I sold DaVita in February because dividends felt more real than buybacks. It returned 70% while I watched. Here's what I learned about corporate cannibals.

A cannibal is a company retiring its own shares aggressively - but not the fake kind where buybacks just offset employee SBC dilution. NET buybacks. The kind where the share count actually shrinks.

The math is interesting: if true FCF is growing while shares shrink, the yield available to you as a shareholder is compounding twice. And here's the psychological twist — if the stock drops during the buyback, that's actually good news. Lower price means more shares retired per dollar, which means higher ownership stake. You almost get to root for the price to drop.

My filters: buyback yield of at least 5% (net share reduction) plus true FCF yield of at least 8% (OCF minus CapEx minus SBC). Here's what came out of my screener.

The standouts: ADBE, CMCSA, DBX, PYPL, DVA, BCO. Profit margin as a moat proxy puts ADBE, CMCSA, FISV and GPN at the top.

The Adobe section of the piece is what I'm most interested in hearing pushback on. The Reddit bear case I link argues freemium is a warning sign. My counter: Adobe is attracting 800 million users and generating creative workflow behavioral data at a scale that Midjourney and DALL-E simply don't have. As a data ontologist by trade, that context corpus looks more like a moat than a threat. Very few people question whether Anthropic's freemium model creates value. Why is Adobe different?

Full piece with jaws of life charts: [https://cavemanscreener.substack.com/p/the-jaws-of-life-finding-stocks-that](https://cavemanscreener.substack.com/p/the-jaws-of-life-finding-stocks-that)