We talked last week about earnings of six of the seven AI kings.

The seventh, Nvidia comes later this month.

And remember, it was two years ago in Nvidia’s May earnings call that Jensen Huang shocked the world, declaring “the beginning of a major technology era.”  He told us there was a “rebirth of the computer industry” underway, where “AI has reinvented computing from the ground up.”

And he told us there was a “retooling” going on across the economy, the beginning of a 10-year transition of the world’s $1 trillion data center, to accelerated computing.

And he had the numbers to back it up.  They grew revenues by 19% that quarter, from just the prior quarter (!), with the outlook to grow over the next quarter by 52% (shockingly huge).

As you can see in the chart, this was the beginning of Nvidia transforming itself into an AI company (growing data center business from 60% of the entire Nvidia business, to now nearly the entire Nvidia business).

We’ve now heard from six of the seven AI kings on Q1.

Remember, just months ago there were questions about whether the huge capex plans from this group would go forward, after the DeepSeek disruption of late January.  

They answered the questions with $300 billion worth of capex planned for 2025. 

Meta just raised its planned capex by 10%.  And Apple just committed to spend $500 billionover four years. 

And Satya Nadella, head of Microsoft, gave several signals this week that the technology revolution is accelerating.   

He said they are bringing on data center capacity (and executing the plan to spend $80 billion in capex this year) but customer demand for those computing resources is outstripping their supply.  

And that has a lot to do with this:  The large language model capabilities are (in his words) “doubling in performance every six months.” 

And clearly there has been a significant breakthrough in the past month or so, as he said Microsoft processed 100 trillion tokens in the quarter, half of that in the past month alone.   

That AI workload processing is up five-fold from a year ago.   

This explosive growth is all about AI agents automating tasks, from designing infrastructure, building and executing marketing plans to developing software. 

With that, we should expect a productivity explosion ahead. 

And hot productivity tends to be very good for economic growth.  

Jerome Powell himself presented back in 2016 at the Peterson Institute (here), that high productivity growth is a driver of a higher long-term potential growth rate of the economy.  

That’s good news.  Because it means we can see real wage growth — the kind that restores purchasing power and quality of life (catching up to the reset in the level of prices). 

And importantly, productivity driven wage gains are non-inflationary because rising wages are offset by rising output. 

So, stepping back from the media and geopolitical noise, this is the big picture — we’re in the early stages of an industrial revolution.

If you want to own the stocks of the companies building the infrastructure to power AI, the companies delivering the capabilities of AI to hundreds of thousands of businesses, and the companies that will best leverage the productivity enhancements from AI, you can find them in our carefully curated AI-Innovation Portfolio.

If you haven’t joined us yet, now is a good time.

In my note yesterday, we talked about the likelihood of a negative GDP number and a soft inflation number in this morning’s data. 

We got both.

So, with this morning’s inflation number, the Fed now has rates set at 200 basis points above the rate of inflation — a historically tight monetary policy stance.  By design, that puts downward pressure on an economy that just contracted in the quarter.

Next up is Friday’s jobs report — where a negative surprise is quite possible.  Remember, government job cuts from the past three months still haven’t shown up in the monthly labor data. 

Of course, a weak jobs number would compound a deteriorating economic picture.  And on that front, we got a warning shot this morning from a soft ADP private payrolls report. 

So, with the above in mind, the Fed meets next week. 

Let’s revisit what Jerome Powell said in March about conditions for adjusting policy, or not:  

“If the economy remains strong and inflation does not continue to move sustainably toward 2 percent, we can maintain policy restraint for longer. If the labor market were to weaken unexpectedly or inflation were to fall more quickly than anticipated, we can ease policy accordingly.”

Also, remember the Fed has been telling us for the past year that signs of “cracks” in the labor market would be a condition to “react” (i.e. with rate cuts). 

We get the first look at Q1 GDP tomorrow.

And it should be negative.

Going in, the latest reading on the Atlanta Fed’s GDP model has the economy tracking at a 2.7% contraction in the first quarter (the green line).

But the Atlanta Fed also estimates that Q1 GDP has been dragged down by more than 5 percentage points because of the impact of imports.

GDP equals consumption + investment + government spending + net exports.  In the current case there has been a massive pull forward of imports to get ahead of tariffs.  And that means a big negative “net exports” drag for GDP.

The good news:  If we normalize for this anomaly in net exports, using 2024 average net exports, GDP for Q1 would be closer to 2% growth (working with the projections in the Atlanta Fed model).

But that won’t stop the media from hand wringing over a negative GDP number.

And then we’ll get the April jobs report on Friday.  And thus far, we haven’t seen the government job cuts reflected in the monthly labor report.  But as we’ve discussed over the past few months, a labor market shock is coming.

Who is on high alert to “unexpected weakness” in the labor market? 

The Fed.

Add to that, we’ll also get the Fed’s favored inflation gauge tomorrow, PCE for the month of March (its inflation target is 2% headline PCE).

And it should be disinflationary. 

Remember, the March CPI came in earlier this month at 2.4%.  And the last time we saw a number that low was September of last year — which happens to be the month the Fed kicked off its easing campaign, with a 50 basis point rate cut.

And based on the CPI and PPI reports, Fed Governor Waller has already given us the Fed’s estimate of where PCE will come in tomorrow.  In a prepared speech on April 14th, he said PCE should be flat for March, and fall to 2.3% for the year-over-year number.

That said, the Fed meets on rates next week.  The market is now pricing in a rate cut for June (a resumption of the easing cycle).  But a negative GDP number, a soft inflation number and a negative surprise on jobs over the next few days would make the Fed meeting next week more eventful.

We heard from Telsa and Google last week, on Q1 earnings.  And we’ll hear from four more tech giants this week.

Remember, just months ago there were questions about whether the huge capex plans from this group would go forward, after the DeepSeek disruption of late January.

But they didn’t flinch.  Instead, they all pressed the accelerator.  In aggregate, they announced $300 billion worth of capex planned for 2025.  Importantly, all in response to “signals of demand.”

Where do they stand now?

For the most recent quarter, Google and Tesla reaffirmed the big capex plans.  And we should expect more of the same from the cohort this week.

From that point, the focus will turn to Nvidia’s earnings, which will come in late May.

With that, the Trump administration has recently blocked Nvidia’s exports to China, which is thought to be about a 10%-15% revenue hit.

But remember, the headwind for Nvidia is supply, not demand.

On that note, Nvidia announced two weeks ago that new capacity for its most advanced chips has just started production in Arizona, and production in Texas is due in 12-15 months.  That will add to global chipmaking capacity which means Nvidia will be able to fulfill more of the demand backlog.

Amazon’s CEO (Andy Jassy) has said that the AI business could be growing faster if not for chip supply constraints.  So, this U.S. onshoring of chipmaking will effectively press the accelerator on the AI revolution.

With all of this, Nvidia remains the most important company in the world, which makes the trend in this chart below (sparked by the “ChatGPT moment”) the most informative about the current stock market environment:  it’s a correction within a long-term structural trend.

Relative to the ChatGPT moment, Nvidia is growing revenue three times faster, is twice as profitable, and yet its shares are cheaper (forward P/E of 24).  Meanwhile, Nvidia’s market position has never been more dominant, and demand has never been more insatiable.

Trump has been turning up the heat on Jerome Powell to restart the easing cycle.

With that, let’s revisit our discussion from earlier this month on the influence of central bank action on turning points in markets. 

Remember, we looked at this five-year chart of the S&P 500 in my April 3rd note …

Each turning point over the past five years was driven by some degree of central bank “dial turning” (i.e. monetary policy adjustment).

And that’s consistent with the history of major stock market turning points — they tend to be directly influenced by central banks.

The most recent turning point (the top in the S&P 500, the global barometer of risk-asset sentiment) also coincided with a central bank event:  a pause in the Fed’s four-month old easing cycle.

So, will the next turning point for stocks be marked by a central bank event?

Likely.

The bigger questions:  

>Will the eventual Fed action come in acknowledgment of their mistake, in misjudging the inflationary outcome of tariffs (which is demand-destructing)?  

>Or will the eventual Fed action come when something breaks in the financial system, resulting from their overly tight policy (which includes extracting $2 trillion of liquidity from the system over the past two years).

Keep in mind, the Fed’s job is to keep the ship in the channel (price stability and full employment), regardless of external forces.  

But the Fed tends to be reactionary — which means they tend to be late.

And because major central banks, generally, are late to adjust policy, they become a primary contributor to market excesses — in both directions.

With all of the above in mind, what did Chicago Fed President Austan Goolsbee say in the midst of the market correction back in August?  “If the economy deteriorates, the Fed will fix it.”

Yesterday, we talked about what seems to be a “Mar-a-Lago Accord” in the offing.

It’s been fourteen days since “Liberation Day” (reciprocal tariff announcements).  And it has been said repeatedly by the Trump administration that over 75 countries are clamoring to do a deal. 

But no deals done. 

Why?

Are they planning on doing a grand coordinated deal, all at once (and probably over a weekend)?  Maybe.

What would it look like, based on what’s been guided by key Trump advisors:  Tariffs get slashed, in exchange for countries opening up their markets, boosting their defense spending, and committing to buy more from the U.S., invest in American manufacturing, and buy our Treasuries — and isolate China.

If we look at the behavior of gold and the dollar, the market seems to be sniffing out such a deal, to include an agreement to devalue the dollar. 

Gold is up 13% since Trump’s 90-day pause on tariffs just one week ago.  The dollar is down 4%. 

That said, in a prepared speech last week, Trump’s top economist (Stephen Miran) dismissed the conventional economic view that trade deficits get fixed, naturally, through currency depreciation.  Not when you have the world reserve currency.   

Instead, he views tariffs as the solution — not currency manipulation. 

 

While stocks have sharply corrected under tariff fears and the Fed’s missteps on policy, make no mistake the AI revolution continues to charge forward.

Over the weekend, the Trump administration signaled exemptions for critical AI inputs from China, while Nvidia announced it has started production on its most advanced AI chips in Arizona, and will be building “supercomputers” in Texas in 12-15 months. 

This “onshoring” is new global capacity, and that will be like pressing the accelerator on the AI revolution.

And with that in mind, one of the handful of people working on the frontier of this technology revolution said this about the state of AI back in February … 

Elon is talking about teetering on the edge of artificial general intelligence (AGI).  It’s where the machine can reason like a human or better – in any domain.  The next stage, it can constantly improve itself and with no downtime.

Related to that, remember, earlier this year Jensen Huang (Nvidia founder, CEO) said “the ‘ChatGPT moment’ for general robotics is just around the corner.“

He said over the next several years, the combination of 1) agentic AI robots, 2) autonomous cars and 3) humanoid robots will become “the largest technology industry the world has ever seen.”

So, as the crowd obsesses over tariffs, tight money and political mudslinging, the real story is relentless ascent of world changing technology.  

And as we’ve discussed often here in my daily notes, there is a wartime-like effort, by the U.S., to win the AI arms race.

With that, if you can look through the noise and pushback surrounding the Trump agenda, this broad market correction is an opportunity to buy the key stocks that are powering the AI revolution, on sale.  

But isn’t this another tech boom/bust cycle?

For perspective, the AI revolution is as transformative, if not more, than the advent of the internet, but there are clear differences in the investment.  

The dot-com bubble was fueled by a speculative mania. 

Money was being poured into companies with little-to-no revenue (like a few hundred thousand dollars of revenue) – and valued on the basis of eyeballs (web traffic).

On the other hand, as a proxy on the AI revolution, if we step through our AI-Innovation Portfolio — we’ve built a 24 stock portfolio since launching in June of 2023, and ALL but one are profitable — not accounting tricks, legitimate net income.

That’s a big difference.   

And the one company that isn’t profitable, is expected to be profitable this fiscal year.

And all have positive net cash (most in the billions of dollars).

This is a very different technology boom, and it’s still early.

If you want to own the stocks of the companies building the infrastructure to power AI, the companies delivering the capabilities of AI to hundreds of thousands of businesses, and the companies that will best leverage the productivity enhancements from AI, you can find them in our carefully curated AI-Innovation Portfolio.

Consider this:  We added Nvidia in June of 2023, just when the AI revolution was kicking off.  We exited on the catalyst of the 10-for-1 stock split announcement, after the announcement fueled a further surge in the stock. 

So, we sold it last May for 184% return, at a P/E in the high 40s.  

We recently re-entered Nvidia at a forward P/E of 21, down a third from its peak, trading back to its pre-split level — BUT, the company is now producing twice as much revenue, and twice as much profit, and the state of the AI-driven technology revolution is far more advanced than it was just 11 months ago.

So, this correction in stocks revalued Nvidia to a multiple very near the broad S&P 500 market multiple.  That’s extremely cheap for a company with 78% year-over-year revenue growth and operating margins greater than 60%, with $35 billion of net cash. 

The best investors in the world have built their careers in markets like this, buying when others are selling.  

This is an opportunity to buy the technology revolution on a dip.

Here’s how you can join me…  

✅The AI-Innovation Portfolio is about allocating to HIGH-GROWTH.

✅For $297 per quarter ($99 per month), you’ll gain exclusive access to my in-depth research, expert analysis, and timely investment recommendations focused on the generative AI revolution — all email delivered to your inbox. 

✅You can join me by clicking here — get signed up, and then keep an eye out for Welcome and Getting Started emails from me.

Best,

Bryan