| Home | About | Archives | RSS Feed |
The Retired Investor: Corporate profits better be good
Third-quarter corporate profit results kick off this week. A lot is riding on what company management tells Wall Street. Both the stock market's present and future fortunes depend on what they say.
It has been an extraordinary year for revenues, profits, and profit margins for many companies listed in the benchmark S& P 500 index. In the second quarter alone, earnings results were 50 percent higher than they were last year. This time around, investors expect a further 29 percent to 30 percent increase over 2025. That’s good news and makes great headlines, but the truth is that a relatively small handful of stocks are responsible for such stellar results.
None of us should be surprised that artificial intelligence is the engine of growth behind this remarkable performance. And while these technology stocks will generate the lion’s share of earnings, they won't generate all of it. You see, while companies like Micron Technology and Nvidia benefit greatly from the AI infrastructure buildout, AI is boosting business for just about everything from construction companies to copper miners.
One caveat the reader should note, however, is that most companies on the receiving end are booking the $166 billion in tariff refunds occurring now (due to the illegality of Trump’s first round of tariffs) as profit. In a past column, I predicted that little, if any, of this money would be returned to the consumer (who paid 90+ percent of these tariffs). As a result, these refunds are partially responsible for record profit margins throughout the economy.
As readers know, I use the S&P 500 Index as my benchmark when analyzing the market’s fortunes, since most other markets key off its performance. Thanks to AI, the technology sector now represents about 40 percent of the index market capitalization. However, neither Meta nor Amazon, two of the largest hyperscalers, are included in that index. If you include them, the AI-related weighting of the S&P 500 is almost 50 percent.
But let’s get back to earnings. In the second quarter, the top contributors to the massive earnings gain were Amazon (+242 percent) and Alphabet (300 percent). To bank those gains, a lot of people would have had to spend most of each day online buying everything Amazon offered. As for Google, I’m not sure it would have even been possible- without an accounting twist. For both companies, much of their profit came from huge, non-operating, unrealized gains on their private equity stakes in other tech companies (read: AI darlings Anthropic and SpaceX).
How would corporate results have fared without that accounting treatment? Second-quarter earnings would have dropped from 47.7 percent to 28.8 percent if you excluded just these two companies’ profits from these investments. Rest assured, they weren't the only two companies that used this accounting convention. I’m sorry to get all wanky on this, but that’s not the only accounting smoke and mirrors involved in the AI investment boom.
General accounting practices dictate that a company's wages and other expenses are deducted from its sales. What is left over is profit. However, large capital expenditures are treated differently. An AI company can plunk down $300 billion for data center or three, but it doesn't subtract that cost from revenue. Instead, it's listed as property on the company’s balance sheet. As such, they recognize the costs only over time as the value of this property declines due to wear and tear, etc.
And now the plot thickens. At the same time, all these billions of dollars are being spent on infrastructure; the Nvidias of the world that sell semiconductor chips, GPUs, CPUs, and the like to big hyperscalers like Microsoft and Alphabet get to book their sales and profits immediately.
Bottom line: the huge AI infrastructure price tag is dramatically understated and pushed into the future. That accounting standard softens the impact on earnings, while the profits of the various AI spenders are booked immediately and are overstated.
Together, these accounting principles create a virtuous circle: more AI spending guarantees higher sales and profits for companies. That translates into higher prices for the individual stocks, both private and public. In the case of private company shares held as investments by some hyperscalers, more unrealized gains can be booked as income.
How does this merry-go-round end? This bubble will burst, but only after investors and Wall Street analysts alike start looking under the hood and not just the headlines, do the work, and start coming up with their own conclusions. It took years during the early boom-and-bust cycle of the Dot-com era for this to happen. In this case, some early calls for caution are beginning to be heard.
But it is not only the stock market where the AI mania has caught hold. Next week, I will discuss how AI companies have been forced to either raise new rounds of equity or tap the bond market for funds now that their cash flows have been depleted. I also look at another player that has entered the fray and what that may mean for AI in the near term.
