It was all about AWS when Amazon reported results after the bell on Wall Street Thursday.
Of course, it’s always about AWS when Amazon reports, but the cloud readout took on an extra sense of urgency in the wake of Microsoft’s sharp, ongoing post-earnings rout.
As a quick, but relevant, aside, the Microsoft selloff feels like an overreaction at this point to the extent it’s predicated purely on an Azure print which was hardly bad, and a cloud outlook that, while not gangbusters, probably didn’t call for a ~20% haircut in the stock. If you want to cite too much spending and/or broader concerns about software, I’m sympathetic. But the idea implicit in analysts’ Azure color — namely that you can grow a business at a 40%+ clip in perpetuity — is patently absurd.
Coming back to Amazon, AWS sales in Q4 were $35.58 billion, up 24% YoY. That’s a beat and the quickest in years, but given investors’ growing intolerance with the admittedly stark contrast between escalatory capex and cloud growth which, while healthy, isn’t breakneck, I doubt it’s enough. Consensus wanted $34.90 billion.
Overall, sales at Amazon last quarter were $213.39 billion, up 13.63% and above the midpoint of the company’s holiday season guide. But here again, the beat wasn’t that pronounced. Analysts were expecting $211.33 billion.
The guide was… I don’t know, ok I guess. Taking the midpoint, current-quarter sales will be $176 billion, Amazon indicated. That’s a narrow beat versus expectations, and it’d represent 13% growth.
The main issue here’s the spending. Amazon said Thursday it expects capex to be $200 billion in 2026. That’s 35% more than analysts expected and it’s going to drag profit. Amazon sees current-quarter operating income between $16.5 billion and $21.5 billion. The Street was looking for more than $22 billion. So, even if income prints at the high-end, it’d be a miss to the pre-guide consensus.
I think the comparison between cloud growth and capex is apples to oranges (although they’re obviously connected given the spending’s aimed at juicing growth), but analysts and investors are obsessed with that juxtaposition. And here we have Amazon tipping spending far higher than expected against AWS growth printing only a small-ish beat and the guide midpoint tipping flat overall sales growth for a fourth straight quarter.
In explaining the spending plan, Andy Jassy essentially said there’s too much at stake — the possibilities are too vast — to risk underinvestment. He cited “seminal opportunities like AI, chips, robotics and low earth orbit satellites” and, in a half-hearted attempt to calm what I assume he knew would be frayed nerves, said Amazon “anticipate[s] strong long-term return on invested capital.”
Again, if what happened to Microsoft over the last six sessions (five of which were down days) is any indication, the market’s going to frown at the capex guide absent some reason to believe AWS growth is going to accelerate dramatically in the near-term.
Is AMZN able to fund this without reducing their share buybacks?
Hence expectations for $2 trillion in IG supply this year.
Speaking of Msft, you said after earnings that the sell off was a trend and not necessarily a point at which to buy.
Good call.
This reminds me of when Jerry Sanders of AMD would tell us “real men have fabs” and we investors ate it up until we didn’t.
My lovely wife “convinced” me to start attending yoga classes.
I told one instructor “Real men don’t use mats.” That only merited a quizzically raised eyebrow.
Throwing away approx $75 million on Melania and destroying the Washington Post brand are just drops in the bucket I suppose.
You mean the documentary masterpiece with a surely not manipulated 99% rating on rotten tomatoes?! The piece of art that will transcend humanity and be the sole representation of human culture on the next Voyager golden record?!
Yep, that one. Just flat out bribery.
There are two kinds of errors that can be made by decision makers. One is to go all in on something which turns out to fail. That one will cost you actual money can easily end one’s career. The other is to not do something that could have succeeded. That one creates an opportunity loss only. Few have lost their jobs over this one. It strikes me that we do not yet know enough about AI’s potential to be in such a rush to fail. I just acquired a new Windows 11 computer. I’ve now had it for a week and a half during which time it has crashed numerous times, it frequently spews AI vomit all over my screen making virtually impossible to complete the tasks for which I purchased it. In fact, it Just threw up making me start over. God, I hate AI.
I can relate. I decided to keep my older laptop (from 2020) that had Windows 10 (no longer serviced by Microsoft) and it doesn’t have the configuration capacity for Windows 11. I pretty much only use it at home (on secure wifi) to update some excel spreadsheet and do my tax return ( I download the tax software and do not store the return in the cloud). If my laptop dies, it dies. I’ll deal with it then. I did renew my security subscription for good karma.
I can’t stand the AI on my phone, but I’m learning how to shut it down and work around it.
I bought a new Windows box just this week as well. I’d blocked my old PC from updating Win10, but no longer.
The secret is to shell out the extra cash for the Win Professional edition. They don’t stuff all the ads and other garbage in the face of what they’re presuming will be a corporate/professional work station.
With Win10, they pushed plenty of ads, but you could block most of it and ignore the rest. With Win11 it’s nearly impossible. (I did stumble on a second method with my kid’s PC. If you have parental controls on, they don’t push ads or AI slop at you.)