One of the market's hottest stock themes is buying everything AI can't replace

Roundhill launched the HALO ETF (LOHA) on Thursday, targeting U.S. companies with heavy physical assets and low AI obsolescence risk. The theme is being positioned as a beneficiary of AI disruption, with cited holdings such as FedEx and ExxonMobil up about 30% year to date and Coca-Cola up about 17%. The fund follows a rapidly gaining AI-resistant trade, but the article is primarily thematic commentary rather than a direct corporate catalyst.

Analysis

The HALO bid is less about “AI winners” than a capital rotation into companies with pricing power, asset intensity, and lower earnings-disruption convexity. That favors cash-flow compounding businesses with embedded replacement value: transportation, auto aftermarkets, industrial engines, rail, and HVAC tend to absorb AI as an efficiency layer rather than a demand destroyer. The second-order effect is that passive and thematic flows may keep compressing the quality discount on these names even if near-term fundamentals only improve modestly.

The biggest crowded-risk here is not the concept itself, but the ETF wrapper becoming the marginal buyer. Once thematic flows become explicit, factor investors often overcrowd the same “durable” basket, which can push valuation to levels where the downside is no longer fundamental disruption but multiple mean reversion. That creates a subtle vulnerability in names with strong business quality but cyclical earnings elasticity: if macro growth softens, the market may discover that “AI-resistant” does not mean “recession-proof.”

The losers are the software franchises already under scrutiny, but the more interesting dynamic is competitive: if enterprise buyers reallocate spend toward physical-world efficiency and away from broad software seat growth, then legacy application vendors can see budget pressure while industrial distributors and logistics firms gain share in customer workflows. Meanwhile, AI can actually widen the moat for asset-heavy operators by improving routing, maintenance, and yield management, increasing returns on existing infrastructure without changing the underlying need for assets. The trade is therefore a relative-duration trade, not a pure growth-vs-value call.

Contrarian view: the move may be underappreciating how much of HALO is already embedded in quality-factor exposure. The more investors label a basket “AI resistant,” the more the market may re-rate it like a defensive growth trade, not a cheap cyclical one. That means the best opportunities are likely in the names that benefit operationally from AI but are still priced like old economy businesses, rather than the obvious HALO leaders already up significantly this year.

Key Decisions for Investors

• Go long FDX / short ADBE in a 6-9 month pair trade: thesis is that AI improves logistics optimization while software multiples remain vulnerable to budget scrutiny; target 15-20% relative return with stop if software sentiment stabilizes broadly.
• Build a basket long CMI, CSX, and TFII on pullbacks over the next 2-4 weeks: these names have the cleanest mix of physical asset leverage and AI-enabled productivity uplift, with 10-15% upside if HALO flow persists.
• Sell out-of-the-money call spreads on KO and AZO for 1-3 months: both benefit from defensive rotation, but the thematic crowding can cap upside; attractive if implied vol stays elevated and the market continues chasing the same basket.
• Pair long XLI/industrial-logistics exposure against short NOW or CRM for 3-6 months: monetizes the divergence between asset-backed earnings and software valuation compression; watch for any AI monetization headlines as the main risk.