The office maturity wall is an extension trade with a limit

Yardi Matrix counts 14,000 U.S. office properties carrying $289.2 billion of loans that have already matured or come due by 2028, a 33.5% share of total loan volume. The aggregate invites the crash frame, but the vintage is what a credit desk can price.

Nearly 59% of those loans originated before 2021, according to a report from Yardi Systems' CommercialCafe, written under the baked-in assumption that high demand for office space would sustain the obligations through maturity; hybrid work has since put down persistent roots, office-using employment has been declining since mid-2023, and what demand remains has concentrated largely in the highest-quality buildings. Set those three facts against an origination window that closed five years ago and the wall stops behaving like an event; it becomes a repricing with years still to run, in which the loans repriced first were the ones with the least room to absorb it.

The densest stretch of the schedule sits ahead rather than behind, with maturities expected to peak over the next few years, and Peter Kolaczynski, a director at Yardi Research, does not hedge the read: "The office debt problem is not going away yet. With interest rates and stubbornly low physical occupancy both adding to the headwind, the expectation is for increased delinquencies and distress."

More delinquencies is a different forecast from more sales, and the distance between the two is where this cycle's office debt gets resolved. The refinancing wall is sorting into short bank paper for earned income and equity for the forecast, and the extension trade runs out exactly where sponsor equity isn't there to meet it. Commercial real estate has already run the experiment in another property type: as we reported in August, modifications have kept CLO delinquencies below 1%, which is a measure of how far a maturity date can be pushed when a lender prefers a modification to a workout. The catch travels with the technique, because every pushed maturity becomes a claim on a later year that will need refinancing on then-current terms, and the exits that don't happen now build the test that arrives later.

Office is a harder version of that test than a CLO, because in a CLO the collateral's cash flow held roughly intact while the capital structure moved; in office, the demand moved first. A loan sized before 2021 was underwritten against a breadth of tenant demand that no longer exists in the same shape, so refinancing gets answered building by building rather than by a sector average. That is also why the $289.2 billion reads as granular rather than concentrated: divided across the 14,000 properties Yardi identifies, it works out to roughly $20.7 million of debt per property. Debt that size does not produce landmark defaults; it produces a steady drip of extensions, forbearances, and quiet transfers, and that is the pattern the sector has already been showing.

A further problem sits in the report's own framing. If demand has concentrated in the highest-quality assets, then the loans most likely to refinance cleanly sit on the buildings least likely to need help, and the population that most needs an extension is the population with the weakest collateral behind it. Extensions are cheapest for the sponsors who need them least. That inverted distribution, more than the $289.2 billion total, is what makes the schedule a credit problem rather than an accounting one.

The mechanics are unforgiving in this vintage: a refinancing today is likely to produce a smaller loan than the one it replaces, because the value underneath has moved, leaving a gap that has to be filled before any lender closes, and filling it is what the extension trade is for. A sponsor can cover a modest shortfall once. When the same loan comes back needing another check, the arithmetic stops working for both sides, and that is the point at which the wall stops rolling forward and starts producing the sales that set marks.

Evidence of that transition would show up first at the loan level, and the loan-level record is thin. Connect CRE's weekly rundown of distressed debt returning to lenders named no loans at all in the edition we covered in August, less a reading on the volume of distress than on how much of it is being resolved through modification and forbearance before it reaches a title transfer. Maturity schedules describe a queue; they say nothing about the order in which the queue clears.

Office's clearing price has been set one trade at a time, and the next mark will come from owner-user notes and leasing spreads rather than from appraisals. Maturing debt is where that argument takes credit form: the coupon a lender will accept on an extension is a more current read on a building's value than an appraisal produced for a loan that does not mature until 2028, and the gap between the two is where this vintage's losses get recognized: by the borrower in cash, by the lender at the extension, and only later in a published mark.

Yardi expects delinquencies and distress to increase between now and 2028. Whether the resolution arrives as sales or further extensions is the whole question for office credit at the moment, and the trackers that name actual loans and the extension coupons lenders agree to will answer it well before the aggregate does.