Artificial Intelligence Might Also Mean Artificial Growth

Probably the best way to describe the current market is resilient. We are about to enter the second week of the first major government shutdown since 2013, and equities remain unfazed, even as gold has surged past $4,000 an ounce for the first time.

The jobs markets have clearly started to falter in recent months. The August Bureau of Labor Statistics jobs report showed only 22,000 positions created during the month and the BLS also revised original job estimates for June and July sharply lower. The drama in D.C. kept the BLS from issuing its initial September assessment last Friday as scheduled, but investment firm Carlyle Tuesday estimated that only 17,000 jobs were created last month. This was a great deal better than the ADP September jobs report that came out last Wednesday. It projected 32,000 private sector jobs disappeared last month.

Artificial intelligence, meanwhile, is powering most of the growth in the economy and the markets. Over 60% of all venture capital raised over the past 12 months has gone into AI-related investments. This is crowding out other opportunities in the economy. Since late 2022, when ChatGPT first appeared, some 75% of the returns in the market have come from the companies in the AI vanguard like Nvidia Corporation (NVDA) , Amazon (AMZN)  and Alphabet (GOOGL) .

Lately a lot of this rise appears to be powered by the return of vendor financing, a core driver of the Internet Boom. This can be a virtuous circle, until it isn’t. Oracle (ORCL)  recently surged following a somewhat tepid quarterly earnings report. The trigger for this massive rise was a huge increase in Oracle’s backlog. Primarily from tens of billions of dollars a year pledged by OpenAI. Money OpenAI will likely have to raise itself in the future. Despite a $500 billion valuation, the non-public firm will generate only a projected $13 billion in revenue this year and that is bleeding cash.

And this is for capacity Oracle hasn’t built yet and recently partially financed via a $18 billion bond deal. In addition, this new capacity will take the output of four average nuclear plants to supply electricity to operate. What return on investment these massive investments eventually generate is indeterminate and opaque. What possibly could go wrong?

Meanwhile, the rest of the economy is not faring well. After a brief rebound into and just after the election, the Leading Economic Indicators index is declining again. The U.S. residential real estate market, the largest asset class in the world, looks heading into another difficult year in 2026. Outside the top 10% of income producing households that account for roughly half of consumer spending, the consumer is really struggling. Delinquency rates are rising for many loan types and student loans are looking like a train wreck after payments were restarted this year. The jobs markets seem like they are flat lining and the impacts here from AI are just starting to be felt.

I am more convinced than ever that the overall market is in an Internet Boom-type bubble. The hard part is discerning if we are closer to the 1996 "irrational exuberance" part of this bubble or near the late 1999 stage just before everything went to hell in a hand basket as my later father liked to quip. Therefore, roughly a quarter of my portfolio remains allocated to short-term treasuries with the rest going to covered-call positions around the few stocks I am finding in this market still sporting reasonable valuations.

At the time of publication, Jensen was long AMZN