In unlikely-to-pay exposures secured by property the borrower is still paying, the asset is still producing and no enforcement has begun. What follows is a negotiation with rules of its own, governed by a prudential clock that allows a single pause. The path of one position, from classification to exit.
A northern Italian property company, fourteen million euros of mortgage-backed bank debt, a mixed-use building in a provincial capital: two income-producing floors, the rest vacant since the last tenant left. Remaining rents cover the interest but not the amortisation. The bank classifies the position as unlikely to pay. What follows is not one specific transaction, but a sequence the Italian market has repeated many times under different names and numbers.
The classification is about the borrower, not the building
An unlikely-to-pay classification expresses a specific judgment: the bank considers full repayment improbable without enforcing the collateral. It does not say the borrower has stopped paying, nor that the property is worth little, and it opens no proceedings. It is a forward-looking assessment of debt service capacity, made while everything else still functions.
The distance from a bad loan is substantial, and the figures bear this out. According to Banca Ifis's Market Watch, Italian banks held 49 billion euros of non-performing exposures at the end of 2025, an NPE ratio of 2.6%. Within that total, unlikely-to-pay exposures account for roughly 27 billion against 16 billion of bad loans. Most of what remains on Italian balance sheets is not credit to be recovered but credit to be managed.
Two valuations of the same building, both defensible
At classification two numbers coexist, and they almost never match. The first is the bank's estimated recovery value: expected cash flows discounted over an enforcement horizon, net of auction discounts, costs and procedural delays. The second is the property's market value as a going concern, with the vacancy filled and the works completed.
The gap between them can exceed fifty per cent without either appraisal being wrong. They answer different questions. That differential is the space in which the negotiation takes place, and whoever measures it better captures the larger share of it.
The prudential clock allows one pause
Time is the factor that orders everything else, and the borrower does not set it. EU Regulation 2019/630, applicable to loans originated from 26 April 2019, imposes minimum coverage levels that rise with the age of the classification, progressing more slowly for exposures backed by real estate collateral than for unsecured ones.
The detail that matters sits in Article 47c. Granting a first forbearance measure suspends the coverage step-up for twelve months; once the year expires, the percentage is recalculated as though no measure had ever been granted, based on the original classification date. Rescheduling pauses the clock, it does not reset it. A bank granting a moratorium buys twelve months of capital relief and nothing more.
The building, meanwhile, keeps its own agenda: a permit to obtain, a tenant to renegotiate, a site to complete. None of these respect the supervisory calendar. The mismatch between the two clocks is what actually generates loss in positions that deteriorate further.
Exit routes and their invasiveness
Rescheduling is the minimum intervention and buys time rather than solving anything. New money to complete works or fill vacancy is the most effective route and the hardest to obtain, since it asks the bank to increase exposure to an already classified position: this is where negotiated settlement and super-priority mechanisms make a practical difference. Further downstream sit datio in solutum, sale of the property with partial debt release, and single-name sale of the loan.
One further route has built a market of its own: contributing the position to a dedicated fund, converting the credit into units and handing asset management to a specialist. The platforms built by doValue with Aurora RE, the Prelios-Luzzatti fund and the restructuring vehicle set up by Sagitta and Europa Investimenti answer the same need, separating whoever originated the credit from whoever has to work the bricks.
What the price says
UTP portfolios traded at around 43% of gross book value in 2025, against roughly 30% for secured portfolios and 14% for unsecured ones. The formally least compromised category commands the highest prices, which is initially surprising. The explanation lies in what is being bought: a UTP purchaser is not acquiring an enforcement procedure but a business plan that can still be executed.
The figure deserves caution. Banca Ifis notes that UTP pricing shows pronounced heterogeneity, reflecting the features of individual portfolios traded: the average indicates a direction, not the value of a position. Collections moved the same way, rising from 3.1% to 3.8% of residual gross book value, while the share of judicial action fell from 74% to 71%.
Between the 2015 peak and 2027 the cumulative reduction in Italy's non-performing stock is estimated at 108 billion euros, and what remains is not a scaled-down version of what came before. Average portfolio size has fallen from around one billion to 190 million, single-name deals are increasing, the market has turned granular. Germany is travelling the opposite way: commercial property deterioration rose from 2.2% to 6.9% in three years, while the same Italian indicator fell from 7.7% to 4.3%. The expertise built here over a decade of property workouts is beginning to find demand elsewhere.