Record leasing. Record vacancy. Both are true. Here’s what that actually means.

The headline writers don’t know what to do with this one. Office leasing just posted its best quarter since mid-2018 — 120 million square feet signed in Q1 2026, a 25% jump year-over-year, the first time this decade that quarterly volume has cleared the pre-pandemic average. Meanwhile, vacancy hit 21% — a new all-time record, up 60 basis points in a single year, 400 basis points above where we were when COVID started.

Both numbers are real. Both are being used to argue opposite conclusions. And most of the commentary I’ve read is missing the actual story.

Let me give you the operator’s read.

The vacancy number tells you what’s already broken.

Ten of the last thirteen quarters posted net contraction in occupied office stock. Q1 alone shed 5.2 million square feet nationwide. Oakland-East Bay led the carnage at 944K SF of negative absorption — a market I’d avoid for office exposure for the foreseeable future. Moody’s Analytics expects vacancy to keep climbing as more leases expire through the rest of 2026.

This isn’t a blip. This is a structural reset still in progress.

The leasing number tells you where capital is actually moving.

Financial services firms are carrying Charlotte and New York above their pre-2020 leasing averages. SF is getting a real AI-driven lift — sublet availability is down, and tech tenants are back. These aren’t phantom deals. Tenants are signing.

But here’s the catch: average new lease sizes are still running roughly 15% below pre-pandemic norms — a pattern that has held since early 2023. Tenants are moving fast into spec suites and pre-builts, committing to shorter terms. They’re not planting flags. They’re hedging. The leasing surge is real; the commitment depth behind it is not.

Meanwhile, the distress is hitting numbers that should make every investor pay attention.

A 485,000 SF building in Chicago — sold for $68 million a decade ago — just traded for $4 million. The Denver Energy Center, a two-building complex that went for $176 million in 2013, just cleared a foreclosure at $5.3 million. The GSA sold a 940,000 SF federal building to a residential converter for $24 million. Even Green Street, which tracks the better-quality stuff, has Class A values down roughly 35% from peak.

Lenders and owners held on as long as they could. They waited for the turnaround. Now they’re taking the loss.

The buyers picking these up aren’t idiots — they’re underwriting to a completely different use case. Conversion. Redevelopment. Land value plays. At $4 million for nearly half a million square feet in Chicago, the question isn’t whether office works. It’s whether the structure can be repurposed for something that does.

Here’s the framework I’m using to read this market:

Record leasing and record vacancy aren’t contradictory. They’re a portrait of bifurcation.

The top 10–15% of buildings — well-located, heavily amenitized, in markets with real demand drivers — are getting leased. Rents in those buildings are holding or rising. Buyers of that product can still make money.

Everything else is repricing toward its next-best use, which in many cases isn’t office at all.

The ceiling on the recovery is coming into view. CoStar flagged it directly: the return-to-office movement may be approaching its peak. Job growth is tepid. Energy costs tied to geopolitical volatility are a real demand-side headwind going into the second half of 2026.

If you’re underwriting office value on the assumption that absorption trends keep improving, you’re probably wrong. If you’re buying distressed office at 10 cents on the dollar with a conversion thesis and patient capital, you might be looking at one of the better risk-adjusted plays in the market right now.

The market isn’t recovering. It’s sorting. Know which bucket your deal is in.

Daniel Kaufman is the Principal & CEO of Kaufman & Company, a Los Angeles-based real estate development and investment firm.