These AI stocks are hitting new all-time highs

- By Stephen McBride, Chief Analyst at RiskHedge

Hey friends, all is quiet here in Abu Dhabi.

I was in the US for two weeks, and while I was gone, the city flooded while the kids were stuck at home remote learning. The missiles were the least of my wife’s worries!

This chart from my friend Marko Papic over at BCA Research shows the attacks are down “bigly” since the war began.

Source: BCA Research

The S&P 500 and Nasdaq rebounded in recent days but are still down on the year.

What if I told you there’s a group of stocks making all-time highs right now?

The most undervalued artificial intelligence (AI) stocks…

That’s what I said when I first talked about optical stocks back in December.

Despite the war, tariffs, and AI uncertainty, this group is ripping higher.

Applied Optoelectronics (AAOI) has surged 280% since January. Lumentum Holdings (LITE) has soared 140%. Ciena Corp. (CIEN) has more than doubled. And Corning (GLW) is “only” up 90%. Talk about bucking the trend!

You might be wondering...

“What are optical stocks?”

You know big tech giants are spending record amounts of money building AI data centers right now.

Inside these massive data centers, there are hundreds of thousands of AI chips (GPUs) constantly “talking” to each other. This is how AI models are trained and run.

More than nine-tenths of the time an AI model takes to respond to a query is spent shuttling data back and forth between chips. If the pipes between GPUs are slow, the chips sit idle. A $40,000 GPU doing nothing is the most expensive paperweight in tech.

These chips used to be able to communicate with each other through copper wire. But as the number of AI chips has grown (the compute needs of top AI models has surged 1 billion percent since 2010), so too has the need for a faster way for those chips to communicate.

Now we’re ripping out the copper wiring and replacing it with fiber optics.

Optical cables always lived outside the data center...

Think submarine cables, coast-to-coast fiber, and telecom towers.

AI is dragging optics inside the data center: into the racks, the network fabric, and even the links between GPUs.

In Nvidia’s (NVDA) latest GB200/NVL72 systems, the optical components alone can amount to more than $500,000 per rack.

For years, fiber was a boring, cyclical business with terrible margins and terrible customers. It was the poster child for the dot-com blowup. Remember all that “dark fiber?”

But fiber does one thing extraordinarily well. Instead of sending data as electricity over copper, fiber turns data into laser light and shoots it down glass threads thinner than a human hair.

Corning's CEO put it perfectly: Over short distances, photons are 3X more efficient than electrons. Over long distances, it's more like 20X.

If you want tens of thousands of GPUs to act like one brain, optics is the only way.

As hyperscalers build bigger AI clusters and push network speeds higher, more copper connections get replaced by optical ones.

The big takeaway here is that AI is the megatrend of the decade.

If you want to make money from AI, you must own the winners.

The great thing about AI is that it’s like a big, hot ball of money minting new winners every few months. Thankfully, my team and I have been on the AI “gravy train” since 2018, when we first recommended Nvidia.

The first wave of winners was obvious: GPUs. When the world ran out of graphics processing chips, Nvidia exploded more than 1,000%. It was the only company supplying the shovels in a gold rush.

Then AI hit its next limit: power. Training giant models requires enormous amounts of electricity. That pushed utilities—one of the dullest sectors in history—to the top of the performance charts in 2024.

Next came memory. As models grew from billions to trillions of parameters, the system needed far more high-bandwidth memory. Micron Technology (MU) and SK Hynix surged as AI companies bought every memory chip they could produce.

Now the bottleneck is shifting again.

Tens of thousands of GPUs linked together only create value if they can exchange data fast enough to behave like one machine. That’s why fiber optics stocks are hitting new highs despite the fact that the stock market is struggling.

It helps to think of AI like a python swallowing a pig. The bulge moves slowly through the system, from one bottleneck to the next. First chips, then power, then memory… and now, data flow.

The tidal wave of AI spending doesn’t hit one corner of the market and stop. It keeps rolling on, lifting entire sectors investors had written off decades ago.

We think the next major AI bottleneck is…

The companies that make the chipmaking machines.

ASML Holding NV (ASML), the first company we ever recommended, is the grandaddy of this sector. But there are so many more opportunities.

These stocks are about to make big moves because most of the current machines can't produce the cutting-edge chips the world needs by the millions. They need to be extensively upgraded.

Billions of dollars in new investment will pour into this space in 2026. It’s already started.

That’s why I believe semi-caps are the next big opportunity in AI.

______________

Stephen McBride is Chief Analyst at RiskHedge. To get more ideas like this sent straight to your inbox every Monday and Friday, make sure to sign up for The Jolt, a free investment letter focused on profiting from disruption.