Spend five minutes with the average investor and all you’ll hear are the same three lines on repeat. AI stocks are in a bubble. Semiconductors are uninvestable. Value investing is dead. Fair enough, when companies tied to artificial intelligence keep printing new highs while stocks trading at 8 or 10x earnings keep making new lows, it’s easy to conclude the market has lost its mind entirely.
I’d argue the market hasn’t lost anything. And value investing hasn’t stopped working, but the problem is that we’ve spent so long hunting for cheap stocks that we forgot what value actually means in the first place.
Intel, Nvidia, And Why Investors Should Think Differently
Three years ago, value investors thought they’d found an obvious mispricing. NVIDIA Corp (NASDAQ: NVDA) traded at a multiple they considered absurd. Intel Inc (NASDAQ: INTC) traded at a fraction of it, threw off billions in profit, paid a healthy dividend, and showed up on every screener built to find bargains. Buy Intel, avoid NVIDIA, the trade wrote itself.
Then the AI boom hit, and the math completely inverted. Hyperscalers poured hundreds of billions into infrastructure, demand for accelerated computing exploded, and NVIDIA kept selling every chip it could physically manufacture. The valuation everyone feared got easier to justify with each passing quarter, because the business underneath it refused to stop compounding. Intel lived the opposite story in real time – AMD chipped away at the share it once considered untouchable, the manufacturing lead disappeared, the AI cycle arrived, and Intel showed up late to it, and investors kept finding fewer reasons to believe tomorrow would beat today.
One company looked expensive because the market believed its best years were still ahead. The other looked cheap because the market had quietly started pricing in the opposite. That’s not a coincidence or a temporary mispricing correcting itself. In fact, that’s exactly how markets are supposed to work, and Intel versus NVIDIA is the cleanest demonstration of it in a generation.
The Cheapest Stock On Your Screener Might Be The Most Dangerous One
Every investor loves finding a bargain. The mistake is assuming a bargain always arrives wearing a low P/E. Walgreens Boots Alliance Inc (NASDAQ: WBA) spent years proving exactly how wrong that assumption can get. Investors pointed at its single-digit earnings multiple and fat dividend as evidence the market had mispriced a stable business. Meanwhile, reimbursement pressure kept compressing margins, store traffic kept eroding, Amazon kept pushing deeper into healthcare, and management eventually announced plans to close hundreds of locations.
The low multiple was never the opportunity. It was the warning, the market pricing in deterioration that most retail investors hadn’t noticed yet, because they were too busy admiring the dividend yield to look at why it was being offered in the first place.
That’s the blind spot a P/E ratio creates on its own. I’ve spent tons of hours studying 30+ earnings reports in the past 2 months and what I’ve discovered is that two companies can report identical earnings in one quarter and deserve completely different valuations, because one is compounding into a stronger competitive position and the other is quietly losing relevance underneath a number that still looks fine.
Wall Street was never pricing last quarter’s results. It’s always pricing what the next several years are likely to look like, whether the multiple agrees with that or not.
Follow The Money, Not Where The Headline Points
While flipping through the pages of most reports, scouring opportunities in the AI ecosystem, one line I never forgot for a second was: “Don’t forget construction and infrastructure.”
Because the truth is, most investors hear AI and think NVIDIA. But the market is pricing something considerably bigger than that. Every data center has to be designed before a single GPU goes in it. It needs power before it needs compute. It needs transformers, cooling systems, switchgear, fiber, and mechanical infrastructure built out long before a chatbot answers its first query – and that buildout is exactly why names like Vertiv have quietly become some of the largest beneficiaries of this entire cycle without building a single model themselves.
My point is, the money behind a transformative technology never stops at the headline company. It spreads through every business that makes the headline company’s existence physically possible. It happened during the railroad expansion, the smartphone buildout, and it’s happening again right now while investors keep arguing over whether NVIDIA’s or even SpaceX’s multiples are too rich. Time is better spent finding where the next layer of value is actually accumulating instead of doubting if “value investing” is getting phased out.
Value Investing Was Never The Problem
Blaming the market for irrationality is the easy move. Questioning whether our own definition of value has gone stale is the harder one, and it’s the one that actually pays. Markets have always rewarded businesses capable of expanding earnings power faster than consensus expects. Sometimes that business trades at 8X earnings, sometimes at 35, and the multiple was never the variable doing the deciding. The business was.
Intel and NVIDIA made that lesson impossible to ignore any longer. One looked like the disciplined, defensible choice. The other looked reckless to anyone still anchored to a spreadsheet. Years later, the investors who bought the cheaper stock weren’t rewarded for their discipline. Those who correctly identified which business was becoming more valuable were. So no, value investing isn’t dead. The definition most investors have been using for it expired a while ago, and the market has been waiting patiently for everyone else to catch up.
