Facing a space where valuation runs far ahead of delivery — like embodied AI — most people bet by picking one scenario: either believe humanoids ramp in three years and go all-in on body upside, or call it all a bubble and touch nothing.

Here’s the point: betting on one scenario means that, across bull, base, and bear futures, at least one will break you. There’s a structure that survives all three — the barbell.

Why a barbell

Allocate to the two ends, avoid the middle, like a barbell:

• Left end (certainty): upstream “shovel sellers.” Whichever humanoid company wins, the providers of compute, models, and simulation benefit — certainty beta, no bet on a single winner.
• Right end (moat + upside): downstream real scenes and data flywheels. Names already billing, with switching costs and scene lock-in — moat plus upside.
• Avoid the middle: body assembly is thinnest and most commoditized, squeezed from above and below and cost-crushed by the Chinese supply chain (see “the smile curve”) — visible but untouchable.
• Theme stocks only as small option trades: you can’t buy pure humanoid upside with certainty, so express it in small options with hard caps.
• Always hold cash: the barbell’s third pillar, and the most overlooked.

The key test: it doesn’t break in any of the three scenarios

The barbell’s real edge is that it doesn’t depend on you “guessing the future right”:

• Bull (humanoids accelerate): the right end’s upside and the theme options catch the move; the left-end shovel sellers rise too.
• Base (slow grind): the left end’s certain cash flow and the right end’s moat compound steadily; you don’t panic.
• Bear (L3 stays late, narrative deflates): you avoided the most fragile middle, theme names are only small options, and you hold cash — so when others are forced to sell, you have ammunition to buy discounted quality.

A single-scenario portfolio always has one future that kills it; the barbell lives through all three — and lives best in the worst one. That’s its fundamental edge over “all-in on one belief.”

In a space where even experts can’t call when L3 arrives, the smartest bet isn’t to guess more precisely — it’s to build a structure that survives without needing to guess right. Two ends, avoid the middle, small theme options, always cash. Not sexy — but it keeps you at the table.

Predicting the future is hard; building a portfolio that “doesn’t break under any future” is more doable. Don’t bet the scenario — build the structure.

Does your portfolio look more like “one scenario” or “a barbell”? How much cash are you holding?

(Independent industry and educational research, not investment advice. Companies named are illustrative examples, not recommendations. Data is from public sources and may change. The author and affiliates may hold positions in securities mentioned.)

— Adapted from Embodied Intelligence Investing, Ch. 22 (Barbell Framework) & Deep Dive XI

#AssetAllocation #EmbodiedAI #Investing