Meta: The Ad Machine and New Cloud Business Hiding Behind the Capex Fear

Let's start with what scared the market

The moment Meta guided to $125-145 billion in capital expenditures for 2026, investors sold.

On the surface, I get it. That is a staggering number, and when a company starts burning cash at that scale, short-term margins get squeezed. So the stock pulled back hard — as I write this it trades around $582, roughly 21 times forward earnings.

But here's what the sell-off is missing: Meta's core business is still one of the best money-making machines on the planet.

The ad machine hasn't cooled off

Here's the headline: advertising across Facebook and Instagram grew 33% year-over-year at a 41% operating margin.

That means for every dollar coming in, 41 cents drops to operating profit. A company this size growing at this pace while holding that kind of margin is genuinely rare. Gross margin sits at 82% — every additional unit they sell, 82 cents of it is profit.

What I like even more is the quality of that growth. Over the last five years Meta has spent only about $5 billion on acquisitions, yet revenue has climbed impressively across every window — three, five, and ten years. They didn't buy their growth. They grew it. I weight that heavily.

And there are cards still unplayed. Overseas ad revenue for Facebook and Instagram isn't maxed out, and WhatsApp monetization is only just beginning. Half the planet uses these apps almost every single day.

July's new catalyst: Meta is selling cloud

Then a new catalyst dropped in July: Meta is building a cloud business.

Think Amazon Web Services, but the Meta version. They're going to rent out all that excess AI compute they've been building to outside developers. Those billions poured into infrastructure turn from a cost line into a revenue and profit line.

While the market fixates on the capex, I'm watching the point where that capex converts into a new revenue stream. That's the heart of the story.

Meta by the numbers

I don't judge a company by its share price ($582); I look at the market cap. This is a ~$1.5 trillion business. Enterprise value is about $1.58 trillion, and that ~$70 billion gap is essentially net debt.

Why does that matter? Because you compare it to cash flow. Meta produced $48 billion in free cash flow last year. That $70 billion of net debt is less than two years of free cash flow. They pay a dividend too, but it eats only about $5.5 billion of that FCF.

One observation I find interesting: with an 82% gross margin, you'd expect net margin to keep improving as the company scales — but Meta's profit margin has been remarkably stable across the last one, five, and ten years. Overhead and taxes have grown right alongside the business. That isn't necessarily bad — it can mean better service or future growth — but the "margins expand automatically with scale" story hasn't shown up here yet. Worth remembering.

So what would I pay?

The question is always the same: not whether it's a great business, but what you pay for it.

Here are my assumptions for the next ten years: revenue growth of 7/10/14%, net margins of 29/31/33% (with FCF a touch lower given the capex), and a P/E ten years out of 18/22/26. The market average is 15-16; I don't consider Meta average. High returns on capital, products half the world uses daily, and steady growth justify an above-average multiple.

My desired return is 9% — that's a pure intrinsic value, no margin of safety. And I want to be clear: if you buy individual stocks, build in a margin of safety. If a 9-10% return is all you want, just buy a low-cost ETF. Taking on single-company risk should demand more.

Run those inputs and I get a low of $550, a high of $1,400, and a midpoint of $850. At $590, if my middle case holds, that's about a 13.7% discounted-cash-flow return. To me that says "worth more research," not "back up the truck today."

Getting paid to wait

I don't just sit on my hands until my price shows up. I sell cash-secured puts.

In plain English: I commit to buying the stock at a lower price than today, and the market pays me for that commitment. On Meta, selling a put about a month out at a $520 strike pays roughly $7.87 per share. Repeat that monthly and you're looking at a cash yield in the 17% range. If Meta falls below $520, I buy at the price I wanted and keep the premium; if it doesn't, I keep the premium and wait for the next setup. Either way, I've built a structure where I don't lose.

The risks and the counterargument

I have to look at the other side too.

The biggest risk is that the capex doesn't come back as profit. Burn tens of billions in the AI arms race without commensurate revenue, and today's margins and valuation get hard to justify. The cloud business, too, is still a "we're going to" — not proven revenue yet.

And as I noted, this is a company whose net margin hasn't expanded automatically, so a generous multiple leans heavily on future growth. The moment that growth slows, 21x may not look cheap.

The conclusion is simple. I have no doubt Meta is a great business. The question, as always, is price. I only buy great businesses when the price makes sense.