For most of this year, the AMD story has been simple. Beat estimates, rip higher, repeat.

That pattern held through six straight quarters of acceleration and a 142% year-to-date run.

Then last week AMD posted its best quarter yet, record revenue, data center sales more than doubling, and the stock got sold hard anyway.

Not because the business slipped. Because the price already assumed perfection, and a very good quarter isn't the same thing as perfection.

That's what One Big Idea digs into today: what happens to a great business once its own stock price becomes the biggest risk to owning it.

Missed Friday's Playbook on Palantir's 29% surge?

Greg Abel is starting to show his own capital-allocation style at Berkshire. Berkshire’s cash and Treasury pile fell from about $380 billion at the end of March to $365 billion at the end of June, while the company bought back roughly $4.5 billion to $4.6 billion of its own stock and put nearly $20 billion net into equities, including Alphabet. In plain English, Abel is still patient, but he is no longer just sitting on Buffett’s cash mountain.

Johnson & Johnson remains one of the steadier income names in the market. J&J raised its quarterly dividend by 3.1% to $1.34 per share, marking its 64th straight year of dividend increases, which keeps it firmly in Dividend King territory. The stock may not be exciting, but for income investors, the appeal is exactly that: durable healthcare cash flow, a long dividend record, and a yield that still sits above the broader market.

Microsoft’s earnings reaction has held up better than feared. Recent earnings data showed Microsoft beat expectations with $4.27 in EPS versus a $4.06 consensus, while full-year earnings are still expected to grow strongly next year. The bigger point is that investors may still worry about AI capex, but Microsoft’s core earnings power, Azure momentum, and enterprise AI positioning are giving the stock enough support to avoid a deeper post-earnings hangover.

That's what Warmly proved. They defined ICP, scored buying intent, and surfaced the right accounts before a human ever touched a lead. HubSpot noticed.

On August 12, Max and Keegan are rebuilding it live in HubSpot — and showing you how to replicate it this week. HubSpot Credits included when you join HubSpot for Startups.

Grab Your Spot Now

AMD reported record revenue of $11.54 billion last week, up 50% year over year, and beat Wall Street on both the top and bottom lines. The obvious read is that the stock should have ripped higher.

Instead it fell nearly 9% in after-hours trading and kept sliding the next day, closing the week around $483, down from a pre-earnings high near $519.

AMD didn't miss. It just didn't blow the doors off enough for a stock that had already run 142% this year heading into the print.

The real question this week isn't whether AMD's business is fine. It clearly is. The question is what it actually takes to move a stock once the entire market already agrees it's a winner.

Data center revenue more than doubled to $6.7 billion, up 107% year over year, and now makes up 58% of AMD's total sales. Q3 guidance came in around $13 billion, implying 41% growth, well above where AMD was modeling just two quarters ago.

The customer list backing that growth is real. Anthropic has committed to up to 2 gigawatts of MI450 GPUs running in AMD's Helios server racks, OpenAI has its own 6 gigawatt arrangement, and Meta is designing custom silicon based on the MI450 architecture.

CEO Lisa Su put it plainly on the call: EPYC demand is accelerating, Instinct deployments are scaling, and Helios is just starting to ramp. This is a company entering the part of the cycle where deals signed a year ago start showing up in the revenue line, not one running out of momentum.

Here's what the bulls don't want to hear. AMD walked into this print with a put to call ratio of roughly 0.60, meaning options traders were positioned overwhelmingly for a blowout, not just a beat.

Short interest sits at just 2.44% of the float, which tells you almost nobody was left betting against this stock. Gross margin guidance came in flat at 56%, and cash conversion was weaker than expected, and neither number is alarming alone, but together they shifted the conversation from how fast AMD is growing to how much of that growth turns into cash.

That's a far less flattering question for a stock trading near 44 times forward earnings. Not a broken thesis. A trade that ran out of new buyers before the story ran out of good news.

AMD spent the days before earnings doing exactly what a crowded trade does, climbing even without new information, gaining 7% the day of the print alone and touching $518.58 just hours before the numbers came out. Then the results landed, and instead of confirming the run, they became the excuse to take profits.

The stock dropped 8.8% in after-hours trading to $472.94, kept bleeding into the next session, and settled around $483 by the end of the week. That puts it roughly 17% below its 52-week high of $584.73, even after a year that's still seen the stock roughly triple.

What matters now is whether $470 to $475 holds as a floor. That level acted as the first real resistance during the after-hours plunge, and a clean bounce from there would suggest this was profit-taking rather than a genuine repricing, while a break below it opens the door toward the low $400s, a zone the stock hasn't traded at since before this year's rally really got going.

Flat 56% margin guidance was the actual trigger for this selloff, more than the revenue number itself. Watch whether margins expand even modestly next quarter as Helios ramps and mix shifts further toward data center, since that's the single number most likely to reopen the bull case.

Management called this quarter the start of the Helios ramp, not the middle of it. The next two quarters of disclosed deployment volume will tell you whether the Anthropic and OpenAI commitments are converting into revenue on schedule or slipping into next year.

A put to call ratio this skewed doesn't reset in a single week. Watch open interest over the next few weeks for signs the crowded call positioning has actually cleared out, since that unwind is often what creates the next genuine buying opportunity rather than round two of profit-taking.

I've been tracking this rivalry since AMD hit its all-time high back in July, and the SpaceX-Nvidia exclusivity comment this week is a reminder the AI infrastructure race isn't a two-horse coronation for AMD yet. Because that race is still contested, competitive commentary from Nvidia's own next report matters more than usual right now, and any sign of slowing data center demand there would hit sentiment on AMD too, fair or not.

Back in May I flagged that AMD was starting to look overpriced, and the stock proceeded to run another huge leg higher before finally cracking this week. Watch whether this pullback is the valuation reset that call was actually about, or just a pause before the next leg of a trade that's stayed crowded far longer than the multiple alone would suggest it should.

I flagged AMD as overpriced back in May, and the stock ignored me for three more months and another huge leg higher before this week finally happened.

That's not a failure of the call. It's a reminder that crowded trades don't correct on a valuation model's schedule, they correct when the last incremental buyer runs out, and this week that's exactly what happened.

The business is not the problem here. Data center revenue more than doubling, Helios ramping, and three of the biggest names in AI locking in multi-gigawatt commitments is about as strong a fundamental story as this sector produces.

I'm not selling AMD, but I'm also not pretending this drop was irrational. A stock priced for perfection needs perfection, and a great quarter that was merely great was always going to be the thing that finally punctured it.

Tax season doesn't have to mean wondering if you have the right forms, second-guessing your deductions, or scrambling to pull everything together before the deadline.

With BELAY’s Tax Prep Checklist, you can start preparing for tax season with confidence.

Download the Free Checklist

The most dangerous place to own a great business is right after everyone else agrees it's great.

AMD's fundamentals didn't change this week. The price everyone was willing to pay for those fundamentals did.

That distinction matters more than almost anything else in investing, and it's the one most easily forgotten during a run this strong.

A crowded trade can be right about the business and still be wrong about the entry point.

Both of those things were true for AMD this week, and only one of them is actually AMD's fault.