AMD is becoming a data-center company. The chip designer reported $13.4 billion in quarterly sales, beating Wall Street's estimate. Sales of data-center chips nearly doubled from a year earlier to $8.5 billion. A year ago that segment was 47% of revenue. Now it's 64%. The stock rose 8.5% after the report. And yet.
Here's the whiplash. AMD's stock is up nearly 120% this year. It's also down 23% over the past month, as fears about an AI infrastructure bubble intensified. Same company, same quarter of strong results, two completely different stories depending on where you start counting. Big partnerships with AI companies and strong new products drove the year's run. A wave of worry about overspending drove the month's drop.
The megadeal and the charge
The quarter included a $1.1 billion accounting charge tied to a megadeal struck last year with an AI company. Under that deal, the partner committed to pay tens of billions of dollars to buy up to 6 gigawatts of computing power using AMD's most advanced chips. Deals like that are why the stock soared. Their circular nature, AI companies and chipmakers leaning on each other, is exactly what cautious investors like Elliott have flagged as a risk. AMD's results were genuinely strong. The debate isn't about this quarter. It's about whether the AI spending that drives it can last.
We hold AMD in one of our model books, at a small weight. Our model books changed little this week: a CoreWeave trim and two small swaps, nothing in the North Star. A name this volatile earns a small position, not a big one.
Our read
This is an Investments and Risk story, module IN04, and AMD is a perfect teacher because it's the same stock telling two stories. Up 120% for the year is the story a buyer from January tells. Down 23% in a month is the story a buyer from four weeks ago tells. Both are true. Which one you live in depends entirely on when you bought and how much you own. That's not a fact about AMD. It's a fact about position size and your own nerves.
The practical lesson is the one we keep landing on, because the market keeps teaching it. A stock that can move 23% in a month, up or down, belongs in a small slice of a portfolio, not a big one. The math is simple. If a name is 2% of your money and it halves, you lose 1% of the whole. If it's 20% and it halves, you lose 10%, and more importantly you panic and sell at the worst moment. Here's the check this week. Find your most volatile holding, the one whose price you check most nervously. Figure out what percentage of your total it is. If a 50% drop in that one name would genuinely change your life, it's too big, and the fix is to trim it toward a size you'd barely feel. We hold AMD small on purpose. If you want help right-sizing a volatile position, that's fifteen minutes.
