Can't Classify It? Maybe It's a Rule Breaker Stock

David Gardner just floated one of his favorite investing signals on X, and the striking thing is how little it has aged. His idea: When the experts can't agree on what industry a company belongs to, they may be "looking at tomorrow through yesterday's taxonomy."

One of my favorite investing signals:

If the experts can't decide what industry a company belongs to...

...they may be looking at tomorrow through yesterday's taxonomy. 😎🌈

(From more than 25 years ago.)#RuleBreakerInvesting pic.twitter.com/wdZrsugiI2

Where to invest $1,000 right now? Our analyst team just revealed what they believe are the 10 best stocks to buy right now, when you join Stock Advisor. See the stocks »

-- David Gardner (@DavidGFool) July 21, 2026

The wink at the end -- from more than 25 years ago -- is the point. He first wrote a version of this in Rule Breakers, Rule Makers, back when the internet was still a single, confused category on Wall Street's spreadsheets.

That passage is worth sitting with. David describes how the first great internet brands were bundled into a catch-all bucket labeled, unhelpfully, "Internet." America Online, Yahoo!, Netscape, and Amazon (NASDAQ: AMZN) all landed in the same pile -- a service provider, a search engine, a software maker, and a bookseller -- treated as peers because nobody had a better box for them. Was Amazon a bookseller, to be measured against Barnes & Noble? Or an internet-commerce company, to be measured against long-forgotten names like CDNow? The honest answer was neither, and that was the tell.

The Confusion is the Opportunity

Team Rule Breakers

His counterintuitive conclusion: The confusion is the opportunity. When you struggle to place a company, it's often because the company is inventing its own category -- which is precisely the first of David's six traits of a Rule Breaker stock: a top dog and first-mover in an emerging industry. The others tend to travel with it. Getting there first can harden into a sustainable competitive advantage, and a business the market can't price is one the financial media loves to call wildly overvalued. The classification problem and the "overvalued" verdict are often the same symptom wearing two hats.

Hindsight has been kind to the argument. Amazon never settled into the bookseller box; it built cloud computing, logistics, and advertising businesses the 1990s taxonomy had no words for. The analysts reaching for a comparison were, quite literally, describing the future in the vocabulary of the past -- and the investors who did best were the ones who let that winner run instead of trimming it every time it looked expensive. It is the working case for two of David's habits: hold for at least three years and add to your winners rather than doubling down on your losers.

The Same Puzzle Keeps Returning in New Clothes

Team Rule Breakers

The tweet resonates now because the same puzzle keeps returning in new clothes. Think of the companies today's commentators can't quite file: is Tesla (NASDAQ: TSLA) a carmaker, an energy company, or a robotics bet? The debate now is whether Tesla is a "story stock" or a "story-plus-data stock": the automotive business is measurably better than a year ago, and the new businesses are producing real data for the first time, but the bull case no longer rests on Model Y deliveries -- it rests on whether the Austin robotaxi network can scale economically and whether Optimus can generate real revenue before the end of the decade. Moreover, the market can't agree whether Optimus is a product line or a call option on the future of labor...

Elsewhere, investors are asking whether Nvidia (NASDAQ: NVDA) is a chipmaker, an AI-infrastructure platform, or something without a name yet. Jensen Huang has more or less stopped calling it a chip company at all, preferring to describe data centers as "AI factories" that convert energy into intelligence -- a category the 1990s spreadsheets, and most of today's, have no column for. When a semiconductor maker starts sounding like a utility, a software platform, and an industrial supplier at once, "chipmaker" is plainly the label that's out of date.

And is Palantir (NASDAQ: PLTR) a software company, a defense contractor, or a data analytics consultancy? This one is stranger still, because it's arguing about its own box: A company long filed under "defense contractor" now insists it's an enterprise software platform, even as its identity stays welded to intelligence agencies, the Pentagon, and border enforcement. Here, the disagreement isn't only among analysts -- the company and its critics can't agree on what they're looking at, either!

The Disagreement Itself is Information

Team Rule Breakers

Notice how the other Rule Breaker traits cluster around these names too -- fierce brand loyalty, dramatic past price appreciation, backers the market takes seriously -- and how dependably each has been branded overvalued along the way. The disagreement itself is information. A business that fits neatly into an existing sector is, almost by definition, competing on someone else's terms. One that breaks the filing system may be writing new rules.

None of this is a formula. Plenty of hard-to-classify companies are merely muddled rather than visionary, and a category-of-one can still turn out to be a category-of-none. That is why the habits matter as much as the traits: David caps a new position near 5% of a portfolio and expects to be wrong on roughly 40% of his picks, trusting the rare giant winners to more than cover the misses. The signal is a lens, not a screen -- a prompt to look harder at the thing that doesn't fit, not to buy it for failing to match a template.

Which may be the most durable idea in the whole post. Taxonomies describe the world as it has been; the best investments tend to describe the world as it is about to become. When the experts start arguing about which box to use, it's worth asking whether the box is the problem.

The Motley Fool has positions in and recommends Amazon, Nvidia, Palantir Technologies, and Tesla. The Motley Fool has a disclosure policy.