In my last piece, I tried something different — breaking down market narratives version by version. The response was unexpectedly positive.

I found the approach genuinely useful myself. It makes three things visible at once: what the market has already priced in, what might happen next, and which scenarios remain unpriced — and therefore worth investigating.

Without that structure, debating whether "Meta really has network effects" is just noise recycled endlessly.

Back to The Trade Desk. Put simply, this is a stock the market has written off — almost no one believes it can grow again. But that's exactly the question: could its next AI story actually hold up?

That uncertainty is the only reason we're writing this at all.