🤖 Big tech's AI spending spree

Hey hey, happy Monday! 

A 6-minute update on our favorite tortured tech execs.

But actually, how the biggest tech companies can afford to spend $400 billion on AI ventures – and where the risk lies in their financial underpinnings. That money has to come from somewhere.

Before we get to the good stuff, I’ve got an announcement for all the financial advisors out there.

🚨 WE WANT YOUR THOUGHTS 🚨

We’re writing a research paper on how you use crypto in your practice, and we need your help! Give us five minutes of your time to take our survey and be entered to win a $100 Amazon gift card! Here’s the survey, thank you in advance!

Smash the button below to share OptimistiCallie with a friend 😊

Bonds with a winning edge

It’s been a while since we’ve checked in on the men of the Magnificent Seven.

Jeff Bezos is a married man after renting a whole Italian island for his wedding. Elon Musk has given up his government chainsaw after his precious Tesla’s stock price got cut in half. Mark Zuckerberg is still rocking that broccoli haircut, desperately trying to stay relevant when your grandparents are the only people using Facebook. The others seem to be lying low, unless there’s a tabloid I’m missing in my mostly financial media diet.

To be sure, they’re all tied up in mid-life crises on behalf of their companies, which are trying to pivot from the business models of subscribers, advertising and social clout that are sooooo 2010. Right now, every leader is throwing AI spaghetti at the wall with mixed results on what seems to be sticking.

Oh, and that experimental spaghetti? Super expensive, to the tune of $385 billion this year and $457 billion in 2026. The same amount of money as Denmark’s GDP this year, if you’re into the pointless size and scope sort of thing.

I can’t blame them for trying. AI seems to be all that Main Street and Wall Street care about these days. Announce a partnership with OpenAI and boom! – your stock rockets higher. Instant wealth generation, economy/tariffs/humanity be damned.

But in all the hype, you have to wonder – how are these guys blowing hundreds of billions of dollars on an unproven business model? Their pockets are deep, but are they that deep?

The answer: it’s complicated. Yes, they can all technically afford it through a mix of incoming money, savings and debt. But the way they’re paying may make a huge difference in how markets interpret their big, bold AI plans (that largely aren’t paying off yet).

Right now, the details matter with increasing scrutiny. A lot of the biggest AI players are making investments with each other, which could bind them all to the same fate if one project fails. We also don’t know what these companies may be giving up by putting all of their chips in what’s still a speculative venture.

Now, companies have lots of options when it comes to raising money.

More than we non-corporate plebeians have. They can tap all sorts of pockets and investors. Sell stock, sell debt, sell rights to stock or debt. Each funding source is a strategic choice rife with pros and cons.

We’ll focus on three today: cash flow (the income from business operations that companies can direct towards investments), cash hoards (the cash they’ve saved up for years), and debt (the ability to borrow money through public markets).

Also, not every Magnificent Seven company is spending gobs of money on AI.

Several of the biggest tech businesses are playing winner-take-all in building huge data centers to house networks, power and storage (via the cloud) for LLMs (large language models, the robots) to use. These companies are called hyperscalers, and they’re the big spenders in question.

This group of hyperscalers doesn’t include Apple or Tesla. Tim and Elon, you’re off the hook for now.

Cash flow

This is the easiest pocket for companies to pull from – spending money that’s coming through the door each year. You don’t have to tap your hard-earned savings or beg investors to finance your projects, phew! But in spending your incoming cash, you risk taking away from what your precious shareholders desperately want: proof of profits.

A company’s free cash flow is the purest way to look at how much hard cash it’s bringing in. Free cash flow is calculated outside of capital spending, so changes in this metric can show you how much business spending is eating into its future potential.

Look at free cash flow among the top five hyperscalers, and you’ll see a few patterns.

For one, they’re all very good at making money. Few of them have ever recorded periods of negative free cash flow, which means they’re all adept at generating more cash than they spend.

But the binge on AI projects could push cash flows into dangerous territory. Oracle sticks out like a sore thumb this year and next, as it’s expected to post negative free cash flow, the stark opposite of its fatter 30% free cash flow to sales per year since 2001. This is especially unusual for such a large and established business – only 6% of S&P 500 companies are projected to have negative free cash flow next year. Meta and Microsoft’s free cash flows relative to sales are expected to fall to multi-decade lows in 2026. 

Less free cash flow is a hard pill to swallow. When investors see slower – or negative – free cash flow generation, they often become more skeptical about how companies are using their money. Oh, you want to splurge on this project with no guaranteed payoff versus this other one or giving us a dividend? Interesting choice, Satya.

Savings accounts

The Magnificent Seven firms also have huge piles of cash in savings and short-term investments. The leaders may have their issues, but saving isn’t one of them. In the second quarter, these five hyperscalers alone were sitting on $341 billion in cash, or nearly 13.7% of all S&P 500 company savings.

Don’t get me wrong, $341 billion is a lot of money. However, these balances are dwindling when you compare them to other publicly traded companies. In fact, that 13.7% share of S&P 500 cash is the lowest in nearly 12 years. While the cash cushion is thick, it’s thinning.

It’s hard to say how much of these savings are being used for AI initiatives. Share buybacks and dividends among tech companies have stayed fairly constant over the past five years, so I’d guess lower cash hoards have to do with higher investment. And for what it’s worth, you want these companies to spend on the future – their innovation is what makes them magnificent.

Still, the right amount of savings can be a delicate balance. As cash on hand falls, the company’s quality decreases with it. Lately, big tech has earned a reputation for its big cash piles and overwhelming competitive advantages. If tech pulls from its savings to splurge on AI, it’s threatening this perception of quality.

Microsoft and Google’s cash on hand have dropped the most in the past four quarters, by the way. Nothing alarming, but trends worth watching.

Debt deals

Ah, debt. While the idea of borrowing money can ruffle feathers, you have to accept that companies (wisely) tap this option frequently. Borrowing is often a smarter financial strategy than spending cash reserves or crimping your cash flow (for reasons I mentioned above).

But there are costs to consider here, too. You’re ultimately paying this money back plus interest. Luckily, tech hyperscalers can borrow money at impressively low rates. These are some of the most creditworthy firms on the market at a time when investors are demanding the lowest interest rates for corporate debt relative to similar Treasuries in decades.

They’re in a good position to borrow, too. Google and Meta actually have enough cash on hand to cover all of their debt obligations today if they wanted to. Oracle is the only name out of the five with a net debt (or debt minus cash) to earnings ratio higher than the average S&P 500 or Nasdaq constituent.

Nasdaq 100 companies overall are holding about half of the debt they had in the early 2000s when you account for earnings and cash on hand. These certainly aren’t the go-go days of the tech bubble, baby.

But again, perception matters when it comes to debt. Investors can be cool with debt-financed projects until they’re not. A spike in interest rates can make it tougher for companies to raise money. Higher debt levels raise eyebrows.

There’s a workaround, though. Sort of.

Companies have the option to raise debt through special purpose vehicles – or entities controlled by the company that aren’t financially intertwined. It’s like asking your friend to run a GoFundMe for herself, but she’s really funneling all the money to you (less nefarious, though).

Tech execs are getting creative now, too, according to several reports. Meta just gathered $30 billion in the largest private deal on record – none of which will sit on its own books.

In these cases, the companies pay more to borrow money, but the off-the-books financing is less messy to an investor’s eye. The debt still exists, though, and at least one firm sits on the other side. If these AI ventures don’t work out, the risk could be more widespread than you think.

So what’s the point here, Callie?

None of these tech companies is going to blow up tomorrow.

Neither will any of these infamous execs go broke, absent a major scandal.

Even so, there’s still a lot of risk to think about here as tech takes over your investments.

Financial stability has levels to it, as we can see with each hyperscaler. And as these company execs stumble through their mid-life crises, they’re willing to do almost anything to grasp this AI moment. Including weakening their own financial standing to spend billions on data centers, while AI applications remain elusive for most businesses.

AI will change the world, but how long are investors willing to wait for this experiment to pay off?

I’m not sure – and it’ll differ by company. Fewer financial levers to pull, more scrutiny on your AI shopping sprees.

I’ll leave you with this chart.

When prices have stretched so far beyond actual profits, there’s a lot of room for disappointment.

Don’t be surprised if shareholders choke on higher spending and lower profitability in the coming months.

Thanks for reading!

Callie

Like what you just read? Share it with a friend, pretty please 😊