P2P Lending: The Global Lesson

Peer-to-peer lending has become something of a buzzword in Nepali finance. Nepal Rastra Bank has published a consultative document; this year’s monetary policy has opened a framework for non-bank lending, and there is talk of pilots with the central bank.

That enthusiasm is not misplaced. This is an exciting space and one Nepal should want to be in. But the idea is 20 years old and has been tried at scale across the globe, so it is worth being clear about what problem we are solving and what has and has not worked.

A Price Problem

Peer-to-peer (P2P) lending emerged in the mid-2000s as a disruptive alternative to traditional banking. Britain’s Zopa pioneered the model in 2005, followed soon after by Prosper and LendingClub in the United States. Their premise was straightforward: banks paid savers minimal returns while charging borrowers, particularly credit card users, interest rates exceeding 20%. P2P platforms sought to bridge that gap by connecting borrowers directly with investors, reducing the role, and cost, of the traditional financial intermediary. 

Remove the bank and the two sides could split it. It served primarily people who already had credit and were overpaying, not people locked out of it altogether. But it became clear that the spread was not the bank’s margin. It was the price of absorbing loans that fail, and the retail lenders who took it inherited the defaults along with it. Zopa ended up taking a banking licence. LendingClub bought a bank. Both were founded to make banks unnecessary, and both decided that being one worked better.

An Access Problem

Closer to home, both of Nepal’s neighbours tried the same structure on a bigger problem. For them it was outright exclusion, not just price. Neither country was short of liquidity. What was missing was institutional appetite for lending to people the banks could not read, so savers were recruited to take the risk institutions refused.

In China, state banks lent to state companies, leaving the private economy starved of credit. Peer-to-peer platforms opened around 2007 and proliferated quickly, with minimal regulation for nearly a decade. As competition for savers intensified they began guaranteeing returns, which quietly turned them into deposit-takers without a licence. Lenders who trusted the platform stopped examining the loans underneath, and a platform did not strictly need to have any. Left unsupervised, a market like this does not merely underperform. It turns fraudulent. China’s regulator later found that 40% of platforms were effectively running Ponzi schemes. Of roughly 6,000 platforms, fewer than 30 have survived.

India recognised the need for regulation early and wrote its first peer-to-peer guidelines in 2017. But as the sector expanded, more risk became evident, and in 2024 the central bank came back with a much stronger set of rules. Growth in pure peer-to-peer lending has slowed sharply since.

Lending Is Two Jobs

Any lender does two jobs. Assessing risk, deciding who deserves the loan. And managing it, absorbing the loss when the assessment proves wrong. Peer-to-peer lending assumed it could beat banks at both. Across every market that has tried it, they have usually beaten banks at the first and struggled at the second.

At assessment they can become formidable, reading bank statements, transaction patterns and digital footprints to ask not only whether a borrower can repay but whether they intend to.

At managing risk they are structurally weak. A platform is not a financial institution. It holds no capital and no balance sheet, so when something goes wrong it has nothing to take the shock. A retail saver who thought she held something close to a deposit is worse placed still.

The model can work. The caution belongs with the people whose money is at risk, not with the idea.

What Nepal Actually Needs

Nepal is not short of liquidity. Banks hold more than Rs 1.5 trillion they cannot lend, while roughly 70% of adults still borrow outside the formal system. What it lacks is the ability to judge a borrower with no collateral. The banks have no incentive to build it. A bank that lends profitably against land will not invent a second method, and has no cost structure to support Rs 50,000 loans. That will come from private platforms, because collateral-free lending is their primary business, and because these models only improve by watching their own loans succeed and fail.

These platforms are not only about individuals lending to individuals. The assessment they build can be used by anyone, including the banks and microfinance institutions that already hold the money and the reach. A platform recruiting only retail investors to lend reaches a few thousand people. The same assessment, put to work by institutions, can reach millions. That is where the real value lies, and why this space deserves encouragement.

None of this argues for leaving the sector alone. That is what China did, and Nepal has seen with the cooperative crash what happens when nobody looks after the retail saver. Regulation matters. What matters more is where it is aimed.

Two things need firm rules. First, who is allowed to lend. Savers with genuine surplus who understand the risk should be permitted, with limits on how much any one person can put in and how much any one platform can take in. Second, what a platform may do with that money. No promised returns, no early exits, and no loan risk on its own books. Most markets that regulated this sector have converged on some version of these rules.

What does not need regulating is the credit decision itself. That is the one thing these platforms do better than any regulator, and it is the whole reason to let them in.