For the first time since 2022, more apartments were absorbed than built.

Every multifamily pro forma written between 2023 and 2025 had the same weak point: a lease-up assumption made in a market being flooded with new units. Concessions ate the first year. Rent growth arrived late or not at all.

That pressure is lifting, and the crossover is measurable.

Net absorption was 167,500 units in Q2 2026 alone, nearly double the 84,300 of Q1. Deliveries fell to 77,700, down 14% year over year, and 2026 completions are projected to fall 28% to about 382,000 units. Vacancy came down to 4.3%, below the long-run average of roughly 5%.

Why the rent number looks disappointing anyway

Average asking rent reached $2,257 in Q2 2026 — up only 0.5% year over year, but 1.5% on the quarter. That gap is the whole point. The annual figure is still carrying the damage of the supply wave; the quarterly figure is what the market is doing now.

Year-over-year rent tells you where the market has been. Quarter-over-quarter tells you where it is.

Regional leaders are not where the last cycle's money went. The Midwest led at about 2% annual rent growth, then the Northeast at 1.7% and the Pacific at 1.4% — the markets that never got a supply wave in the first place.

What we would actually check

• The submarket delivery pipeline for the next eight quarters, not the metro one. Metro-level easing means nothing if four hundred units are opening a mile away.
• Concessions still embedded in the comps. A quoted rent with two months free is not that rent, and it will not renew at that rent.
• Whether your lease-up assumption was written during the flood and never revisited.
• Operating expenses, especially insurance and taxes, which did not ease at all while rents were flat.

A market turning is not the same as a market that has turned. Absorption beat deliveries by about four thousand units out of 360,000 — a rounding error that happens to point the right way.