Peer-to-Peer Lending: An Unsecured Loan Book, Held Directly

RISK-FORWARD, and three things belong before anything else. First: you are the lender. Not a depositor, not a saver — a lender making unsecured loans, generally to borrowers who could not obtain bank credit on the same terms, and if they do not repay, the loss is yours. Second: a provision fund is not a guarantee. It is a pot of money set aside by the platform, sized on expected losses, and it can be exhausted — which is most likely to happen precisely when defaults rise, meaning the protection weakens exactly when it is needed. Third: platforms have failed, and a reader should know what happens to their loans if the one they use does. This article explains the mechanics and the arithmetic. It names no platform and recommends nothing.

The economics are legible and the yields look attractive against deposit rates. The comparison to a deposit is the error the whole article exists to correct, because a deposit is a claim on a bank and frequently protected by a compensation scheme, while this is a loan book with the credit risk sitting on the reader.

Where the yield comes from, and what removes it

The portal's credit principle applies unchanged: extra yield is compensation for something. Here the something is specific and identifiable. Borrower credit risk. These borrowers are typically served less well by banks — because of thin credit history, business type, or affordability assessments — so the rate reflects a genuinely higher probability of not being repaid, which is the high-yield principle in a retail setting. No deposit protection. A deposit-taking bank is generally covered by a compensation scheme; a peer-to-peer loan is not, and part of the yield premium is the absence of that safety net. Illiquidity. Loans run to a term, and exiting early depends on a secondary market that may not exist or may be suspended — a facility that can be switched off is not a market, the same point the crowdfunding article made. And platform risk, which is the exposure with no analogue in a savings account. Now what removes the yield, and this is where the arithmetic matters. Defaults are not an accident; they are a rate. A platform advertising 9.5% expects some proportion of loans to fail, and the honest figure is the yield after expected losses and fees. Four things a reader should compute or find before assessing any advertised rate. The default rate and how it is measured — annualised, cumulative, or by cohort, which are different numbers. The recovery rate on defaulted loans, since a 100% loss and a 40% loss are very different. Platform fees, charged on lending, servicing, or both. And the seasoning effect: a new loan book shows almost no defaults because loans have not had time to fail, so a young platform's reported default rate is systematically flattering and improves nothing. That last point is the most useful single item in this article, because it explains how an advertised track record can be accurate and misleading at once.

The provision fund, and platform failure

Many platforms operate a provision fund and it deserves precise treatment, because it is the feature most likely to be read as safety. What it is: money retained from borrower payments, held to compensate lenders for defaults. What it genuinely does: smooths ordinary, expected losses, so a lender does not experience each individual default directly — and that is a real benefit in normal conditions. What it is not: a guarantee. It is finite, it is sized against expected losses, and it is typically discretionary in application. And the structural problem: a fund sized for expected defaults is by construction insufficient for unexpected ones — so it works in the conditions where a lender needs it least and fails in the conditions where they need it most. That is not a criticism of any platform's integrity; it is what a fund of that design can do. Then platform failure, which requires four questions — the same four the portal has now applied to crypto custody, property platforms, farmland vehicles, and crowdfunding. Do you hold the loans or a claim on the platform? Where loans are held directly or through a segregated arrangement, they may survive the platform's insolvency; where the position is a claim on the platform, the lender is an unsecured creditor of a failed company. Is there a wind-down plan, and who administers it? Loans need collecting for years after a platform stops operating, and collection is labour, so someone must be paid to do it or it does not happen. Does the secondary market survive? Generally not. And is the platform regulated, for what activity, in your jurisdiction? What this article will not say is that peer-to-peer lending is a fraud. It performs a real function — funding borrowers banks have retreated from — the disclosure on regulated platforms is substantive, and a lender who diversifies across many loans, understands the default arithmetic, and treats the yield as compensation rather than as interest is making an informed decision. What it will say is that the reader is the lender, the provision fund is not a guarantee, the seasoning effect flatters young platforms, and the correct comparison is not a savings account but a portfolio of unsecured consumer or small-business loans — which is exactly what it is.