The Observation
Small-bay light industrial, multi-tenant facilities in the 150 to 500 m² (1,600 to 5,380 sq ft) bay range serving small and medium-sized businesses, is one of the most consistently performing asset classes in Canada and the United States. Owner-occupier demand is structurally stable. The tenant base is domestically driven. The product is undersupplied in most markets where it exists.
In Mexico, this asset class has not received the same level of attention. Institutional capital followed the nearshoring narrative into large-format logistics and border manufacturing parks. The sub-500 m² (5,380 sq ft) bay category has not been systematically built or tracked. A hybrid small-bay model, in which some bays are sold as condominiums to owner-occupiers and the rest are held for lease, had been identified as a niche with demand but limited supply in several Mexican markets. That combination looked like a thesis worth evaluating seriously.
What the Evaluation Found
Four structural problems emerged in the markets evaluated.
In the markets evaluated, market pricing for small-bay condominium units did not cover the current cost of land and construction. The gap was not marginal: the numbers were significantly underwater. The spread between what an SME owner-occupier will pay and what it costs to deliver the product (land, entitlements, construction, sales) did not produce a viable development margin at any configuration evaluated. Holding more of the bays for lease does not fix this, and it adds problems of its own.
The rental income in a small-bay Mexican asset is denominated in pesos. The return expectations of a North American investor are denominated in US dollars. That currency mismatch does not disappear with structure: it has to be priced, and pricing it further compresses an already thin return.
Exit liquidity is thin, and leasing the bays does not change who can buy them. Institutional buyers such as Mexico's FIBRAs need tenant quality and scale that small-bay product does not offer, so whether the bays are sold or held for lease, the buyers are still limited to owner-occupiers and domestic operators. With business borrowing in Mexico typically costing 10 to 15% a year, two to three times what the same borrower pays in Canada, fewer of them can actually close.
Related to this, the interest rate environment in Mexico fundamentally changes the risk calculus in ways that are not obvious to a North American investor arriving with Canadian assumptions. Mexican government treasury rates (CETES) peaked at approximately 11% in 2023 and 2024, more than double the equivalent Canadian rate at the time, and they remain in the 6–7% range as of 2026. That is the risk-free benchmark against which local capital prices every deal. A development return that looks attractive by Canadian standards may be thin or inadequate against the Mexican risk-free rate. Any return modelling for a Mexican real estate project that does not account for this comparison is incomplete.
What Was Observed in the Market
Small-bay and micro-industrial lots are being actively marketed and sold to retail investors in certain Mexican markets, including in at least one major urban centre observed. Units are being marketed on the basis of projected rental returns. Whether those returns are achievable at current market rents is, in the view of this evaluation, an open question. In some cases the land was acquired years ago at prices that no longer reflect current replacement cost: the development margin existed at entry, not at today's cost basis. Neither observation changes the conclusion. Demand exists, though in the market observed it is now softening, with supply saturating and prices coming down. The question is whether the numbers work for a developer entering the market today at current land and construction costs.
The Decision
The thesis as structured did not clear the standard in the markets evaluated. The gap between achievable sale prices and the cost of land and construction is the primary problem. Holding for lease does not solve it: the currency mismatch, thin exit liquidity, and the interest rate environment compound it.
What Would Need to Change
Land at a price point that lets the cost of land and construction be recovered at achievable sale prices. A peso-denominated return structure targeting domestic Mexican capital would address the currency problem. Both require a different entry point and a different capital structure than currently exists.
Current Status
Monitoring. The white space in this asset class is real, and the macro demand case holds, though it varies by submarket: the softening observed was in one of the few markets where this product already exists. The economics as currently structured do not work for a North American capital base in the markets evaluated. That may change as land pricing, construction costs, and the interest rate environment continue to shift.
I tested these findings with people on the ground before publishing. They reached the same conclusion: at today's numbers, the product is underwater.
This case study is an analysis of a market thesis. It is not an investment offering or a solicitation of any kind.