What the heck is happening to AI stocks?

My Micron (MU) position is down 13% in just over a week.

Nearly every other AI-related stock has sold off as well.

Nvidia… Broadcom… Alphabet… Bloom Energy… Corning.

All have been caught in the downdraft.

But the interesting part?

The fundamentals haven’t changed.

In fact, I'd argue the opposite.

Here’s why I’m sticking with these stocks.

Reason #1: Earnings Estimates Continue Moving Higher

Whenever one of my investments pulls back, the first thing I check is earnings estimates.

Not the headlines.

Not social media.

If Wall Street starts cutting revenue and earnings estimates, that's a sign something may actually be wrong.

But that's not what’s happening here.

Check out how earnings estimates have moved higher over the last two months.

Below are Wall Street's earnings estimates for next year from 60 days ago, 30 days ago, and today:

Micron (MU): $102.72 → $117.95 → $153.74

Nvidia (NVDA): $12.65 → $12.73 → $12.87

Intel (INTC): $1.54 → $1.55 → $1.99

Alphabet (GOOG): $14.44 → $14.50 → $14.72

These aren’t cherry-picked.

You name an AI company and estimates are likely moving higher.

So while the stocks are going down, it’s important to know that the expectations aren’t.

Reason #2: The Biggest Customers Keep Spending

Ultimately, the fate of AI stocks comes down to one thing:

Are the world's largest AI companies still spending?

So far, the answer is a resounding yes.

Last week, Alphabet raised its full year 2026 Capex guidance range to $195 billion to $205 billion, up from their previous estimate of $180 billion to $190 billion.

Tonight, Microsoft and Meta report earnings.

My prediction is that both companies also raise their AI spending forecasts.

Looking specifically at memory, demand remains unprecedented despite the recent negative sentiment.

Deloitte released an article with this title yesterday: “Why the memory chip crunch is greater than expected, and may not ease until 2029”

And Elon Musk literally stopped the Tesla earnings call last week to say this:

“I'd also like to thank Micron for giving us the memory allocation. They're going to make some very tough decisions on memory allocation, and we really appreciate Micron making room for Tesla in the years to come and giving us actually a very significant allocation on reasonable terms given the pretty insane pricing of memory these days.”

To me, this is far more important than a few days of negative price action.

Reason #3: I’m Not Losing Sleep Over China

Seemingly the biggest reason memory stocks have sold off is renewed concern about China.

A Chinese memory manufacturer recently went public. Another Chinese company announced plans to manufacture memory equipment.

Suddenly, investors are acting as if Micron woke up with major new competition.

This reasoning is just flat out wrong.

CXMT is the name of the Chinese memory manufacturer that recently went public.

The company was founded in 2016, so it’s not new to the chip market… just the stock market.

Plus, the company faces a long list of restrictions on international sales because its investor list is a who's-who of Chinese government entities.

And then there’s the fact that we've seen this movie before.

In 2023, China's largest semiconductor company, SMIC, announced it had produced an advanced chip despite U.S. export restrictions.

Investors immediately worried China had caught up to the West.

It didn’t.

The chip cost significantly more to produce and wasn't competitive with the best products coming out of Taiwan and the United States.

Could China become a formidable competitor five or ten years from now?

Absolutely. This is a risk worth monitoring.

But that's a very different question than whether Micron's competitive position fundamentally changed over the past week.

The Real Test is Tonight

The most important question isn't what China does… or how investors are feeling on any particular day.

It's what company management teams say about demand and AI spending, starting with Microsoft and Meta this afternoon.

These companies are two of the largest buyers of AI infrastructure in the world.

If they continue talking about expanding data centers and increasing AI investment, it reinforces the demand story for memory and the broader AI ecosystem.

That's why I'll be paying much closer attention to capital expenditure guidance than a few days of falling stock prices.

Because while stock prices can change overnight, business fundamentals rarely do.

And right now, the companies writing the biggest AI checks don't appear to be slowing down.

If anything, they're speeding up.