A stock falls and suddenly investors start questioning the moat they were praising at higher prices. Management gets scrutinized. Risks that were always there start feeling much bigger. Then the stock rallies. Now those same risks become easier to ignore. Management looks brilliant, future growth seems more certain, and paying a higher multiple somehow feels safer. You can see this tension in the AI trade right now. Growth can be real, but that doesn’t mean the capital being spent will generate enough future cash flow per share to justify the valuation. And sometimes the market has a point. Cash flows might arrive later than expected, or the underlying economics might actually be deteriorating. An intact business can still be overpriced. That’s why I try to separate what’s happening to the stock from what’s actually happening to the business. I want to know what changed, what expectations are already priced in, and what return the new price offers for the risk. A falling stock doesn’t automatically mean a broken business. And a rising one doesn’t automatically mean a better one.

The chip can leave the fab long before an AI system is ready to run. $NVDA designs the accelerator. $TSM manufactures the silicon. But someone still has to design the boards, networking, power and cooling systems. Components need to be sourced, racks assembled and everything tested before the infrastructure can be deployed. One failed system test can delay an entire deployment, even when the GPUs themselves work. That is where the businesses behind the chip come in. $TTMI supplies circuit boards. Foxconn and $JBL L manufacture at scale. $CLS works across engineering and manufacturing. $HPE sells complete systems, while $PENG helps get AI clusters running. Different jobs. Different ways to get paid. AMD’s ZT Systems transaction shows why. $AMD D kept the rack-scale design teams while $SANM acquired the U.S. manufacturing business. $TSSI shows another distinction. Most of the components it integrates belong to customers. It earns revenue for putting them together without reporting their full hardware value as sales. That is why revenue growth alone cannot tell us which business captures the most value. I want to know who has the expertise customers depend on, who carries the capital burden and who can keep earning attractive returns. The chip gets the headlines. But making it work is a business of its own. Business Behind the Business | Part 5

Man, @Tape_Vector has been putting out some seriously good research lately. This $KKR breakdown is another fantastic read. I've spent a lot of time digging into $KKR , Brookfield and Apollo myself, so this one was right up my alley. What fascinates me about KKR is how its different capital engines work together. Long-duration capital, management fees, Global Atlantic and investment realizations all feeding into the wider business. But growing the platform and creating value per share aren't necessarily the same thing. I especially appreciate how he gets into the details that are easy to overlook. The $72B not yet paying fees, the K-Series reporting changes and why the USI realization shouldn't be treated as recurring profit. His recent $ROP article was another great read. I've owned Roper before and spent plenty of time revisiting the thesis, so I loved seeing him dig into the Indicor gain, revenue quality and debt-funded buybacks. And these are just two of his articles. He's been covering everything from $KEY to $COCO, while still putting out his market reads, sector rotation research and leadership radars. I've actually been using those radars as another input for finding interesting businesses and sectors myself. I especially appreciate how he separates emerging leadership from confirmed strength. The range of his work is impressive, but what I appreciate even more is the effort he puts into understanding each business beyond the headline numbers. Seriously, my brother has been on a tear lately 😂 If you enjoy fundamental investing, finding new businesses or understanding what's happening across the market, this is an account you really shouldn't miss. Go give him a follow. His research is worth your time.