Lenders modified $2.36 billion in CRE loans between May and July, and only a third were plain maturity extensions. The rest carry reserve requirements, paydowns and rate resets, and every one of those terms becomes a line somebody has to budget.
For most of this cycle, the standard remedy was a maturity extension, and the standard commentary was that lenders were extending and pretending. That is no longer an accurate description of what lenders are doing.
CRED iQ’s tracking, reported by CRE Daily on August 17, counted 82 modified CMBS and CRE CLO loans totaling $2.36 billion between May and July 2026. Maturity extensions were still the largest single category at $802.5 million, or 34% of modified balances. Forbearances followed at $514 million, and combination modifications at $345.6 million. Those three together made up 70.5% of the total.
The interesting number is the remainder. Another $695.4 million, nearly 30% of modified balances, landed in miscellaneous categories: paydowns, rate adjustments, reserve requirements and other remedies mixed to fit the specific asset. Lenders have moved from applying one remedy to negotiating a package.
Two other shifts are worth noting. Multifamily is now the largest source of workout activity, with 35 loans totaling $1.14 billion, or 48.4% of modified balances, as floating-rate resets and slower rent growth catch up with deals underwritten in friendlier conditions. Hotels followed at $493.8 million, retail at $236.7 million, and office at just $226.2 million, or 9.6%. Office problems have not gone away, but the pressure is rotating into sectors that looked more durable.
And the distress is not concentrated in trophy assets. Loans between $20 million and $50 million generated $1.22 billion of modifications, more than half the total. Only two loans above $100 million were modified. The average modified loan was $28.7 million and the median was $23.1 million, which puts this squarely in the range of a single well-run property or a small portfolio, managed by a team of a few folks rather than a workout department.
Here is the part that does not get written about, because it happens after the deal desk closes the file and hands it to asset management.
When the remedy was a clean extension, the operating model barely moved. The maturity date changed in the debt schedule and nothing downstream cared. A negotiated workout is different in kind, not just in degree, because most of the terms being negotiated now are terms that consume cash on a monthly basis.
A workout that requires funding replacement reserves or a capital escrow creates a real, recurring monthly cash requirement that has to sit in the model at the property level. It is not an accrual and it does not net out. If it is modeled as an annual lump somewhere below the property, the monthly cash view is wrong all year, which matters most in the months it is least convenient.
A modified rate, a new floor, or a change in the index means the debt service line you budgeted in the fall is not the line you will actually pay. This is precisely the case a reforecast exists for, and it is why the reforecast matters more than the budget in a workout year. A budget that cannot be re-run cheaply, with the assumption changed in one place, turns a rate reset into a week of work instead of an afternoon.
A lockbox or cash sweep provision means operating cash is released on a schedule and against conditions. The expense budget may be entirely correct in total and still be unfundable in the month it was planned. Timing becomes a real constraint rather than a presentational choice.
Most workout packages carry a debt service coverage test, and a DSCR test is a claim about future net operating income that somebody now has to defend on a quarterly cadence. That changes the standard your projection is held to. An internal budget can be directionally right. A number that feeds a covenant calculation needs to tie to the underlying rent roll and expense detail, and needs to still tie three months later when someone asks how it was built.
This is the shift worth internalizing. Before the modification, the budget was an internal document. Its audience was the asset manager and the owner, both of whom already knew the assumptions because they helped set them.
After the modification, the budget has a reader who was not in the room, is not inclined to be generous about ambiguity, and will be reading it again next quarter. That reader wants the same number to mean the same thing every time, wants to see what changed since the last submission, and wants the build to be traceable from the reported figure back to the lease and the expense account it came from.
Most teams can produce that once. The strain shows up on the fourth submission, when the model has been edited by three different folks, the assumption that produced last quarter’s number is no longer written down anywhere, and reproducing it means rebuilding it. Reporting cadence is the real cost of a workout, and it is almost never in the plan.
There is one technical detail worth flagging, because it is easy to get wrong and expensive to find late.
Capital funded through a workout is frequently work that is also recoverable under your leases. Say a required reserve funds a $60,000 parking lot resurfacing that your leases let you recover over three years. That has to become a $20,000 annual adjustment in the recoverable expense pool, entered once, in one place, linked to the right expense account.
Two failure modes follow. The first is double counting: the capital shows up as a funded reserve on the cash side and again as a recoverable expense, and the property looks worse than it is. The second is more common and harder to spot. The amortization adjustment gets entered as a manual add-back and then stays in the pool after the three years are up, because nothing removes it. It will keep contributing $20,000 a year to your recoverable pool until somebody notices, and the tenants who paid it will have been overbilled for as long as it sat there.
If you are entering amortized capital this budget season, write the note next to the number: what it is, what it recovers over, and the year it comes out. Your successor will not know, and in three years, neither will you.
The headline version of this story is that CRE distress is diversifying, which is true and which every outlet ran. The operational version is that the remedy has changed shape. When workouts were extensions, they were a capital markets event. Now that they carry reserves, resets and tests, they are an asset management event, and the work lands on the same small team already running budget season.
Kardin has spent 30 years on that team’s problem. Kardin Portal is a purpose-built budgeting engine that works alongside your accounting system rather than replacing it, where the rent roll, expenses and recoveries sit in one dependency chain, so changing an assumption re-runs what depends on it instead of starting a rebuild. Portfolio-level reporting includes a Cash Flow Analysis with DSCR, which is a useful thing to have when the number your lender is testing is the number you have to forecast.
If a modification is in front of you and the 2027 budget is due at the same time, it is worth seeing what that looks like on your own numbers. Request a walkthrough and we will use your rent roll, not a demo file.
Budget with clarity.