Correct, and Completely Useless

You need to start with what market concerns right now.

You spend two weeks researching a company. The moat is clear. Revenue is up 30% year over year. Gross margin has been climbing for years. You buy.

Six months later, it’s down 20%.

You reread your notes. Every single line still holds. You can’t figure out what went wrong.

The problem isn’t your analysis. The problem is that the thing you analyzed, the market already knew three years ago — and priced in long before you showed up.

I call this being correct, and completely useless.

You’re Grading a Test You Wrote Yourself

Most people start their research by opening the earnings report and checking boxes. Did revenue beat? Is gross margin up or down? Was guidance raised or cut? You work through every line, and you conclude: not bad.

There’s nothing wrong with the exercise itself. The problem is that you never asked the obvious question: whose question are you actually answering?

You wrote this test yourself. You set the questions, you filled in the answers, you graded the paper. At no point was anyone else in the room.

And the market has never once looked at it.

So the First Step Isn’t Opening the Filing

It’s figuring out what the market is asking right now.

And you won’t find that in the earnings report. You’ll find it in how the stock reacted.

Go back through the last four to six earnings days and write down what the market actually cared about each time. Which quarter did it fall? Which one did it rally? What was the trigger — which sentence, which number?

You’ll notice something fast: the market’s attention has a focal point, and that focal point moves over time.

Find it, and you finally know what to research. Miss it, and everything you do is just grading your own paper.

Netflix: Every Box Checked

On July 16, 2026, Netflix reported Q2 results. EPS of $0.80 beat the $0.79 consensus. Revenue of $12.56 billion grew 13.4% year over year, a hair below the $12.58 billion expected. Total global view hours for the first half of 2026 hit 97 billion, up 2% — faster than the 1.5% growth in the same period of 2025, and achieved while the Winter Olympics and the World Cup were pulling eyeballs away.

An earnings report that beat on EPS and showed accelerating engagement. The stock fell as much as 12% intraday and closed down roughly 8%, an 18-month low.

Someone will point out that Q3 guidance was soft — $12.86 billion in revenue against $13.0 billion expected, EPS guidance of $0.82 against $0.84. That’s true. But a gap that size doesn’t explain an 8% single-day drop.

What the market was worried about had nothing to do with this quarter’s margins. It was one thing: is anyone still watching? And in that same release, Netflix announced it would cut its semiannual viewership report down to once a year starting in 2027, on the reasoning that the focus should return to primary financial metrics like revenue and operating income.

The market asked, “Is the audience still there?” The company answered, “That question, I’ll now address once a year.”

Which is as good as conceding the point.

If your research at the time had been built around margins and revenue growth, you’d have concluded the quarter was fine. Correct, completely useless, and actively dangerous — because it would have given you a reason to average down for entirely the wrong reason.

Meta: Every Box Failed

In 2022, Meta lost 64% of its value. The market had exactly one question for it: how much more are you going to burn on the metaverse?

On February 1, 2023, Meta reported Q4. Check the boxes and not one of them passes.

Revenue of $32.17 billion was down 4% year over year — the first quarterly revenue decline in company history. EPS came in at $1.76, down 52%, against consensus of $2.20. Total costs and expenses of $25.8 billion were still growing 22%. Operating margin compressed from over 30% the prior year to 20%.

Revenue falling, costs rising, earnings cut in half. Anyone grading line by line writes down “fundamentals deteriorating” and moves on.

The stock rose more than 20% the next day.

The reason fits in one sentence. On the call, Zuckerberg declared 2023 the “year of efficiency,” cut full-year expense and capex guidance, and announced an additional $40 billion buyback.

The market asked, “Will you stop burning cash?” The company answered, “Yes, and here are the numbers today.”

That ugly earnings report scored a perfect 100.

Two Answer Sheets

Netflix passed every box, withheld the one piece of evidence the market wanted, and fell 8% in a day.

Meta failed every box, answered head-on the one thing the market was asking, and rose over 20% the next day.

If the quality of an earnings report determined the stock price, one of these two things could not have happened. Both did — three years apart, in different industries, at different scales.

So the quality of the report and the direction of the stock were never the same question. Only one thing decides the second:

Did this report answer what the market is actually worried about right now?

That “not bad” scorecard in your hand measures the first question. Your money is riding on the second.

Three Things You Can Do Tonight

The first step in research isn’t opening the filing. It’s finding out what the market is asking. And the answer lives in the price reaction, not the report.

Pull up any stock you own. No Excel. No paid terminal.

One: review the last four price reactions, not the last four reports. Find the quarter that beat and fell, or missed and rallied. That asymmetry is the question. The line item the market cares about will reveal itself in the reaction.

Two: list what management is dodging. On the call, find the thread analysts keep pulling and management keeps deflecting. The number of times it gets asked is the weight of the question.

Three: look at the premium to peers, then ask what that premium is buying. Peers at 25x, this one at 45x — the market is telling you plainly that it expects this company to do something its peers cannot. The question isn’t “is it expensive.” The question is “what exactly is that something.”

When you’re done, you should be holding a single sentence: what the market fears most about this company is one specific thing, and next quarter’s data can answer it.

That sentence is the question. Research starts there.

Miss the focal point and everything else is just grading your own paper.

Four Questions Before You Size a Position

• What is the market most worried about right now? One sentence. Two lines maximum.

• Do I disagree with the market on that specific point? What’s my evidence? How big is the gap?

• When does the next answer arrive? Write the date.

If you can’t answer, don’t bother with valuation.

You don’t know what the other side is thinking, which means every number you produce is just you playing chess against yourself.

Most of the time you’ll find you don’t actually disagree with the market. That’s normal — and not buying is the correct execution.