Every time I tell someone that Nvidia is the most valuable company in the world, they nod. Everyone knows that by now. It was the first company ever to cross the five-trillion-dollar mark, and it sits at the top of the list, ahead of Apple, ahead of Google, ahead of Microsoft. But when I add that this same Nvidia is also one of the cheapest stocks in the market, I get a confused look back. How can a company worth almost five trillion dollars be "cheap"?
And the whole story starts right there, in that little bit of confusion. Because "an expensive company" and "an expensive stock" are not the same thing, and most people have never stopped to think about the difference. This week we're going to stop on it, because once you understand it, you can look at any stock in the world and decide for yourself whether it's expensive or cheap. Not by feel. With a tool.
First, what is "market value," or market cap? It's simply the size of the company — what the whole thing is worth if you bought all of it. Nvidia is enormous, the most valuable on earth, and deservedly so: it sells the chips that the entire artificial intelligence revolution runs on. If you remember the letter about the AI gold rush (Letter #07), we talked there about the people who sold the picks and shovels to the gold diggers instead of digging themselves. Well, Nvidia is the biggest seller of picks and shovels the world has ever seen.
But here's the first thing you have to absorb: the size of the company tells you nothing about whether the stock is expensive. Those are two completely separate questions. And to answer the second one, professionals have one simple tool they look at before anything else. It's called the price-to-earnings ratio.
Let's explain it without a single piece of jargon. Imagine you're buying a hot dog stand. The stand makes 100,000 dollars a year, clean. The owner wants 3,000,000 dollars for it. How many years of profit are you paying to buy it? Thirty. You're putting thirty years of profit on the table up front. That's the price-to-earnings ratio — how many years of current profit you pay for the business. This stand trades at a multiple of 30.
Now imagine that further down the street there's a second stand, nearly identical, that also makes 100,000 a year. But they're only asking 1,000,000 dollars for it. Ten years of profit. The second stand is far cheaper, right? Same profit, a third of the price. And that's exactly the point: both stands have a "for sale" sign, but one is three times as expensive as the other — not by the sticker price, but by what you actually get for your money. The sticker alone tells you nothing. What tells you something is the ratio between the price and the profit.
Back to Nvidia. Its current multiple is about 30, exactly like the expensive hot dog stand. Thirty years of profit. At first glance, expensive. So what am I talking about when I call it cheap?
This is where the most important distinction in this letter comes in, and it's the difference between the current multiple and the forward multiple. The multiple of 30 is calculated on Nvidia's profit over the past year. But Nvidia is not standing still. Its profit is growing at a pace that's hard to grasp: its revenue jumped about 65 percent this year, and in the last quarter alone sales rose roughly 85 percent compared with a year earlier. When profit grows that fast, next year's profit is expected to be far bigger than last year's. And when you divide the same price by the bigger, forward-looking profit, the multiple drops. Nvidia's forward multiple is about 20.
Go back for a second to that expensive hot dog stand, the one at a multiple of 30. Suppose you discover that its profit doubles every year, because a new tech campus just opened next door and the lines keep getting longer. In two years it'll already be making several hundred thousand a year, and the price they asked, three million, suddenly doesn't look like thirty years of profit but more like four or five. The "expensive" stand turned out to be a pretty good deal, not bad at all. That's exactly what the forward multiple is trying to capture: not how much you earned yesterday, but how much you're about to earn.
And now for the last piece of the puzzle, without which you can't tell whether any number is expensive or cheap: comparison to similar companies — what's called the industry, or the sector. Is a multiple of 20 expensive or cheap? There's no answer to that in a vacuum. You need something to hold it against. So here it is: the rest of the world's chipmakers trade at a median forward multiple of about 38. Nvidia, at 20, trades roughly 40 percent below its own sector. And even within the club of tech giants, the "Magnificent Seven" (Apple, Microsoft, Alphabet, Amazon, Nvidia, Meta, and Tesla), it's one of the cheapest there is.
And here the paradox finally resolves. Nvidia is both the most expensive company in the world — by size — and one of the cheapest stocks in the market — by its forward multiple against its sector. Both facts are true at the same time, because they aren't measuring the same thing at all. One measures how big the company is. The other measures how much you pay for every dollar of profit it will produce.
But I wouldn't be EcoMan if I stopped here and let you feel clever. Because there's a catch, and it matters. The low forward multiple is not a gift and not a free lunch — it's a bet the market has already made, the bet that Nvidia's profit really will explode upward the way everyone expects. Notice what happened here: the risk didn't disappear, it just moved. It shifted from the price to the profit. You're no longer at risk of having paid an inflated price; you're at risk that the future profit won't show up.
Let's be precise for a second, because this is where it's easy to get confused. Nvidia's current multiple, 30, isn't really a cheap multiple, even if it's below its sector. What's actually cheap is the forward multiple, 20. But notice what that 20 is: it rests on profit that hasn't happened yet, profit Nvidia is expected to produce but hasn't produced. In other words, the buyer is getting a cheap multiple — but cheap on the forecast. And that's exactly the trick: it's a cheap that comes with an asterisk. It's cheap on condition — on the condition that Nvidia really does keep growing the way everyone expects. If the growth shows up, it'll turn out you paid a little for a lot. If it doesn't, that pretty 20 evaporates, and the price you're paying today will suddenly look very expensive.
What does this mean for your pocket?
If you own a fund that tracks an index — one that follows the S&P 500 (the 500 largest companies in America) or the Nasdaq — you already hold a nice slice of Nvidia without ever buying it directly. It's the single heaviest weight in those indexes. What moves in it, moves in your portfolio, whether you noticed or not.
If you've ever looked at a stock's price and said "too expensive," or "cheap, it's only a few dollars" — it's worth knowing that this simply tells you nothing. The price alone won't tell you whether the stock is expensive or cheap. A stock at 1,000 dollars can be cheap, and a stock at five dollars can be very expensive. What decides is the multiple, not the number on the screen.
If you're comparing two stocks to decide between them — don't compare prices, compare multiples. Preferably the forward one, and preferably against that company's own sector, not the whole market. Apple to apple, not apple to watermelon.
And if you hear on the news that "the stock is historically expensive" or "it's a bubble" — ask immediately: expensive against what? Against yesterday's profit or tomorrow's? Against which companies? Most of the time you'll find that the dramatic headline melts away the moment you put those three questions to it.
So here's the tool, in short, for every time someone tells you a stock is expensive or cheap: first, the multiple, not the price. Second, the forward one, not just the current one. Third, against its sector, not in thin air. Three questions, and you already know more than most of the people talking about the market over dinner.
The real question with Nvidia, then, was never whether the stock is expensive. It's whether the growth will hold. The next quarter, which Nvidia expects to bring in about $91 billion in revenue, will tell us whether the market was right to price the growth in advance. If the numbers arrive, the "most expensive in the world" will keep looking cheap. And if they don't, we'll find out how fast "cheap" turns back into "expensive." That's exactly what we'll be watching.
We shall see.
See you in next week's letter, or sometime during the week, if I can't hold out until then...