AI Could Crash Most Businesses, But Real Assets Are Immune

The article argues that AI could pressure a much broader set of businesses than investors currently expect, with SaaS potentially being only the first sector affected. It suggests real assets may offer relative immunity to AI disruption, implying a defensive rotation rather than a broad risk-on trade. The piece is largely thematic commentary rather than event-driven news, so near-term market impact is likely limited.

Analysis

The market is still pricing AI as a software profit-shift story, but the bigger second-order effect is cost compression across white-collar workflows, which tends to hurt businesses with low switching costs, high labor content, and weak data moats long before it helps the model providers. That means the vulnerable cohort is broader than SaaS: outsourced services, agencies, recruiting, basic legal/admin software, and information intermediaries can see margin pressure within 2-4 quarters as customers demand lower prices or unbundle seats. The winners are not just "AI-native" companies, but any business that owns scarce physical bottlenecks rather than human throughput.

Real assets have a different defense: they are scarce, inflation-linked, and harder to replicate with code, but the immunity trade is not uniform. Asset-heavy names with contractual escalators and high occupancy or capacity utilization can benefit if AI accelerates productivity and corporate consolidation, yet real assets tied to secular demand decay or financing dependence remain exposed if AI-induced disintermediation slows tenant growth or weakens credit quality. The key second-order risk is that lower labor costs can actually reduce replacement demand for some logistics, office, and light industrial footprints even as it supports margins elsewhere.

From a positioning standpoint, this is more of a multi-quarter relative-value trade than an immediate catalyst event. Consensus is likely underestimating the speed of repricing in labor-arbitrage-sensitive industries and overestimating how much of the value accrues to incumbents versus customers. The market usually waits for one or two earnings seasons of downward guidance before fully marking the losers, so the best risk/reward is to fade expensive workflows and own hard-asset cash flows with visible contractual protection.

The contrarian angle: "AI immunity" may become crowded just as investors realize some real assets are not immune at all — they are simply later-cycle casualties or beneficiaries depending on lease duration, leverage, and tenant mix. In other words, the right trade is not real assets broadly, but select real assets with pricing power, low refinancing risk, and limited substitute exposure. That nuance matters because the initial move could overshoot into a blanket rotation, creating opportunities to buy quality hard-asset cash flows on weakness and short structurally exposed service names on strength.

Key Decisions for Investors

• Short a basket of labor-arbitrage-sensitive software/services names over 6-12 months; target 20-30% downside as AI-driven pricing pressure shows up in renewal rates and headcount growth decelerates.
• Own high-quality real asset cash-flow compounders with contractual escalators for 12+ months; look for 8-12% upside from multiple support plus mid-single-digit NOI growth, but avoid names with heavy refinancing needs.
• Pair trade: long select real estate/real-asset beneficiaries vs short enterprise workflow software; aim for relative outperformance if the market starts treating AI as a margin-disruption theme rather than a pure productivity story.
• Prefer balance-sheet-strong operators with low leverage and long-dated liabilities; in a 6-18 month window, they have the best chance of surviving pricing compression while weaker peers face dilution or covenant stress.
• Do not chase a broad "real assets are immune" basket; use a 5-10% drawdown in crowded names to initiate, since the trade can get crowded before fundamentals fully reset.