t overbuilt and doesn’t incur a brief painful correction.” Despite all the debt, and the fact that companies are keeping large swaths of it off their balance sheets using methods of conduit financing, actual AI revenues are in fact materializing. First-quarter cloud revenue for Alphabet jumped 63% annually, the company reported last month . Amazon’s AWS cloud unit clocked 28% revenue growth in the first quarter, compared to the year-ago period, with AWS segment sales reaching $37.59 billion. Microsoft’s cloud revenue increased by 40%, with its division that includes Azure reporting $34.68 billion in revenue for the fiscal third quarter. Those numbers came as a salve for the broader equity market, in which gains are being increasingly concentrated in semiconductors and AI infrastructure stocks, suggesting the current equity boom may have a weak foundation. Even as broader market measures like the S & P 500 have been surging since the end of March, the ratio of companies gaining to companies losing has been declining, according to a Wednesday note from Piper Sandler. “The advance-decline line shows a stark divergence as the SPX makes record highs, indicating that leadership has become more concentrated, namely in technology,” Craig Johnson wrote for Piper.
<
