The numbers are real. REO completions climbed 42 percent year over year in April 2026. Foreclosure starts are up 20 percent annually, with 82,631 properties entering the process in Q1 2026 alone. Every outlet that covers real estate ran those figures Thursday and treated them as a housing market story. What they are, structurally, is a work order story. And the work orders are not arriving.
That disconnect is not an accident. It is the product of several compounding mechanisms that every player above the field level understands and none of them discuss publicly. This piece names them.
The Loss Mitigation Holding Pattern
A property does not generate a preservation work order the moment a borrower misses a payment. It does not generate one when the servicer files a notice of default. Under federal mortgage servicing rules, a servicer cannot make the first foreclosure notice or filing unless the borrower is more than 120 days delinquent and there is no complete loss mitigation application pending. That 120-day floor is the beginning of the delay, not the end of it.
Once a loss mitigation application is submitted, the servicer must evaluate the borrower for all available options and provide a written decision within 30 days. If the borrower appeals a denial, that process adds another layer. If the servicer offers a forbearance, the clock extends further. Fannie Mae’s servicing guidelines allow forbearance plans in increments of up to three months, extendable to a cumulative term of 12 months. A borrower who enters forbearance on day one of delinquency and receives the maximum term can be 13 months behind before a foreclosure referral is even issued.
The FHA pipeline adds its own structure. The servicer must evaluate FHA borrowers through a sequential loss mitigation waterfall — repayment plans, forbearance, partial claims, loan modifications — and once a borrower qualifies for any option, the evaluation stops. FHA’s new permanent loss mitigation framework, accelerated under Mortgagee Letter 2025-12 with an October 2025 effective date, now limits borrowers to one permanent loss mitigation option every 24 months. That restriction shortens the runway for serial modification applicants, but it does not accelerate the timeline from default to referral for first-time applicants. It only means the second cycle moves faster.
The practical result is that the 577-day average timeline from default to completed foreclosure recorded in Q1 2026 is not a servicing failure — it is the system working as designed. What enters that pipeline today does not reach REO status for a year and a half. The work orders visible in the ATTOM numbers right now were generated by defaults that occurred in late 2024 and early 2025. The borrowers going delinquent today will produce work for the field in 2027, assuming the loss mitigation waterfall runs its full course.
That lag is structural. It is baked into federal regulation, GSE servicing guides, and investor agreements. No amount of foreclosure volume acceleration at the front end changes the timeline at the back end unless the entire loss mitigation apparatus compresses simultaneously. It has not.
The Occupied Property Problem
Foreclosure filings are a count of legal events. They are not a count of vacant properties requiring maintenance. A property in active foreclosure with an occupant — whether the defaulting borrower, a renter, or a holdover — generates no preservation work orders under HUD 4000.1 or standard GSE servicer guidelines. Occupancy inspection orders may be issued, and those go to inspectors, not FSTs, and at rates that have not moved in years.
The properties producing REO-level preservation demand are vacant properties at conveyance. In April 2026, Texas led the nation with 640 completed REOs, followed by California at 515, Florida at 381, Pennsylvania at 346, and Illinois at 340. Those numbers represent properties that have cleared the entire process and landed in lender hands. Those are the work orders. Five thousand of them nationally in a single month, distributed across a network of nationals, sub-contractors, and regional vendors. That is not volume. That is a maintenance schedule.
The 42,430 properties with active filings in April are not 42,430 work orders. Most of them are occupied. Most of them are in some stage of loss mitigation review. Most of them will not produce a field service order for 12 to 18 months.
The National Aggregator Layer
When a property does become vacant and enter preservation status, the work order does not travel directly from servicer to FST. It travels from servicer to national vendor — Guardian, ServiceLink, MCS, Altisource, Safeguard — and then through whatever subcontracting structure that national has built. Each layer extracts margin. The rate that reaches the FST at the bottom of that chain reflects what remains after every layer above has taken its cut.
This publication has documented that dynamic for over a decade. What is different now is that the labor pool that used to absorb discounted rates has contracted severely. The 80 percent attrition figure from NAMFS is not an estimate. It reflects a workforce that found better pay in other industries and did not come back. The nationals who spent those years compressing rates and expanding scope requirements helped create that attrition. They cannot now expect that a 42 percent increase in REO completions will reconstitute the workforce they destroyed.
The properties cycling through in 2026 will encounter a thinner field than the one that existed in 2019. Fewer technicians, fewer inspectors, fewer qualified vendors with the equipment, insurance, and platform registrations required to accept orders. Some of that work will not get done on time. Some of it will not meet conveyance standards. HUD will charge back. Nationals will pass chargebacks down. The FST who completed the work for $40 will absorb the penalty.
The Question Nobody in the Industry Asks
Every servicer executive quoted in trade press this week cited rising volume as a capacity challenge. Every one of them framed it as a vendor management problem — scaling oversight, tightening timelines, managing documentation. None of them mentioned that the capacity problem has a cause, and the cause is the pricing structure they built and defended for fifteen years.
The foreclosure numbers are real. The work they will eventually generate is real. The question is whether the workforce capable of performing it still exists in sufficient numbers, in the right geographies, at the point when those 577-day timelines finally expire. Based on everything this publication has documented about labor attrition, stagnant rates, and the compounding cost burden on independent FSTs and inspectors, the answer is almost certainly no — and when that becomes undeniable, the industry that caused the shortage will be the first to call it a labor market problem.
It is not a labor market problem. It is a pricing problem that became a capacity problem that is about to become a conveyance problem. The ATTOM data just started the clock. And in light of NFR’s recent demand for EIN from folks working under their Social Security number, we are undoubtedly entering into a new phase of employee misclassification litigation.
