The peer-to-peer (P2P) lending trap

A reader lines up two numbers. A bank fixed deposit is offering around 7% a year. An app is offering a peer-to-peer lending product at "up to 12%". Same money, nearly double the return, and the app looks every bit as polished as a banking app. The obvious question forms on its own: why would anyone leave money in the deposit?

That question is the trap, because it assumes the two 12% and 7% figures are the same kind of number. They are not. A deposit rate is a contractual return from a regulated bank, sitting behind deposit insurance. The P2P 12% is an expectation on a bundle of unsecured personal loans made to strangers - and every rupee a borrower fails to repay is subtracted from the lender, not from the platform. The higher yield is not the answer to a question. It is the invitation to inspect a different set of risks.

This chapter exists because yield language can make credit risk look like savings. The word "returns", a monthly payout, and a clean interface all borrow the feeling of a deposit. Underneath, peer-to-peer (P2P) lending is simply retail lending: an online platform matches individual lenders with individual borrowers, and the lender is now a small, unsecured moneylender to people they will never meet.

The protective habit for the whole of this part is a refusal to analyse the return too early. A high advertised return is not yet an investment case if you cannot say who is regulated, who bears the loss, and what route exists if it goes wrong. A safe-sounding promise hides the real question: who is taking risk, who is paid upfront, and who bears the loss later. On a P2P platform the answer is uncomfortable once you say it plainly: the borrower takes the cash, the platform is paid its fee whether or not the loan is repaid, and the lender bears the default.

There is a second reason this is the book's closing chapter. It gathers every earlier habit in one place. It needs the scam-reading instinct from the chapter on guaranteed returns, the "who protects me and how far" instinct from the chapter on regulators, and the plain arithmetic of the whole shelf: a bigger headline number is not more money in hand.

Read a P2P offer by unbundling the single yield number into the separate risks it hides. There are four, and they do not disappear because the app calls the product "investment".

The first is credit risk. The loans are almost always unsecured - no house, no gold, nothing pledged behind them. If the borrower stops paying, there is little to seize. On a portfolio of such loans, a certain fraction defaulting is not a rare accident; it is the base case, and the advertised yield is a gross figure before those losses are taken out.

The second is platform risk. In India a P2P platform must be a non-banking financial company registered with the RBI as an NBFC-P2P. Crucially, that licence lets it only match and service loans. It cannot lend from its own book, it cannot promise or guarantee a return, and it cannot provide credit enhancement. Lender money must move through a bank-promoted escrow account that the platform never directly touches. So the platform is a marketplace, not a bank - and if the platform itself fails operationally, someone still has to service and recover live loans.

The third is liquidity risk. Your money is inside real loans with real tenures. You get it back as borrowers repay, not on demand. Any "withdraw anytime" or "instant exit" language is doing heavy lifting it is not entitled to; the RBI's 2024 tightening specifically bars platforms from promising liquidity or offering secondary-market-style exits.

The fourth is recovery risk. When a loan goes bad, recovery is slow, partial, and expensive - collection calls, legal notices, and often cents on the rupee. That gap between the loan's stated value and what actually comes back is where the advertised yield quietly leaks away.

The misfire is speed. The reader sees one clean number, feels the pull of the extra 5%, and skips the boring bridge where all four risks live. Registration tells you the platform is allowed to exist; the escrow and disclosure records tell you the money is trackable; the recovery process tells you what your loss path actually looks like. Read those before the yield, not after.

The arithmetic here is not "which product returns the most". It is "what does the advertised number become after the losses the advertisement leaves out". Every figure below is a constructed teaching example, not advice, and not a claim about any real platform. illustrative

Suppose you lend ₹1,00,000 across a P2P platform advertising 12% a year. Because the RBI caps how much a single lender may give a single borrower at ₹50,000, and because sensible lending is spread thin, your money is split across, say, 50 borrowers at ₹2,000 each. That spreading is protective - no single default can sink you - but it does not remove the average. illustrative

Now run a default rate. If 8% of the loan value goes bad over the year and recovery on those bad loans is only a quarter of face value, the loss is roughly 8% × 75%, about 6% of the pool eaten. The gross 12% does not arrive; something closer to 6% does, before the platform's fees are taken, and before tax. The interest you do receive is taxed as income from other sources at your slab rate - while the defaults you suffer usually give no clean tax relief against it. A 30%-slab lender's 6% can quietly become closer to 4% in hand. illustrative

So the working rule is plain: the advertised yield is the top of the range and the worst honest case is well below it. If a platform will not show you its historical default and recovery numbers, you cannot do this subtraction - and a number you cannot reduce for losses is a number you cannot trust.

Read one situation slowly. A reader is comparing a P2P yield to a bank deposit yield. The visible facts are a higher stated yield, borrowers who may default, limited liquidity, and guarantee language that is unclear. Their instinct is to treat the higher number as a better version of a deposit. illustrative

Line up what is actually known. The yield is higher because the risk is higher, not in spite of it - an unsecured loan pays more precisely because it can go unpaid. The liquidity is limited because the money is inside real loans. And the "guarantee" language is the loudest warning of all: a registered NBFC-P2P is explicitly forbidden from guaranteeing returns or protecting capital, so any such promise means either the platform is not registered, or it is breaking its own rules. Read together, these say one specific thing: this is a loan portfolio wearing a savings label.

The repair is to read the offer as what it is. Confirm the platform is a registered NBFC-P2P and find the registration. Check that funds route through a bank-run escrow. Ask for the historical default and recovery record and reduce the yield by it. Confirm there is a named complaint route - the SCORES-style regulated grievance path or the RBI ombudsman - before, not after, money moves. If any of those is missing, the read stops there. This case is built to isolate one reading error; it is not a statement about any real platform.

This reading cannot remove credit risk or market risk. Even a fully registered, honestly run P2P platform lends unsecured money to strangers, and some of them will not repay. Registration makes the loss visible and recorded; it does not make it disappear. Deposit insurance does not stand behind these loans.

It also cannot judge a specific platform. This module teaches why a P2P product should not be read like a deposit; it does not rate any operator, and nothing here says any particular platform is safe or unsafe. That judgement needs the platform's own registration, its audited default and recovery history, and terms that may change.

Most of all, it cannot freeze the rules. Regulation of P2P has already tightened once and can shift again. The durable skill is not memorising the ₹50 lakh cap or the T+1 rule - it is the sequence: name the product honestly as unsecured lending, find the registration and the loss bearer, reduce the yield by realistic defaults, and write down what would change your mind. Verify the current framework from official RBI sources before acting.

People get fooled when the yield is compared without comparing who bears the default. The pleasant half - a number bigger than the FD rate - is read on its own, and the quieter half - an unsecured loan book with the lender at the bottom of it - is dropped.

The traps repeat in a recognisable shape. Studying the promised return before checking registration and recourse. Accepting early, on-time payouts as proof of a real return engine, when they are also exactly what a Ponzi produces and what any loan book shows before defaults surface. Treating a polished, banking-style interface as if it were a bank. And believing a "guaranteed" or "capital-protected" claim that a registered platform is not even permitted to make - which means the claim itself is the red flag.

The repair question is the same one that runs through this whole part: what would change your mind, and which document would prove it? For P2P the document is the registration, the escrow arrangement, and the default-and-recovery history. If those cannot be produced, the yield is not an investment case yet.