29 Forbes articles in September. Nike. GameStop. Micron. McDonald’s. Meta. Wendy’s. Anthropic. Different companies. Same question: What changed beneath the headline—and what has the market still failed to price? My September Insights.

📣 BIG ANNOUNCEMENT I’ve been selected by Forbes to write its flagship analysis of the upcoming Anthropic IPO. A privilege, and one I’m genuinely excited about. But this won’t be another article about AI hype. I’ll be digging into the valuation, the risks, the ownership structure, and what investors need to understand before putting a dollar to work. The real question isn’t how big Anthropic could become. It’s what investors should be willing to pay for it. My analysis will be free to read. Subscribe below so you don’t miss it. 👇 forbes.com/sites/jimosman/

At the dentist. Reception: “We recommend cutting down on sugar. It’s terrible for your teeth.” Also reception: A Halloween bucket containing enough sugar to bankrupt a small nation. I’m beginning to understand their business model. Create the problem. Sell the solution. Genius. 😂🦷

The uncomfortable truth in semis: the best results often finance the next problem. TSMC plans $60–64B of capex this year. ASML’s customers are already talking about 2029–30 orders. Rivals are spending to keep up. Each move is rational. Together, they create more capacity.

Waste-collection trucks and landfills don't sound like an AI story. That may be exactly the point. GFL Environmental owns trucks, routes, and landfills. Not glamorous. Not Silicon Valley. But AI routing software can keep trucks out of traffic. Computer vision can separate recyclables more cleanly. Small improvements repeated across thousands of routes add up. Waste shares have also recently moved differently from headline AI names. #Investing #AI #GFLEnvironmental

I stopped expecting people to show up for me years ago. Not out of bitterness. Out of experience. The hardest lesson in life is realizing that some people you’d move mountains for wouldn’t cross the street for you. So I learned to deal with my problems quickly, quietly, and on my own. Never put your peace in someone else’s hands.

SpaceX spent $48 BILLION in 9 months. Operating cash flow? $13 billion. A $35 billion gap. Meanwhile, 84% of capex is going into AI, not Starlink. I remember the telecom boom. The technology changed the world. Investors still lost fortunes. Growth isn't a return on capital. $SPACX forbes.com/sites/jimosman/20…

SpaceX just crushed earnings. $14.6B revenue. Up 161%. $4.4B net profit. $8.8B adjusted EBITDA. Wall Street will celebrate. But here's the number I'd watch: **$19.6 BILLION in quarterly capital spending.** Extraordinary growth. Even more extraordinary spending. The question isn't whether SpaceX can grow. It's whether shareholders will earn enough on the capital required to get there. $SPCX

FYI prweb.com/releases/the-edge-…

Stock pickers think the AI selloff is finally their moment. Maybe. But 70% of active managers are still trailing. Avoiding an expensive stock is not an investment thesis for whatever you buy instead. Cheap without a catalyst can stay cheap for years. My latest: forbes.com/sites/jimosman/20…

THE EVERYTHING ELSE RALLY

Glad Macron is taking control now …

I do my research. I don’t need machines to help me have a view.

The AI selloff is getting more dangerous. First, investors sold the companies drowning in debt. Now they’re selling the profitable ones with strong balance sheets. That’s when you should start worrying. Because when the market stops distinguishing between the weak and the strong, the problem is no longer just the companies. It’s what investors are suddenly willing to pay for them. My latest Forbes Editors’ Pick. forbes.com/sites/jimosman/20…

Micron doesn’t need a 2023-style bust to disappoint. At projected revenue, every margin point is worth $600M+ of quarterly gross profit. A drop to 75% keeps it hugely profitable but could still erase billions in annual earnings. Shareholders are betting supply stays tight.

Nvidia's order book tells you somebody is spending money. It doesn't tell you who makes money from using what Nvidia sells. Mid-caps are trading at significant discounts to large caps, while earnings quietly accelerate from 10% to 13-15% growth. Multiples fell from 20x to 17x. The market is paying less for a business that may be getting better. That gap is where the next AI winners might actually live. #Investing #AI #MidCapStocks

One reason I left Britain. Tell someone in the UK you’ve got a big idea, and they’ll give you ten reasons it won’t work. Tell someone in America, and they’ll ask how big it could get. I love Britain. Always will. But you can’t build a future by constantly explaining why you can’t.

The Is AI demand real? debate is over. Customers are fighting for chips, memory, and the machines that make them. The new question: how much is priced in, and what will these profits push the industry to build next? Micron’s 87% gross margin is fantastic. It’s also an invitation.

Oracle trades cheaper than Microsoft and Amazon. That comparison gets less reassuring once you check how each funds its AI buildout. Microsoft and Amazon finance AI investment with cash generated elsewhere. Oracle is selling shares, adding debt, and asking customers to fund the hardware. A lower multiple isn't cheap when the business needs far more capital to produce it. #Investing #Oracle #Valuation

I spent hours studying balance sheets, credit spreads and AI valuations. Then this gentleman arrived with a compelling counterargument. I’ve cancelled my research subscriptions, fired the analysts and put him in charge of risk management. We’re launching a hedge fund on Monday.