It’s Full Steam Ahead for the Hyperscalers

Managing Editor’s Note: What if you had a way to pocket more gains in ONE day than the S&P 500 typically delivers in an entire year… again and again… without having to buy, hold, or sell?

Our colleague Larry Benedict’s “Zero-Stock Retirement Plan” strategy lets you do exactly that… and according to him, all it takes is a single ticker and a simple five-minute morning trade.

If you want to learn more about this strategy, you can go here to automatically sign up to join him next Thursday, May 7, at 8 p.m. ET, where he’ll explain exactly what this strategy is all about.

Wow! What a big week this week has been for artificial intelligence.

The investment and expansion continue to grow and accelerate, with the hyperscalers naturally leading the charge. And they continue to revise their forecasted capital expenditures (capex) higher to support growth in AI.

We can see it in the numbers…

• AI spend from just the top four AI hyperscalers – Microsoft, Alphabet, Amazon, and Meta, all of which reported earnings this week – is projected to exceed $700 billion in 2026. This is just the top four. Combined Q1 2026 spend alone hit $133 billion, reflecting a roughly 70% increase versus reports from Q1 2025.
• Microsoft (MSFT) reported $82.9 billion in revenue, up 18% year-on-year, driven by growth in its Azure cloud services division.
• Alphabet (GOOGL) killed its earnings with Google Cloud pulling in $20 billion on AI demand – a 63% jump – and a backlog that nearly doubled from the previous quarter. Also raised 2026 capex guidance to $180–190 billion, largely driven by this expansion of its Google Cloud business as well as other AI project development.
• And Amazon (AMZN) beat forecasts due to Amazon Web Services growth – its fastest clip in 15 quarters. Net sales also increased 17%, outpacing expectations.
• Even Meta (META) beat revenue forecasts, with revenue climbing 33% year-over year. It even increased forecasted capex spend on AI to $145 billion… just this year.

And if all that weren’t enough…

Google committed $40 billion in a deal with Anthropic… That’s $10 billion to be invested now at a $350 billion valuation, and $30 billion more to be invested as Anthropic reaches certain targets. In exchange, Anthropic is committing up to 5 gigawatts of TPU compute over the next five years.

As a reminder, Google’s tensor processing units (TPUs) are Google’s own custom-designed AI application-specific semiconductors that it uses TSMC to manufacture. TPUs are an alternative to GPUs and are far more efficient for AI applications.

And if that weren’t enough, Anthropic is rumored to be in the process of raising a new $50 billion round at a $900 billion valuation.

It’s not slowing down.

It’s speeding up.

Have a wonderful weekend,

Jeff

SpaceX and Spectrum and Starlink… Oh, My!

Dear Mr. Jeff Brown,

[Some questions for you]… Does the Starlink phone need an antenna dish to be connected to the satellite? Re: Starlink [buying] EchoStar. Is EchoStar’s spectrum better than GlobalStar’s?

FCC wants Starlink and Amazon (GlobalStar) to stay in their own approved and licensed spectrum. So is Starlink at a disadvantage to Amazon?

Apple has a new technology that allows the Apple iPhone to be connected directly from the satellite to the Apple iPhone without an antenna dish. So, is Apple better than the phone that Elon Musk wants to come up with?

And now that Amazon bought GlobalStar with Apple in the deal, Amazon should be in a better position to compete with Starlink. How good is GlobalStar’s future?

– Alice L.

Hi Alice,

Quite a few interesting questions there.

First off, there is no Starlink phone. Starlink, part of SpaceX, is the largest constellation of satellites in history, and the first space-based internet infrastructure – literally a world wide web – and something that I refer to as orbital web services (OWS).

Musk and his team at SpaceX designed Starlink to operate in low Earth orbit (LEO) and in such a way as to enable standard Android OS or iPhones to be able to send/receive signals directly to/from Starlink satellites.

There are a few companies that have enabled this kind of service in addition to Starlink, such as AST Spacemobile (ASTS), Lynk, and Globalstar. Each of these companies is trying to partner with as many mobile network operators (MNOs = wireless operators) around the world as possible to grow their businesses. Starlink has had the greatest success. It’s not even close.

Starlink, by far, has the superior spectrum. I’d like to encourage you to read The Bleeding Edge – The Battle for Orbital Web Services, where I provide a lot more detail about this exact issue around the value of the spectrum.

Amazon got caught flat-footed – as we discussed in Clash of the Orbital Titans – as has Google with regard to orbital web services. It’s kind of astounding when we think about it.

Amazon’s Amazon Web Services (AWS) and Google’s Google Cloud are the two largest terrestrial web services businesses on the planet. Both companies have one single constraint to growing their business – energy production. And the solution has always been obvious…all the free, clean energy that we could possibly use is in orbit. Free cooling as well.

This is why Amazon “complained” to the U.S. Federal Communications Commission (FCC) after SpaceX successfully acquired EchoStar’s spectrum. It realized Starlink as a competitive threat to its web services business. It was just trying to slow SpaceX down.

It didn’t work.

SpaceX’s acquisition of EchoStar’s spectrum precipitated Amazon’s acquisition of Globalstar (GSAT). It was the next best spectrum that Amazon could acquire, and Globalstar knew it. Amazon had to overpay for the acquisition. Additional details are in The Bleeding Edge issue that I referenced earlier.

To be clear, Apple (AAPL) hasn’t come up with any special technology that enables direct-to-cell-phone services from space. The most important technology is what Starlink developed, and the spectrum that it uses to connect with mobile phones.

Elon Musk has been pretty clear that he has no interest in developing a smartphone, and that it would only be under extreme circumstances that he would even consider doing so. An example would be if both Google through Android OS and Apple blocked all SpaceX and Tesla applications from working on their phones, or if both companies engaged in widespread censorship.

Now, Google absolutely did engage in widespread censorship, manipulation, and attempts to influence the outcome of U.S. elections in the past, so this is absolutely a possibility. Apple’s censorship was on a much smaller scale, but there still was some, unfortunately.

The situation is better now, and freedom of speech has been restored on X, but we could certainly see it swing back to a totalitarian regime with complete censorship, even worse than what we experienced during the pandemic, and worse than what is happening in most of Europe today.

For all of our benefits, I hope it never comes to that.

One wildcard is the entrance of OpenAI into the smartphone industry. I wrote about that this week in The Bleeding Edge – OpenAI’s Agentic AI Smartphone.

If this product were to gain traction at scale, and if OpenAI were to engage in censorship of information, information manipulation, and the banning of SpaceX/Tesla services – which I see as highly likely – it could force Elon Musk to do something about it.

As for Amazon, it is so far behind SpaceX that it will take more than a decade to catch up. It has almost no infrastructure in space to provide OWS. SpaceX has more than 10,000 satellites. With spectrum, Amazon now has a path to start building and offering OWS, but it will take a long time.

As for Globalstar, it has been acquired. So the stock will delist, and all assets will be owned by Amazon. Over time, Amazon will exit much of Globalstar’s legacy business and simply put the spectrum to work for the purpose of provisioning OWS.

And what about SBGI?

Hi Jeff,

In the current fight for spectrum, where does SBGI stand? At one time, they had a glut, right?

– John G.

Hi John,

That’s the right kind of question. You’re thinking about how the spectrum of other companies may be of value.

There is always a lot of nuance with respect to radio frequency (RF) spectrum. It’s always important to know exactly what the spectrum is, what it can be used for, if it is harmonized, and what it might be worth.

Sinclair (SBGI) is most well-known for holding UHF television spectrum in the 470–698 MHz range. It has a large portfolio of terrestrial broadcasting stations designed for….you guessed it… linear TV broadcasting. That just means the broadcasting of TV stations. Transmitting, not receiving.

The one exception is that Sinclair has a small slice (35 MHz) of contiguous spectrum in the 2 GHz band.

With all that said, none of Sinclair’s spectrum is considered cellular compatible bands, so it is not useful for the provision of direct-to-cell services like what Starlink can do. Sinclair’s spectrum is really only valuable for terrestrial TV broadcasting and perhaps for forms of datacasting. Other than that, not very interesting.

Terrestrial TV broadcasting is obviously becoming less and less important, given the proliferation of streaming services and wireless networks. For Sinclair’s UHF spectrum to become more valuable, the FCC would have to repurpose that spectrum for other uses, like cellular networks.

Why We Don’t Always Immediately Buy on an IPO Day

This email is for Jeff Brown, whom I first started listening to almost a year ago.

My 95-year-old Dad had invested in NVIDIA (NVDA) several years ago when Jeff was strongly recommending this stock. He made a lot of money on your advice and gave him some real confidence again that he once had.

Fast forward, not even knowing that my Dad had gotten his information from Jeff, I was becoming more and more impressed with the amount of research your company was doing, and very impressed with how you were going about it.

Now, tonight, I was reading your report for “2026 AI IPO Blueprint. At the end of this report, after explaining that [there’s a significant AI company planning to IPO in June], you said, “to be clear, we aren’t recommending that investors chase any of these companies at the time of their IPO. This should be considered a top-tier watchlist.”

So needless to say, now I am very confused…

– Patrick C.

Hello Patrick,

Thanks for writing in and sharing your story. Also, please say hello to your father and congratulate him on his investment gains (from me). That makes me very happy.

To clarify, the point that I was trying to make in the report that you are referring to is that we shouldn’t just blindly invest in an IPO without understanding the details.

Since you’ve been reading my work, I know that you know how important valuation is in any investment thesis. Valuation is core to understanding whether or not a stock is “cheap,” a great value, or overpriced.

What makes valuation hard to understand is that reasonable valuation multiples can change over time in any given sector, given the dynamics at any given time.

A valuation multiple that might be considered expensive today could be seen as cheap next year because a sector becomes hot and institutional capital is flooding into the sector. The opposite can be true as well.

Understanding valuations and valuation multiples must be done with context and considered at a given point in time. I know of no one in the industry who focuses on valuation in the way that I have over the last decade plus. It is one of the key reasons my investment strategies have been so successful for my subscribers.

I also ignore P/E ratios. They are accounting metrics and can be manipulated. For well-established companies with revenue and cash flows, all that matters is revenue growth, EBITDA, and free cash flows.

For pre-product revenue and/or companies that are in heavy R&D spend mode, valuation is more of an art form.

As for IPOs, the reason that I never blindly recommend any IPO is that while we know the IPO is coming, we don’t know the valuation, and we don’t have the final IPO SEC filing with the details about its financials.

What makes IPOs harder is that we can only gain an understanding of the details of the IPO and where it might price when the IPO roadshow begins in the days leading up to the IPO. This is where companies pitch their IPO to investment banks, assess levels of interest, and determine where they will price their IPO when it starts trading.

From that information, we can assess valuation and determine if it is way overpriced, underpriced, or reasonable. This informs an investment or trading strategy. We also need to consider market sentiment, momentum, interest levels, etc.

What makes the IPO markets unfair for normal investors is that allocations in hot IPOs are tightly controlled by investment banks. And the investment banks give the most shares to their largest clients and high net worth individuals, leaving only scraps, if anything at all, for retail investors.

And then, oftentimes the IPO is opened up on the mark a large percent above the actual IPO price. I can’t tell you how many times I’ve watched the fast money (hedge funds) run up an IPO only to dump shares on unsuspecting retail investors just to watch the stock crash in the days that follow.

The other key window to watch is the lock-up expiration, which happens six months after the IPO. We often see a lot of selling in the weeks leading up to the lockup expiration, as the expectation is that those with shares that have been locked up will sell for liquidity once their shares are free to trade.

I hope that provides some context. There are a lot of things my team and I weigh when considering what companies to add to our model portfolios. That goes for well-established companies, companies that recently went public, small capitalization companies, funds, convertible bonds, etc. It doesn’t matter what the asset class is… everything is carefully measured and taken into consideration.

I do, on occasion, recommend IPOs, but my team and I do that only when we have a complete picture of the company’s financials, valuation, status of business, both present and future, and a picture of IPO interest.

I hope this is helpful.

Jeff

P.S. Hello, Jeff’s managing editor here!

On the note of significant IPOs and the ways we at Brownstone like to play them…

A major AI company IPO is coming up this summer… the one Patrick alludes to above, to be precise.

And while that IPO is going to be huge – paradigm-shifting, even – for the AI industry… buying the company directly isn’t the only way to make the most of this historic IPO.

Really, it’s only one way to profit, which is why Jeff has identified a number of smaller companies with exponential potential that are also tied to this opportunity.

We have the replay available still for a short while longer if you want to learn more. You can go here to watch it…