The AI ​​arms race costs money

Good morning 10-year Treasury yields rose to a nine-month high of 4.7 percent yesterday after a robust ISM services survey and a Jolts report showed more vacancies. Both inflation breakevens and the dollar rose. To us, it looks just like the market is pricing in a hot economy. Send us cooling thoughts: robert.armstrong@ft.com and Aiden.reiter@ft.com.

Like 7 Capex

Accounting is boring but necessary. Particularly necessary: the difference between a capital expense and an operating expense. A capital expense (e.g., purchasing a big piece of apparatus) just isn’t charged on to profit on the income statement, as is an operating expense (e.g., paying a salary). Instead, a capital expense appears on the income statement over time, theoretically balancing the lack of profits with the lifetime of the capital assets. This distributed expense appears on a line labeled “Depreciation.”

I see you sleeping within the back. But I'm dragging you thru this annoying point since the world's most significant firms, the “Magnificent 7 Big Techs,” are making huge investments, mostly in artificial intelligence data centers. That's money today, however the expense will only be reflected in earnings per share over time. The AI ​​arms race has not yet fully achieved the winnings. The query is whether or not the market has digested the incontrovertible fact that it can need to do that soon.

Here are the capital expenditures for the five out of seven which might be faltering to a greater or lesser extent (at Apple, capital expenditures are stable; at Nvidia, investment money is available in and doesn't come out):

These are frightening numbers, they usually are still increasing.

Amazon appears to be an outlier, but that's not entirely true. You need to align expenses with the remaining of the corporate's financials. For example, you possibly can take a look at capital expenditure as a percentage of sales:

At Meta and Microsoft, every fifth dollar that is available in is spent as capital expenditure; At Alphabet it’s one in seven. And the trend is increasing (Wondering in regards to the big spending pile in 2022 at Meta? Remember the Metaverse?).

Now now we have a way of the size of the spending. But what is very important for our topic is the scale of the investments. That is, how much is capital expenditure relative to depreciation expense? For the last 12 months it looks like this:

Let's take Meta for instance: The depreciation expense is 9 percent of sales, the investment expense is 20 percent. This implies that depreciation expenses can have to extend significantly in the approaching years. This doesn’t mean that the group's operating margins can have to fall by the difference (11 percent) in some unspecified time in the future; Investments in data centers might be spread over 4 to 5 years. However, if current spending levels proceed, the strain on margins might be significant.

How much will Wall Street analysts estimate these five firms' margins will fall over the subsequent few years as AI investments increase? None in any respect. In fact, operating margins in any respect five firms are expected to extend somewhat over the subsequent few years. That is after all possible. With the exception of Tesla, all of those firms are growing their revenues quickly and their models have high operating leverage. But it won't be easy. Optimism about Big Tech's earnings momentum exceeds every part.

Are bankruptcies the canary within the coal mine?

Our colleague Will Schmitt identified yesterday that corporate bankruptcies within the US in 2024 reached their highest level in 14 years. At least 686 firms filed for bankruptcy last yr, 8 percent greater than in 2023 and the very best variety of bankruptcies since 2010:

Is this evidence that higher rates of interest have finally arrived for over-indebted firms? Or has the US economy slowed greater than we thought?

Of the ten largest bankruptcies in 2024, five were personal bankruptcies. One of them, Red River Talc, is Johnson & Johnson's holding company for baby powder liabilities. We checked out the remaining 4 and added two more familiar names: home goods store Big Lots, container maker Tupperware, fabric seller Jo-Ann Stores, tinsel emporium Party City, budget airline Spirit Airlines and Franchise Group, owner of retail chains like The Vitamin Shoppe and Pet Supplies Plus.

The theme might be seen at first glance: all of them are discount stores or franchise stores which might be aimed toward consumers with lower incomes. Like dollar stores and other discount retailers, many suffered as low-income Americans in the reduction of on spending amid high inflation. One of them, Big Lots, blamed this problem directly for its downfall. Taken together, they’re a transparent example of the “K-shaped” recovery within the US.

But these firms and due to this fact the insolvency figures for 2024 usually are not clear evidence of a slowdown. Many were dinosaurs; perhaps surprising that they lasted this long. Brick-and-mortar stores Jo-Ann, Party City, Franchise and Big Lots couldn't compete with Amazon or larger retailers like Walmart and Costco. The same goes for Tupperware and its anachronistic peer-to-peer sales model. Spirit Airlines is a more unique case: While all airlines struggled within the early days of the pandemic, the corporate was unable to stay competitive afterward, had problems with its aircraft and ultimately did not merge with JetBlue.

Were Fed rate of interest hikes the offender? The financial figures paint a mixed picture. All firms were burdened with variable-rate debt, and lots of took on much more debt after the pandemic (particularly Big Lots, Spirit and Franchise). But not everyone saw their interest expense increase: Spirit was in a position to reduce its payments over time after a giant increase in 2019, Jo-Ann and Party City kept interest payments under control, and Tupperware even shrank them.

The problem for many of them was weak profits and non-rising debts. Here is a chart of their net debt to Ebitda ratio. All but Tupperware jumped above six, the ratio that typically sends lenders running:

Judging by the most important bankruptcies in 2024, now we have a one-sided economy and no slowdown.

()

read

Branding issues.

FT Unhedged podcast

Can't get enough of Unhedged? Listen our recent podcastfor a 15-minute take a look at the most recent market news and financial headlines, twice every week. Find out about past editions of the newsletter Here.