Industrial's Long Boom Is Really Two Booms: Cold Storage and Last-Mile

Industrial real estate stopped being one story the moment the pandemic surge cooled. The bulk warehouse that anyone could build and anyone could lease has moderated, with national vacancy holding in the mid-6% to 7% range and asking rents growing under 1% year over year, per CBRE. Underneath that flat headline, two segments keep compounding: temperature-controlled cold storage and last-mile logistics. They are the same NAICS code and a different business. Treating industrial as a single asset class is now the fastest way to misprice it, because the average hides a segment that is soft and a segment that is structurally short.

Key Takeaways

U.S. industrial vacancy held around 6.7% to 7.5% in early 2026 with asking rents up only 0.8% year over year, per CBRE, but the average masks a two-speed market.
Cold storage runs materially tighter than dry warehouse, with national vacancy near 3.4% against roughly 5% to 7% for conventional space, per market commentary.
Cold storage costs two to three times dry warehouse to build, so the barrier to entry that softened bulk warehouse is exactly what protects cold storage rents.
Last-mile facilities compete on drive time to population, not on cheap land, which makes their pricing power a location monopoly rather than a construction race.
The right unit of analysis is the segment, not the sector: bulk warehouse, cold storage, and last-mile underwrite to different demand drivers and different downside.

Why Is Industrial Real Estate No Longer One Asset Class?

Industrial is no longer one asset class because its segments now diverge on demand, supply barriers, and pricing power. Bulk warehouse demand has moderated toward its long-run trend while cold storage and last-mile continue to see structural demand growth, per CBRE. When the sub-segments move in different directions, the sector average describes none of them.

The pandemic masked this. For three years, every kind of industrial leased quickly because e-commerce pulled demand across the whole spectrum at once. That tide has gone out. CBRE reports national vacancy stabilizing in the mid-6% range and asking rents that grew 0.8% year over year to $10.34 per square foot in early 2026, the first rent growth since 2024 but a fraction of the double-digit surges of the boom. A single "industrial is soft" or "industrial is hot" verdict is now wrong in both directions, because the answer depends entirely on which industrial you mean.

What Do Cold Storage and Last-Mile Actually Need That Bulk Warehouse Doesn't?

Cold storage needs refrigeration infrastructure that costs two to three times dry warehouse to build, and last-mile needs proximity to dense population that no amount of construction can manufacture. Both requirements are supply constraints, and supply constraints are what let rents hold when demand normalizes. Bulk warehouse has neither constraint, which is why its rents moved first when the surge faded.

Cold storage is a capital and engineering problem. Terrapin Construction Group's 2026 cost guidance puts refrigerated space at roughly $130 to $210 per square foot, frozen at $200 to $285, and blast-freeze at $285 to $350 and up, against a fraction of that for dry warehouse. Racking, insulation, backup power, and refrigeration are not features a landlord adds later. They define the box. That cost is a moat: it deters the speculative overbuilding that softened bulk warehouse, and it explains why cold storage vacancy sits near 3.4% while conventional warehouse runs 5% to 7%, per market commentary.

Last-mile is a geography problem. The value of a last-mile node is its drive time to the customers it serves, so its pricing power comes from a location that cannot be replicated at the edge of the metro where land is cheap. Build-cost guidance for last-mile with docks, automation, and office runs $150 to $300 and up per square foot, but the real barrier is that the infill parcels close to population are largely spoken for. A tenant pays a premium for the ten-minute delivery radius, not the building.

How Should You Underwrite Cold Storage and Last-Mile Differently?

You underwrite each segment to its own demand driver and its own downside. Cold storage underwrites to grocery and pharmaceutical throughput and to the specialized-improvement risk of a box that only one class of tenant can use. Last-mile underwrites to population density and delivery economics, where the location is the asset and the structure is almost incidental.

Segment	Primary demand driver	Supply barrier	Underwriting watch-item
Bulk warehouse	Regional distribution volume	Low; land plus a shell	Oversupply and rent softening
Cold storage	Grocery, e-grocery, pharma throughput	High; 2 to 3x dry build cost	Tenant concentration and re-use risk
Last-mile	Population density and delivery speed	High; infill land scarcity	Location durability, not construction

The downside cases differ as sharply as the upside. A cold storage box is a specialized improvement: if the operator fails, the pool of replacement tenants who need that exact refrigeration is small, so tenant credit and lease structure matter more than in a fungible dry warehouse. A last-mile facility carries the opposite risk profile: the building is generic, the location is scarce, and the underwriting question is whether the delivery economics that justify the rent survive the next routing and automation cycle. As one industrial capital-markets lead put it: "In cold storage you underwrite the tenant, and in last-mile you underwrite the map." A pro forma that applies bulk-warehouse assumptions to either one is underwriting the wrong risk.

What Does the Two-Speed Market Mean for the Cycle?

The two-speed market means the industrial cycle is no longer a single wave that lifts or drops all boxes together. Bulk warehouse now behaves like a commodity that tracks supply, while cold storage and last-mile behave like constrained assets that hold rent through softness because their supply cannot flex up quickly. The sector will keep printing a flat average while its halves diverge.

CBRE's 2026 outlook supports the split from the supply side: new construction starts remain relatively flat and existing supply is being absorbed, which stabilizes bulk vacancy without reigniting the rent growth of the boom. The constrained segments do not depend on that absorption story. Their scarcity is structural, built into refrigeration cost and infill land, so their rents are far less exposed to the next wave of speculative dry-warehouse deliveries. The operator who reads the sector as one number will keep being surprised. The one who prices the segments separately will not. This is the same lesson our submarket-versus-metro analysis applies to geography, one level up: the average is the wrong unit whenever the parts diverge.

Frequently Asked Questions

Is industrial real estate still a good investment in 2026? Industrial is a two-speed market in 2026. Bulk warehouse has moderated, with national vacancy in the mid-6% to 7% range and sub-1% rent growth per CBRE, while cold storage and last-mile logistics continue to see structural demand and tighter supply. The answer depends on the segment, not the sector.

Why is cold storage tighter than regular warehouse? Cold storage is tighter because it costs two to three times as much to build as dry warehouse, which deters speculative construction and keeps supply constrained. National cold storage vacancy sits near 3.4% against roughly 5% to 7% for conventional warehouse, per market commentary, and that build-cost barrier is the reason.

What makes last-mile logistics valuable? Last-mile logistics facilities are valuable for their proximity to dense population, which shortens delivery times and cannot be replicated where land is cheap. The pricing power comes from the location's drive-time radius, not the building, so infill land scarcity, rather than construction cost, is the binding constraint.

Conclusion

Industrial's long boom did not end. It split. The commodity end, the bulk warehouse anyone can build, has reverted toward a supply-driven equilibrium with thin rent growth and mid-6% vacancy. The constrained ends, cold storage and last-mile, keep compounding because their supply is capped by build cost and by geography, and neither responds quickly to demand.

The operator who underwrites industrial as one asset class is pricing an average that no building experiences. The segment is the unit that matters: cold storage is a tenant-and-improvement bet, last-mile is a location bet, and bulk warehouse is a supply bet. Getting the segment right is the difference between owning the part of industrial that is structurally short and owning the part that is merely stable.