Big tech’s big gamble

Good morning. Gold fell 3 per cent yesterday, and various news outlets (not the FT, I’m proud to say) put this down to hopes for a US-China trade truce, which would (in theory!) reduce gold’s geopolitical risk premium. But the argument that the gold rally is explained by political tension or dollar debasement, which may have been mildly compelling a few months ago, no longer makes any sense at all. Recently, gold has become a creature of retail, momentum, and Fomo. It does not need a political excuse to behave capriciously. Send us your thoughts: [email protected].

Big tech week

This week, five of the world-bestriding Big Tech companies report: Microsoft, Alphabet and Meta tomorrow, and Apple and Amazon on Thursday. Together these stocks account for a fifth of the value of the S&P 500. The numbers they report and the comments of their executives will matter immensely in an ageing bull market that looks just the teensiest bit ragged around the edges.

So, as we do a few times a year, let’s take a step back from the artificial intelligence hype and have a look at the fundamentals. Start with the most important fundamental right now — which is no longer revenue or earnings growth, but rather investment spending. Here is capital expenditure, on a rolling 12-month basis, at the five Big Techs reporting this week:

Four of the five companies have basically doubled capex over the past year and a half, and they are expected to push spending even higher through to the end of this year. The sums involved are staggering. For those four — Alphabet, Amazon, Meta and Microsoft — the biggest determinant of the performance of the shares in the medium-term future will be the evolving prospects for earning an acceptable return on these epic investments.

And, keeping things interesting, Apple, continues to take the other side of the bet. Apple refuses to join the arms race, wagering their device franchise will be able to reap the benefits of AI without paying to build it. Folding while the other four ante up could be brilliant; it could be a disaster; I have no idea.

The inevitable result is that free cash flow at the four big spenders, long the hallmark of their super-scalable business models, is starting to ebb a touch:

The spending will force some choices at Meta. In the past 12 months, it has spent $53bn on dividends and buybacks. It is expected to generate $34bn in free cash flow next year because of all the AI spending. It can’t do it all (without borrowing). Microsoft and Google have more wriggle room, but it’s not infinite. Amazon does not pay a dividend or consistently buy back stock, and has long been happy to burn what it earns.

There will be pressure on earnings, too. We can see this in the ratio of capital investment to depreciation expense at each company. Depreciation is where investment spending shows up (in a smoothed form) as an expense on the profit and loss statement. At a growing company, capex can be consistently higher than depreciation expense, and the timing is often uneven, but over time the two have to be roughly proportionate. At the spendthrift four at the moment, depreciation expense has a long way to rise to catch up with capex. That means pressure on profit in years to come.

None of this to say the AI bet isn’t a good bet (again: no idea!). It’s just to point out that it is a bet. Money has been pushed to the middle of the table, and if the cards don’t come up right, the money will be gone, and will be missed.

The big gamble by the Big Techs’ may or may not have something to do with a modest, but perceptible, change in the way the market is treating the companies’ shares. Since the market’s “liberation day” low in April, all of them save Amazon have outperformed the S&P 500; Alphabet, once the big antitrust case was resolved, has been spectacular.

If you look closely at that chart, however, you will notice that in the past several months, only Alphabet has meaningfully outperformed the market. Big tech is trading differently than it did even a year ago. Look, too, at the five stocks’ premiums to the S&P 500 (the shares’ relative price/earnings multiples):

The Big Techs’ premiums, with the possible exception of Microsoft, have been flat-to-down over the past year or so. Earnings growth, rather than rising valuation multiples, is driving the shares higher. This is healthy enough; valuations cannot expand forever. But it looks like the market is being more pragmatic about the AI gamble than it is getting credit for. Hype will not be enough. Investors want to see performance this week.

One good read

Ancient blood-drinking aliens and the Feds who kill them.