The financial media has found another stock you've supposedly never heard of, and this time the claim is a big one: an optical laser company that could outgrow Nvidia. The argument goes that GPUs are a great business, but the optical components shuttling data between those GPUs may be a better one. Part of that logic holds up. Part of it is just a headline.

The part that holds up: AI data centers move staggering amounts of information between chips, and copper wire runs out of room at high speeds and short distances. Optical interconnects, the lasers and transceivers doing that work, are genuinely a growing business. Demand is real. The companies making these parts have booked real orders from real customers.

What gives me pause is the framing. "Poised to outgrow Nvidia" is doing a lot of work in that sentence, and most of the work is on your imagination. Nvidia grew revenue by triple digits in its hottest stretches. Outgrowing that from a small base is arithmetic that flatters small numbers. A $2 billion revenue company doubling sounds more impressive than a $130 billion company growing 20 percent, but guess which one added more actual dollars. Growth rates from small bases are the oldest trick in the promotion playbook.

I've been doing this since the late '80s, and every hot sector produces its "picks and shovels" stories. Some were right. The companies selling network gear during the dot-com boom had genuine demand too, right up until their customers stopped ordering and the whole supply chain seized. In 2001, the picks-and-shovels argument didn't save optical networking stocks from falling 90 percent. The demand was real. The price paid for it was the problem.

That's the quiet issue with any "under-the-radar" pitch. If a stock is genuinely under the radar, ask why the radar missed it. Sometimes the answer is that the market is slow. More often the stock already ran, the story is late, and you're being handed the baton right before the hill. Articles like this one tend to appear after a nice run, not before it.

If you already own a total market index fund, you own Nvidia, you own the laser makers, and you own whichever supplier wins the next round. You didn't have to pick. If you want a specific position here, treat it as a single-stock bet: size it like money you could watch drop by half without flinching, because component suppliers to a capex boom are cyclical even when the boom isn't a bubble. Ask yourself what these earnings look like if data center spending merely flattens for a few quarters. Not collapses. Just flattens.

There will be winners in optical. There usually are. The question isn't whether the business grows; it's whether you're buying at a price that already assumed the growth. That question doesn't fit in a headline.

As always, this is my perspective, not professional financial advice. Do your own math before your own money follows someone else's headline.