Buyback Guarantee in P2P Lending: What It Actually Covers

Almost every consumer-lending platform in Europe leads with the same reassurance. If your borrower stops paying, someone will buy the loan back from you at full value. It is the single most effective marketing line the sector has ever produced, and it is the reason many people treat peer-to-peer lending as a savings product with a better rate.

It is also the most widely misread term in the industry. A buyback is a commercial promise from one private company, backed by that company’s balance sheet and nothing else. This guide explains what the promise actually says, the four different things that have happened when it was tested for real, and the five checks worth running before you let it influence how much you invest.

📊 CrowdIndex Editor’s Pick: Maclear ranks #1 of the 19 European platforms we track, with realised yields of 14.5% to 14.9% and one borrower default in three years of lending. It offers no buyback at all, which is part of why it appears in this guide: section 6 explains what it offers instead, and where that model has its own gap. Read the platform card → | Visit Maclear →

---

TL;DR

• A buyback is a contractual obligation from the loan originator, the company that made the loan, to repurchase it at nominal value plus accrued interest once the borrower is a set number of days late. On most platforms the trigger is 60 days; Robocash uses 30, the shortest in the market.
• It is not insurance, not a deposit guarantee and not backed by any public scheme. Mintos renamed its own feature from “buyback guarantee” to “buyback obligation” precisely because investors were reading more protection into the word than the contract provides [source: Mintos blog].
• A buyback covers borrower risk only. It does nothing about originator risk or platform risk, and those are the two failure modes that have actually destroyed investor capital in Europe since 2020.
• It has been tested four different ways. It vanished entirely (Grupeer, Envestio, Finko’s Varks), it stalled for years (Creditstar on Lendermarket), it was renegotiated at a haircut (Twino at 80% of capital), and it worked in full under extreme stress (PeerBerry, €51.4 million of war-affected loans repaid by December 2024).
• The useful questions are who signs the obligation, whether a group guarantee sits behind it, and whether that guarantee is legally binding. On Robocash the group-level backing is an operational policy of the holding company rather than a contract owed to investors.
• Some platforms skip the mechanism entirely and secure loans against assets instead. That trades one set of weaknesses for another, as section 6 sets out honestly.

---

1. What a Buyback Obligation Actually Says

Strip out the marketing and the mechanism is short. You fund a slice of a loan that a lending company has already issued to a borrower. That lending company, called the loan originator, signs an agreement with the platform. If the borrower is more than a set number of days late, the originator must buy your slice back at its nominal value plus the interest that accrued while it was delinquent. You get your money and your yield, and the originator keeps the bad loan and the problem of collecting it [source: Mintos help centre; Esketit; P2PMarketData].

Two details matter more than the headline.

The first is who owes you the money. It is not the platform. The platform runs the marketplace and enforces the contract; the obligation sits with the originator. A marketplace carrying dozens of independent lending companies gives you dozens of separate credit exposures, each as good as the company behind it. A platform listing loans from one company that shares owners with the platform itself backs both the loan and its remedy with a single balance sheet, which is a much thinner form of protection than the wording suggests.

The second is the trigger. Sixty days is the industry convention, so a loan can be delinquent for two months before anything is owed to you. Robocash sets the trigger at 30 days on its 30 to 90 day consumer loans, the shortest in the segment. Platforms with long grace periods or repeated extension rights can keep a loan technically current for far longer, which is the loophole section 4 covers.

2. Why the Word “Guarantee” Quietly Disappeared

Mintos, the largest platform in the segment and one of a small group holding a full MiFID II investment-firm licence from Latvijas Banka, used to call the feature a buyback guarantee. It renamed it a buyback obligation, and explained why in its own words: investor feedback showed the word “guarantee” was misleading, because it implied a level of security the arrangement did not carry. Only the word changed. The legal agreements were untouched [source: Mintos blog, “We are renaming buyback guarantee to buyback obligation”].

That is a rare piece of honesty and it is worth taking literally. In everyday language a guarantee implies a third party standing behind an obligation, ideally one that cannot fail. A deposit guarantee scheme works that way; in Switzerland esisuisse covers CHF 100,000 per depositor per bank. A buyback works nothing like that. It is a promise from a company that is in the business of lending money to people with poor credit, and it becomes worthless in exactly the circumstances that would make you want to use it. Most regulated platforms have followed the terminology, so wherever the word guarantee is still used without qualification, treat it as a signal about the platform’s disclosure culture rather than about the strength of the protection.

3. Three Risks, One of Them Covered

The clearest way to see the limits of a buyback is to separate the three ways a P2P investment loses money.

Borrower risk is the smallest of the three and the easiest to manage yourself, because holding 50 loans instead of five already dilutes it. Originator and platform risk are the ones that have wiped out European retail investors, and the buyback is silent on both. At least ten European platforms have failed, frozen withdrawals or attracted a regulator warning since January 2020, from the 2020 Estonian cluster to the 2025 and 2026 Italian enforcement wave, and a buyback protected nobody in any of them [source: P2P Platforms That Failed].

This is the core misreading. Investors treat the buyback as protection against the thing that is most likely to go wrong. It is protection against the thing that is least likely to hurt them.

4. Four Things That Happen When a Buyback Is Tested

The mechanism has now been stress-tested repeatedly across a full market cycle. The outcomes fall into four distinct patterns, and they are worth knowing individually, because platforms tend to quote whichever one flatters them.

It disappears with the originator. In early 2020 Grupeer suspended lending and halted redemptions, citing the pandemic; investors have since alleged fraudulent practice and litigation followed. Envestio shut down in January 2020 and was declared bankrupt that June. In March 2020 the Armenian regulator withdrew the licence of Varks, part of the Finko group, and Finko was reported as unable to cover an outstanding portfolio of around €20 million. In each case the buyback text was still in the contract and still worth nothing, because there was no solvent company left to honour it [source: Kristaps Mors 2020 platform summary; P2P Platforms That Failed].

It stalls without ever formally breaking. This is the pattern investors underestimate. Between 2022 and 2024, Creditstar, the dominant originator on Lendermarket, hit liquidity problems and loans were extended repeatedly, up to six 30-day extensions, so a position could remain outstanding for 240 days before the buyback trigger was reached. Money was not lost, it was immobilised, in some cases for years, while investors held an asset they could neither exit nor spend. Lendermarket 1.0 closed on 31 December 2025 with its pending-payments backlog reported cleared, and version 2.0 now operates under an ECSP licence from the Central Bank of Ireland [source: Lendermarket updates; P2P Empire; Jean Galea, 2026].

It is renegotiated at a discount. When Twino’s Russian book was frozen, the platform offered investors a buyback at 80% of capital plus 100% of accrued interest. Most investors got most of their money back, which is a far better outcome than the first pattern, but the contract said nominal value and the settlement did not. A buyback taken at a haircut is still a loss [source: Mintos vs Twino research].

It works in full, under conditions worse than anything modelled. In February 2022 roughly one third of PeerBerry’s loan book, about €51.4 million, was tied to borrowers in Ukraine and Russia and was frozen by the war. Under the Aventus group cross-guarantee, the group repaid every one of those loans to investors in monthly tranches, in full and with accrued interest, completing on 16 December 2024. PeerBerry holds no MiFID II or ECSP licence and no compensation scheme stood behind it. It paid because the group chose to and could afford to, which is the cleanest evidence anyone has produced that the mechanism can hold [source: PeerBerry vs Robocash].

Read together, those four outcomes say something specific: the buyback text is nearly identical across platforms, and the results were completely different. What varied was the financial strength and the conduct of the company behind it. That is the thing to research.

5. Five Checks Before You Rely on a Buyback

1. Name the obligor. Find the specific legal entity that owes the repurchase and check whether it is independent of the platform. Shared ownership between platform and originator means one balance sheet is standing behind both the loan and its supposed remedy.
2. Ask whether a group guarantee sits above it, and whether it binds. This distinction is not academic. PeerBerry’s Aventus cross-guarantee is contractual and was executed at scale. On Robocash the group-level backing has been confirmed by UnaFinancial as an operational policy of the holding company rather than a contract owed to investors, even though its payment record is clean [source: PeerBerry vs Robocash].
3. Check for published, audited originator accounts. A buyback is a credit exposure to a company. If that company does not publish audited financials, you are extending unsecured credit to an entity whose solvency you cannot assess.
4. Measure concentration. PeerBerry’s book is more than 83% Aventus Group and Robocash’s is 100% UnaFinancial Group. Holding 200 loans across one originator group is one credit exposure with 200 line items, not a diversified portfolio.
5. Read the trigger and the extension rights together. A 60-day trigger with six permitted 30-day extensions is functionally a 240-day trigger, and that gap is where the Lendermarket experience of 2022 to 2024 happened.

6. The Other Model: Collateral and Provision Funds

Not every platform uses a buyback. Business-lending platforms usually secure the loan against an asset instead, and some add a reserve pot funded from the platform’s own revenue.

Our top-ranked platform is built this way. Maclear offers no buyback obligation at all. Instead each loan is secured by a pledge of borrower assets, with Maclear acting as collateral agent for all investors in that loan, and a provision fund is financed by 2% of the platform’s commissions on successful projects and secondary-market transactions [source: Maclear-full §5]. In theory this is stronger, because a claim over a specific asset survives the borrower’s insolvency in a way an unsecured promise from a lending company does not.

In practice it carries its own untested gap, and we would rather state it than leave it out. Maclear has had one borrower default in three years, an Italian company called Vibroedil that borrowed €150,000 in April 2025 and filed for insolvency in July. Investors were repaid in full by November 2025, an outcome almost unheard of in this sector. But the money came from the founders’ personal funds, not from enforcing the collateral and not from the provision fund, so the advertised protection chain has still never been executed against a defaulting borrower [source: Maclear-full §17, §18]. A 2% fund is also a cushion rather than a guarantee, and it would not absorb a broad wave of defaults.

Both models are promises with different weak points. A buyback is a strong promise from a company you cannot inspect; collateral is a weaker promise attached to an asset you can, enforced through courts that are slow and expensive. Neither replaces holding enough separate positions, on enough separate platforms, that any single failure is survivable.

📊 Editor’s Pick: Maclear - realised yields of 14.5% to 14.9%, a self-reported default rate of 0.15%, a €50 minimum per loan and one default in three years, repaid in full. The honest caveats stay attached: its PolyReg SRO membership is an anti-money-laundering regime and not an investor-protection licence, there is no compensation scheme, its 2023 accounts were published 14 months late and unaudited with 2024 still outstanding, the CNMV in Spain has published a notice (a notice, not a sanction), and the collateral chain remains untested. New investors get a €15 welcome bonus on a first investment of €50 or more within seven days, plus a separate cumulative reward of €30 per €500 invested [source: Maclear-full §15]. Read the full review → | Visit Maclear →

Affiliate disclosure: we earn a commission if you open an account through our links. It does not change what we write, and our ranking methodology and full disclosure explain how we handle it. Capital at risk.

FAQ

What is a buyback guarantee in P2P lending?

It is a contractual commitment from the loan originator, the lending company that issued the loan, to repurchase your investment at nominal value plus accrued interest once the borrower is a set number of days late. Sixty days is the usual trigger and Robocash uses 30. Regulated platforms now generally call it a buyback obligation rather than a guarantee, because the obligation is only as strong as the private company that signed it.

Is a buyback guarantee safe?

It is safe against the risk it was designed for, an individual borrower failing to pay, and offers nothing against the two larger risks. If the originator becomes insolvent, the obligation survives on paper and you join the queue of unsecured creditors. If the platform itself fails or freezes, your whole balance is beyond reach regardless of how the loans perform. That second scenario is what caused most European retail losses since 2020.

Has a buyback guarantee ever failed in Europe?

Yes, in several different ways. It disappeared with the originator at Grupeer, Envestio and Finko’s Varks in 2020. It stalled for years on Lendermarket, where Creditstar liquidity problems allowed repeated 30-day extensions and stretched positions to as much as 240 days. Twino settled frozen Russian loans at 80% of capital plus full interest, so investors recovered most but not all of their principal.

Which platforms have the strongest buyback record?

On evidence rather than marketing, PeerBerry has the clearest pass: about €51.4 million of loans frozen by the war in Ukraine and Russia were repaid in full with interest by 16 December 2024 under the Aventus group cross-guarantee. Robocash has an unbroken payment record and the shortest trigger in the market at 30 days, though its group-level backing is an operational policy rather than a contract owed to investors.

Is collateral better than a buyback?

They fail differently rather than one being universally better. Collateral gives you a legal claim over a specific asset that survives the borrower’s insolvency, but enforcing it takes courts, time and money. A buyback is faster and simpler when it works, and worth nothing when the company behind it cannot pay. Before trusting either, check whether the platform has ever executed it and what it actually recovered.

What to read next

• What Happens When P2P Loan Defaults - the full process that starts once a loan goes bad, traced through one real €150,000 default.
• Safest P2P Platforms Europe - how the 19 platforms we track rank on protection rather than yield.
• P2P Platforms That Failed - the platform-risk column of the table in section 3, told as case studies.
• PeerBerry vs Robocash - the two cleanest buyback records in Europe compared side by side.
• Are P2P Investments Safe - the general safety framework these checks sit inside.
• Diversified P2P Portfolio - how many positions it takes for a single failure to be survivable.
• Loan Originator Risk Explained - how to assess the lending company that actually owes you the money.
• P2P Collateral LTV Explained - what a pledge and an LTV ratio actually give you when enforcement starts.