The block is multifamily's clearing price now

San Antonio's Embrey Partners is pricing a $5 billion apartment portfolio block by block, and that sentence should sit on a generalist apartment LP's desk this morning because it describes a change in the unit of underwriting rather than a change in the market. Where an apartment acquisition once cleared against a metro forecast—job growth, in-migration, a rent-growth assumption spread across a 40,000-unit submarket—the discipline now runs corner by corner, and each corner is scored on whether new supply can reach it.

Trepp's latest work supplies the reasoning behind that shift: the high-inflow metros that spent the post-2021 years leading the country in rent growth gave their revenue advantage back to the construction that growth attracted. Migration still tells you where households are going, but it has stopped telling you what an apartment there will earn, because the same force that pulls people into a metro also pulls cranes into the submarkets that received them. PWD's tracking has carried that finding for weeks; what is newer is the number of buyers now pricing as though it were true.

The arithmetic underneath the shift is the 1.2 million lease-up units still working through the pipeline, a figure that is the ceiling on apartment rents as the delivery wave crests and each unit competes for the same renter in a specific place—a specific corner of a specific block. That is why the aggregate rent-growth story has stopped clearing investment committees. A metro average is a smoothing device, and the market it smooths no longer exists.

Nor is the pipeline about to empty: Fengate and Mavrek's Chicago groundbreaking will not deliver until early 2028, which means the ceiling it represents stays in place through the next two leasing seasons and into the year that building starts its own lease-up. An underwrite that assumes the supply pressure lifts before then is assuming away a tower that has not broken ground yet.

Supply discipline is the phrase doing the work here, and it is worth being precise about what it means: the buyer's screen now runs on the parcels adjacent to the asset rather than on the metro's aggregate pipeline, because the two now produce different answers, and the crane count in the metro is no longer the variable that decides the screen. A metro pipeline can be healthy while a specific block is overbuilt, and the renter choosing between two buildings a quarter mile apart does not care about the metro pipeline at all.

The lag between the research and the repricing is not mysterious: apartment underwriting has been organized around the metro—the metro supply number, the metro rent index, the metro population forecast—because those were the numbers that could be assembled at portfolio scale, and because a decade of migration-driven rent growth rewarded the approximation. What the 1.2 million lease-ups changed is that the approximation now carries a specific error term, and the error term is a building under construction two blocks over that the metro number washes out.

What replaces the metro is the block, and the block asks a different question: not whether the metro will grow, but how many units can lease up within a quarter mile of this address and how quickly the next parcel gets entitled. The first is a macro question an LP can delegate to a research shop; the second is a local question that requires a development team's ear at the ground level, which is why the buyers moving fastest now tend to be the ones who build.

This is not a counsel of paralysis: Embrey's book is five billion dollars of conviction, and the West Loop groundbreaking is a real check written against a real corner. What the block-level underwrite requires is a different kind of diligence—walk the radius, count the cranes, price the entitlement queue, and accept that the answer for this parcel may differ from the parcel across the street—because it is slower and more expensive work, and it is the only way to underwrite a market where migration no longer predicts the rent.

LCOR's $73 million purchase of Taconic's lab site is the same logic expressed as a land trade. The site was entitled for life-sciences construction and never produced a building, and it sold $6 million below the construction loan raised against it. Paying apartment-feedstock pricing for the site is underwriting the blocks around it rather than the lab story the entitlement told; the entitlement reflected a metro-level bet, while the price reflected a block-level one.

The Canadian bid into U.S. real estate is following the same channel: the $9 billion total still trails the five-year average, but the share shift underneath it points at apartments. Institutions that spent a decade buying gateway office and industrial are sizing multifamily now, and they are arriving in a market where the metro rent forecast no longer clears an investment committee. That gap—Canadian capital arriving with a metro thesis into a block-level market—is likely to be a more reliable source of mispriced deals than any other single factor over the next eighteen months.

Fengate and Mavrek's 380-unit West Loop groundbreaking reads the same way from the construction side. There is no rent roll; the building opens into early 2028, so the wager is on the corner holding rather than on operations. A location trade priced before a single unit leases is what a block-level underwrite looks like when there is no history to discount, and Chicago's West Loop is a market where the next parcel is already in someone's pro forma.

The Rochester trade is the cleanest read on what a single employer is worth at that level. All-cash bidders priced the building off Mayo's payroll at $7.27 million, small enough that most institutions would skip it, which is exactly why the bid count is the purest available measure of what employer adjacency does to a block's valuation. A Mayo-anchored block does not need a metro forecast; the payroll is the forecast.

Grand Peaks and PCCP bought Beaverton's West End District on agency debt with no price published, and two things matter there: the agency financing is a bet that the block holds a stabilized coupon, and the retail rent roll is where the upside lives. A 2021-22 vintage under Freddie Mac paper is not a rent-growth story so much as a spread story. The $83 million Oregon trade—two complexes at a blended $189,500 a unit, with no published cap rate for either—follows the same pattern of quiet clearance, as public prices thin out when the block-level underwrite stops producing a metro-comparable number an appraiser can bracket.

For three years, the reliable way to lose money in apartments was to underwrite a metro; for the next two, the reliable way will be to underwrite a metro and call it a block. Buyers running a supply-discipline screen at the parcel level will clear their basis, while buyers applying a metro rent-growth number across a portfolio will discover that the 1.2 million lease-up units resolve at the address rather than the MSA, because the concession is negotiated on a specific corner and so is the renewal.

There is a mechanical reason the repricing will arrive unevenly: a block-level underwrite takes longer per asset than dropping in a metro number, so portfolios will trade slowly while single assets trade fast—the reverse of the last decade's pattern, where one MSA story could move a 3,000-unit package through a single committee meeting. The single-asset trades this month with no published price are the front edge of that inversion, and the reason they clear without a syndicated number is that there is no metro comp to syndicate.

Embrey's $5 billion is the frame generalist LPs will spend the next two years catching up to, corner by corner. The first sign will be a generalist fund buying a Sun Belt apartment portfolio and reporting what it paid per block rather than per unit. The second will be a block-level trade that clears with no metro comp at all—no published price, no cap rate, agency paper or none, the Rochester template repeated at scale. When that trade prints, it will settle whether the metro is a retired unit of underwriting or a sleeping one, and it will settle it before the last of the 1.2 million lease-ups has found its first renter.