📦 Amazon made $62.6 billion last quarter. Read the next line.

📊 The Tape: the week in five lines.

📦 Amazon's $62.6 billion: what the headline profit was, and what the business actually made.

🏪 The shop window problem: why the part of a company you can see is rarely the part paying for it.

⏱️ One thing worth doing today: a five-minute check you can run on anything you own.

Friday's close against the Friday before.

Big companies up, small ones down, in a week the 10-year yield hit its highest level since 2007. Borrowing got more expensive for everybody, and the biggest companies shrugged it off.

Amazon reported its second-quarter results on 30 July. Net income was $62.6 billion, up from $18.2 billion a year earlier. Most of the headlines stopped there, and it looked like profits had more than tripled.

They hadn't. The very next line in Amazon's own release says that figure includes $53.4 billion of non-operating, pre-tax income, "primarily from our investments in Anthropic". That's the value of a stake in another company going up. It isn't anything Amazon sold.

The business itself made $27.5 billion of operating income. That's up 43% on a year ago, which is excellent, and a long way from triple.

🔍 Net income: everything that's left at the end, including one-off gains that have nothing to do with the day job.

🏭 Operating income: what the company earned from actually doing its business.

🚩 The gap between them: if net income is far bigger, the explanation is always in the release, usually within a few lines.

Ask most people what Amazon does and they'll tell you it's a shop. The numbers say something else.

Amazon Web Services, the part that rents out computing power to other companies, brought in about 21% of Amazon's sales last quarter and about 60% of its operating profit. The shop in North America turned each dollar of sales into about 8 cents of profit. AWS turned each dollar into about 39 cents.

I call this the shop window problem. You judge a company by the part you can see, and the part you can see is rarely the part paying for it. It matters because the thing you think you own and the thing that actually moves the profit can be two completely different bets.

Pick one company you own. Open its latest results release, which is free on its investor relations page, and answer five questions.

1. Which segment makes the profit? Read the operating income column, not the sales one.

2. Is net income far bigger than operating income? If so, search the release for "other income".

3. Is it spending more than it brings in? Compare cash from operations with capital spending.

4. Is your slice shrinking? Check the diluted share count against a year ago.

5. When does it get tested? Write down the next report date and the range it guided to.

Five minutes. You won't find anything clever. You'll just stop owning things you've never read.

This week's Recap on Substack took Amazon apart properly: the five numbers, what the market's already paying for it, what has to go right and what breaks it, and how it reads to a trader and to a long-term holder. It's free to read.

This free email is roughly weekly and teaches the mechanics. The paid edition of Inside the Trade goes out every Sunday before the open and does something different: the individual names and the levels behind them, written down before the week starts, plus the live board.

Read the Amazon deep dive →

If you want to keep your own numbers straight, the RB Trading Journal works out the R on every trade for you, 20% off with code RBT20.

Everything in one place: start.rbtrading.site

The week that ends with payrolls on Friday 2 October, and which company gets taken apart next. Hit reply and tell me which one you'd pick.

Education, not financial advice.