European P2P default & recovery barometer
Default and recovery disclosure is the least standardised metric in P2P lending. Below is what each of the 19 platforms in this barometer actually publishes - and, just as telling, what it does not. Figures come from our own reviews.
Platforms that disclose a recovery / overdue figure
Platforms that publish no default or recovery data
Source: CrowdIndex platform reviews (19 platforms, 2026) + ESMA. "Not disclosed" means the platform publishes no such figure. Reviewed quarterly. Capital at risk.
P2P lending default rates in Europe
There is no official European default rate for P2P lending. No regulator publishes one, and the platforms that do publish figures are not measuring the same thing. This page shows what each of the 19 platforms in this barometer actually discloses, sets it against the two national datasets that come closest to independent, and explains why the numbers you find elsewhere rarely mean what they appear to mean.
Why default rates are not comparable across platforms
Four different things get called a “default rate”, and platforms pick whichever flatters them:
- Overdue. A payment is late, often counted only past 60 or 90 days. It may still be paid in full.
- Default. The loan is formally declared unrecoverable on normal terms. Each platform sets its own trigger, ranging from 30 to 180 days late.
- In recovery. Enforcement has started, usually against collateral. Money may come back, in part, years later.
- Realised capital loss. The case is closed and investors are permanently short. This is the only figure that describes what an investor actually lost, and it is the one least often published.
A platform reporting “0% defaults” while holding 20% of its book in recovery is not lying, it is choosing the flattering denominator. When comparing two platforms, always check which of the four is being quoted, and over what period.
Buyback obligations hide the underlying picture entirely. On marketplaces where the loan originator repurchases late loans, the investor sees no default at all until the originator itself fails. The 2020 collapse cluster and the 2022 originator failures on larger marketplaces both worked this way: years of clean-looking statistics, then a step change. A buyback guarantee transfers credit risk to the originator, it does not remove it.
Young platforms flatter themselves by construction. A platform three years old with a growing book has most of its loans still inside their term. Defaults surface late in the cycle, so a low rate on a young, fast-growing portfolio tells you almost nothing.
What the 19 platforms in this barometer disclose
The barometer above is built entirely from our own platform reviews. Two findings stand out.
Nine of the nineteen publish no default or recovery figure at all. Not a high number, not a low one, nothing. That includes platforms marketing double-digit returns to retail investors. In a sector where the ECSP framework was supposed to raise disclosure standards, silence remains the most common answer.
Among those that do publish, the spread is enormous. At one end, InRento reports zero defaults across 177 completed projects over five years, and Crowdpear reports zero investor losses over three years with recoveries proven in practice. At the other, EstateGuru carries 60.2% of its loan portfolio in recovery, the highest load among licensed platforms still operating, and Reinvest24 has its entire book effectively in recovery with withdrawals frozen since 2024 and alerts from three national regulators.
Between those poles sit the platforms worth studying, because they publish both sides of the ledger. Capitalia discloses a 1.18% realised capital loss alongside 12.9% in recovery, which is the format every platform should be using: what was actually lost, and what is still in play. Profitus reports a 1.83% non-performing loan ratio with zero investor loss to date. Maclear reports a single default of EUR 150,000, roughly 0.15% of funded volume, repaid to investors in full.
One entry deserves a caveat rather than a number. Debitum’s headline zero-loss claim is disputed and the platform has been the subject of an independent investigation in 2026; we show the figure with an asterisk rather than repeating it as fact.
The two national datasets that come closest to independent
No EU-wide default statistic exists. ESMA’s market reports cover volumes, investor composition and cross-border activity, but not loan performance. Two countries fill part of that gap.
Italy. The POLIMI Osservatorio Crowdinvesting collects self-declared default rates from Italian platforms. In its 11th report, presented in July 2026, the declared rates ranged from 0.65% to 32.43% for loans originated in 2024, and from 6.51% to 37.65% for those originated in 2025. Two things matter here. First, the ceiling: a third of a platform’s 2025 originations in default is not a rounding error, it is a broken underwriting model. Second, the spread: a fiftyfold difference between the best and worst platform in the same country under the same regulation tells you that the licence says very little about credit quality. The same report shows the Italian market contracting 36.8% year on year to EUR 164.19 million, with authorised portals falling from 42 to 37.
France. The 2025 crowdfunding barometer from Forvis Mazars and France FinTech, published in March 2026, reports that 20 to 25% of real-estate crowdfunding projects are in insolvency proceedings, and that one project in two shows significant difficulty. Real estate was 47.9% of French crowdfunding collection in 2025, so this is not a fringe segment. The barometer is an industry survey rather than supervisory reporting, and it covers 51 platforms with estimates for non-respondents, but it is compiled by an auditor rather than by the platforms themselves.
Put together, the picture is consistent: in the real-estate segment across southern and western Europe, project-level distress is running in the double digits, and platform-level headline figures frequently do not reflect that.
How to read a platform’s default disclosure
Five questions that separate a real disclosure from a marketing number:
- Which metric, over what period, on what denominator? Percentage of loans, of volume, or of currently outstanding book? Lifetime or last twelve months?
- Is realised capital loss published separately from recovery? If only one number exists, it is almost always the flattering one.
- How old is the loan book, and is it growing fast? A rapidly growing book mechanically dilutes the default rate.
- Is there a buyback obligation, and who stands behind it? If the originator is a related party of the platform, the guarantee is only as strong as the group.
- What happened in the last actual default? Marketing promises collateral enforcement. The useful question is whether enforcement has ever been tested, how long it took, and how much came back.
That last question is the one platforms answer least willingly, and the one that predicts investor outcomes best.
Frequently asked questions
What is the average default rate for P2P lending in Europe? There is no reliable European average, and any single figure quoted as one should be treated with suspicion. No EU regulator publishes loan-performance data, and platforms use incompatible definitions. The closest independent data comes from Italy, where declared default rates ranged from 0.65% to 32.43% on 2024 originations and 6.51% to 37.65% on 2025 originations (POLIMI Osservatorio, July 2026). The spread matters more than any average drawn from it.
Which European P2P platforms have the lowest default rates? Among the 19 platforms in this barometer, InRento reports zero defaults across 177 projects over five years, Crowdpear reports zero investor losses over three years, and PeerBerry reports an eight-year no-loss record including full repayment of loans affected by the war in Ukraine. Maclear reports a single default of about 0.15% of funded volume, repaid in full. Track record length matters: a zero on three years of data is weaker evidence than a zero on eight.
Does a buyback guarantee protect me from defaults? It moves the risk rather than removing it. Under a buyback obligation the loan originator repurchases loans that fall late, so you see no default until the originator itself cannot pay. When originators failed in 2020 and 2022, investors on buyback-backed marketplaces went from apparently clean portfolios to frozen funds in a matter of weeks. Check who stands behind the guarantee and whether that entity is independent of the platform.
Why do so many platforms publish no default data? Because nothing compels them to. The ECSP Regulation imposes disclosure duties around individual projects and a standardised key investment information sheet, but it does not require platforms to publish portfolio-level loan performance. Nine of the 19 platforms in this barometer publish nothing at all, which is legal and also informative.
What is the difference between a default rate and a recovery rate? A default rate tells you how much of the book stopped performing. A recovery rate tells you how much of that is being chased through enforcement, usually against collateral. Neither tells you what investors lost. Only realised capital loss does that, and few platforms publish it. Capitalia is one that publishes both a 1.18% realised loss and 12.9% in recovery.
Are default rates rising in European P2P lending? The evidence points that way in the segments where data exists. Italian platforms declared higher default rates on 2025 originations than on 2024 ones, and the Italian market contracted 36.8% year on year. In France, one real-estate crowdfunding project in two shows significant difficulty. Overall European monthly funding volumes were down about 24% year on year in May 2026. Contraction and rising distress usually arrive together.
Data sources
- CrowdIndex platform reviews (19 platforms, June to July 2026): every figure in the barometer above, extracted from each platform’s own public disclosures at review time. Each review lists its own primary sources.
- POLIMI Osservatorio Crowdinvesting, 11th report (presented 21 July 2026): Italian self-declared default rates by origination vintage, market volume, authorised portal counts.
- Forvis Mazars and France FinTech, “Baromètre du crowdfunding en France 2025” (March 2026): French real-estate distress rates, market volume and segment split.
- ESMA Market Report on Crowdfunding in the EU 2025: confirms the absence of EU-wide loan-performance reporting; source for market structure figures.
- Regulator records cited for enforcement actions and alerts: Banca d’Italia, Consob, EFSA (Estonia), CNMV (Spain), Finanstilsynet (Norway), Bank of Lithuania.
Figures marked “not disclosed” mean the platform publishes no such number. We do not estimate, model or fill gaps. Capital at risk: past loan performance does not predict future performance.