Listen to the audio version of this article (generated by AI).

Nvidia’s blowout earnings… the “Messy Middle” of AI tariffs… an AI thrift-store play from Eric Fry

Yesterday afternoon, all eyes were on Nvidia (NVDA). Would the chip giant report earnings that lived up to lofty expectations?

Legendary investor Louis Navellier was clear:

I expect NVIDIA to beat; I expect them to guide higher and that’s that.

So, was he right?

Yes – but frankly, even Louis wasn’t prepared for the numbers.

Let’s jump to his Growth Investor Flash Alert podcast this morning:

Wow, wow, wow… It’s incredible.

I’m kind of numb, to be honest with you… I didn’t expect it to be this good… I’m ecstatic.

You have to hold this stock. It’ll be $300 by the end of the year, and $500 by the end of the decade.

Diving into the details, Nvidia’s revenue came in at $96.2 billion for the quarter, up 18% from the prior quarter and up 106% from a year ago. This sailed past the roughly $92 billion Wall Street had penciled in. Earnings landed at $2.22 a share, comfortably ahead of the $2.08 or so analysts expected.

But the big question on Wall Street’s mind wasn’t about “yesterday,” it was about “tomorrow.” Here again, Nvidia didn’t disappoint.

For the current quarter, Nvidia guided to $108 billion in revenue, plus or minus 2% – blowing past both the $104 billion analysts modeled and the psychological $100 billion mark that no company has ever cleared in a single quarter.

Then CFO Colette Kress did something Nvidia almost never does – she guided a full year out. She told analysts to expect roughly 70% revenue growth in fiscal 2028, against a Wall Street consensus closer to 44%. Applied to this year’s numbers, that would point to something like $670 billion in revenue next year – a figure that would vault Nvidia past both Apple (AAPL) and Alphabet (GOOGL).

And here’s the kicker: Kress called that 70% a “supply-constrained” number. When analysts pressed for the real ceiling, CEO Jensen Huang wouldn’t put a figure on it. He’d only say true demand was “a lot higher.”

Huang’s explanation comes down to one shift: agentic AI. He argued that the technology has crossed from novelty into genuine productivity:

AI has reached its inflection point. Its tokens are productive and profitable. Now, compute is revenue.

There was a small blemish with margins. Kress warned that gross margin will slip before bottoming near 71% to 72% in the fiscal fourth quarter, with soaring memory prices the culprit.

Meanwhile, veteran trader Jonathan Rose of Masters in Trading: Live noted that $7.8 billion of income came from investment gains – not chips. And in today’s free episode, he walks through Nvidia’s numbers through a trader’s lens – there’s an angle here you won’t see in the press’s glowing headlines. You can catch it right here.

Still, the market likes what it heard. NVDA is up more than 8% as I write.

But the wider AI complex is dealing with a headwind from a different source this morning

Ordinarily, a blowout earnings report like this from AI’s poster child should have the whole AI trading roaring. While many AI stocks are up, it’s not quite the surge you might expect.

Why?

Politico is reporting that the Trump administration is weighing a fresh round of tariffs on semiconductors – and this time casting a far wider net of products, from laptops to data-center servers to gaming hardware…

Chipmakers have warned, again and again, that tariffs imposed during a supply shortage could choke the very AI buildout Washington says it wants to win. The White House sees it differently – reshoring chip production is a national-security imperative, full stop.

Here’s the thing: both can be true at once.

That tension is what we’ve come to call the “Messy Middle” here in the Digest. As we move deeper into the age of AI, we’ll be forced – repeatedly – to choose between competing “goods.” Cheaper chips or a domestic supply chain. Faster progress or tighter security. One worthy goal advanced at the expense of another.

We simply can’t have everything we want, in equal measure, at the same time. Something gets prioritized. Something else waits.

And that’s the real challenge for you and me. The job isn’t to guess which “good” wins on any given day. It’s to build a portfolio resilient enough to withstand whatever gets prioritized, as the political, social, and economic pressures behind those choices keep shifting over time.

We’ll keep tracking this.

The other side of cheap AI

The cost of AI tokens is falling fast, with every indication that even cheaper AI is on the way.

As we covered in Tuesday’s Digest, eventually, this trend will reach a tipping point – one that punishes the AI companies whose whole business is selling access to raw compute. Their pricing power depends on that compute staying scarce and expensive. As it gets cheap and abundant, that edge will erode.

But increasingly cheap AI will be an enormous gift to a second type of company – the “AI Appliers.” These are businesses that bolt AI onto what they already do. Every time the cost of AI drops, their bill shrinks, and their margins widen – even as that same AI increases what they can produce.

This process has already started.

As AI gets cheaper and more capable, AI Appliers are already doing more – sorting faster, pricing smarter, and automating what used to take armies of people – all at increasingly lower prices.

Which brings me to a concrete example from an industry you’d never associate with AI.

An AI Applier to consider today

Our global macro expert Eric Fry, editor of Fry’s Investment Report, just published a piece pairing AI with the “circular economy” – the business of collecting, sorting, and reselling used goods.

The problem these companies have always faced is chaos. Their raw material shows up in garbage bags and donation bins – one moment a frying pan, the next a cracked smartphone – and until recently, no software could make sense of it.

As Eric puts it:

Software didn’t know how to value a used jacket, grade a bruised apple, or distinguish high-quality scrap metal from low-grade waste. Humans had to do that work – slowly, inconsistently, and expensively.

AI changes that. And here’s Eric with the whole investment thesis in two lines:

Artificial intelligence is finally giving circular-economy companies the missing ingredient: a way to organize chaos at scale…

That shift has the potential to unlock margin expansion every bit as powerful as what AI is doing for digital businesses.

Margin expansion off a shrinking cost base – that’s the Applier trade, exactly.

Eric’s example is Savers Value Village Inc. (SVV), the largest for-profit thrift operator in North America, which processed 1.1 billion pounds of donated goods last year.

It started small, with Automated Book Processing: books ride a conveyor belt while scanners read each ISBN against a database of 55 million titles, and a robotic arm prices and sorts them without stopping the line. That system now runs across roughly 85% of the fleet.

But the bigger tell landed just three weeks ago…

SVV rolled out ThriftIQ, an AI pricing platform aimed at apparel – its core product, not a side category – and has already tested it across 58 stores and more than 25 million items. This is likely to accelerate what’s already been some healthy margin expansion.

Gross product margin has climbed from 50.7% in 2019 to 55.3% last year, as automation and smarter processing worked their way through the business. ThriftIQ pushes that same lever harder – straight at the biggest cost bucket Savers has.

Eric already holds SVV in Fry’s Investment Report. To learn about joining Eric to access his other AI Appliers, click here to learn about joining him.

Last call for the AI Revolution Portfolio

Step back from today’s stories, and one question ties the AI ones together.

Nvidia just told us the buildout isn’t slowing – it’s accelerating. A guide to 70% growth next year, a backlog north of $2 trillion, hyperscaler spending marching toward $1.3 trillion. That’s a tidal wave of capital about to wash across the entire AI landscape.

But a rising tide doesn’t lift every boat here. Between Nvidia’s blowout and the quiet rise of the Appliers, the field is splintering fast into winners and losers – and with hundreds of AI stocks now trading, the hard part isn’t believing in the theme. It’s knowing which names belong in your portfolio.

That’s the question that Eric, Louis, and our hypergrowth expert Luke Lango set out to answer. The three of them spent weeks narrowing more than 200 AI recommendations down to the roughly 20 stocks in their newly rebuilt AI Revolution Portfolio.

Which ideas deserve a spot? Which ones overlap? How much capital should each position get? That’s the problem the portfolio was built to solve.

The last time our experts did this, their portfolio more than doubled Nasdaq’s return. Following its December 2024 rebalance through July 23, the AI Revolution Portfolio gained 58%. Over that same stretch, the Nasdaq rose 25%, the S&P 500 gained 24.4%, and the Dow advanced 19%.

If you want to learn more about the AI Revolution Portfolio, it’s officially last call. The free replay of last week’s event comes down tonight. So, if you’ve been meaning to watch it, now’s the time. Click here to watch before it’s gone.

We’ll keep you updated on all these stories here in the Digest.

Have a good evening,

Jeff Remsburg