Has the artificial intelligence boom gone too far, too fast? Traders recently have decided to sell first and find out later. Is this simple profit-taking or something more insidious?
In my last column, I examined how equity investors have bid up stocks in the AI area for more than a year. A feedback loop of higher earnings expectations in the group triggered higher prices, which reinforced even higher earnings. Looking under the hood, I questioned some of the accounting moves that have driven these earnings in the past.
The market is now looking for a 30% jump in earnings results versus last year as third-quarter earnings are underway. In many respects, the fate of the economy and both the stock and bond markets now depends upon what happens in the AI sphere. That may sound like an exaggeration, but consider this. Goldman Sachs’s head equity analyst expects more than half of third-quarter earnings to come from 10 stocks, with more than one-third of that coming from just two companies.
And what is happening in the stock market is also happening in the bond market. In the corporate bond world, AI-related bond transactions accounted for over 50% of all investment-grade issues as of the end of August. Appetite for these AI offerings has been so strong that, for example, a Microsoft bond issue was preferred over government bonds of the same duration.
Apollo Management estimates that the bond market will raise $1.8 trillion by the end of this year. This is part of why bond yields have risen as much as they have. Between the U.S. government’s need to finance our debt and AI companies in the private sector, the supply of newly issued debt has overwhelmed demand.
That doesn't include the “off-book” borrowings that Microsoft, Alphabet, Amazon, Meta, and Oracle have taken this year. Those obligations amount, at last count, to $1.65 trillion in hidden, off-balance-sheet debt. These include future obligations like long-term operating leases for data centers yet to be built, bulk GPU supply contracts, and joint ventures. Accounting rules say companies don't have to record these liabilities on their balance sheets until after the assets are delivered or go live.
And speaking of pros, guess who has become big players in both AI stocks and bonds as well as U.S. Treasury bonds—hedge funds. This is according to a recent International Monetary Fund (IMF) report. The IMF is sounding an alarm about the growing risk hedge funds represent to the global financial markets.
These risky, private investment firms are leveraged entities that have doubled their assets in the last 6 years to about $13 trillion. A little more than half that amount is borrowed money. They invest alike. They are all up to their ears in the same AI stocks, as well as the same Treasuries and AI bonds. Their share of the U.S. Treasury market is more than 12% right now.
These are not buy-and-hold money managers or pension funds. They borrow money and buy stocks and bonds. If the trade goes against them, they are quick to sell out. Usually, they do so en masse. In the first quarter, their combined assets under management, plus the face value of their derivatives, were estimated at $42.2 trillion. That leaves us with a highly concentrated market, dominated by a mercurial group of players. A large portion of the retail crowd backs them, and are far more focused on the narrative than the numbers.
That narrative, however, is real enough. The global macro picture on AI is clearly on track. Like the internet before it, AI is real and promises to transform society yet again in the years to come. That doesn’t mean the AI space, like the internet before it, can’t get over its skies on occasion.
I don’t think we will see a protracted decline like the dot.com bubble where 50% or more of those companies disintegrated in the early 2000s. Remember, it took years for the boom-and-bust cycle of the Dot-Com era to unfold. What worries me most is the degree of AI concentration risk on both sides of the bond-and-stock market equation. Between the hedge funds and the 50% weight of the big tech players in the S&P 500, let’s hope third-quarter earnings are better than expected.
Bill Schmick is a founding partner of Onota Partners, Inc., in the Berkshires. Bill’s forecasts and opinions are purely his own and do not necessarily represent the views of Onota Partners, Inc. None of his commentary is or should be considered investment advice. Direct your inquiries to his website at www.schmicksretiredinvestor.com. Investments in securities are not insured, protected, or guaranteed and may result in loss of income and/or principal.