Peer-to-Peer Lending

How it works

A borrower applies through the platform, which pulls credit-bureau data, layers on its own scoring model, and assigns a grade that sets the interest rate the loan will carry. Investors do not fund whole loans; they fund small fractions of many loans at once, so a single account is spread across dozens or hundreds of borrowers rather than concentrated in a handful of names.

In the classic US structure, the investor never owns the underlying loan directly. A partner bank originates it to satisfy lending law, the platform purchases it, and the investor holds a payment-dependent note tied to that specific loan and to the platform's own solvency. The platform then services the loan — collecting monthly payments, chasing delinquent accounts, selling charged-off balances to debt buyers, and passing the remainder through to note holders.

Most consumer loans in this space are unsecured, three- or five-year, fully amortizing instalment loans, so cash comes back as a blend of principal and interest every month rather than as a lump sum at maturity. The retail side of the market has thinned out considerably over the past decade — LendingClub closed its retail notes platform in 2020 and converted to a bank — and the platforms still open to individuals differ in whether investors hold notes, whole loans, or fund pools, and in who is allowed to invest at all.

What it pays

Platforms quote two numbers: a borrower interest rate tied to credit grade, and an expected or historical net return after servicing fees and projected charge-offs. The gross rate moves with credit grade, loan term, and purpose, but the net return that actually reaches an investor is driven almost entirely by realized defaults rather than by the headline rate.

Higher grades pay less and default less; lower grades pay much more and default much more, and the spread between the two only behaves as advertised while the broader economy is behaving. Loss curves are back-loaded, meaning a freshly funded portfolio looks unusually strong simply because its loans have not had time to season and go delinquent — an early result that flatters the strategy before it has been tested.

Automated allocation tools spread new cash across incoming loans by criteria the investor sets in advance, which is how diversification is maintained without manually picking individual notes. Because the loans amortize monthly, principal keeps returning throughout the term, and any of it left uninvested creates a cash drag that quietly lowers the realized return unless it is redeployed promptly.

Costs and taxes

The platform takes a servicing fee out of each payment collected, typically a percentage of the payment, plus a share of whatever is recovered on charged-off accounts sold to debt buyers. There is no bid-ask spread at origination, but where a secondary market exists for selling notes early, it usually demands a discount to par.

Interest received is ordinary income for federal tax purposes, reported on Form 1099-INT or Form 1099-OID depending on how the platform structures its notes. The treatment of losses is the asymmetry that catches investors off guard: for a non-business lender, a defaulted loan is generally a non-business bad debt, deductible only as a short-term capital loss and subject to the annual capital-loss limitation, not as a direct offset against ordinary interest income.

That mismatch means interest is taxed at ordinary rates while losses on the same loans are capped and carried forward as capital losses, so after-tax returns can trail pre-tax expectations by more than a simple netting of gains and losses would suggest. Some platforms offer an IRA option that removes the current-year tax mismatch entirely, at the cost of custodial account fees.

Liquidity and time commitment

Notes are illiquid by design. Secondary markets, where they exist at all, have historically been thin, so the practical way to exit is to stop reinvesting incoming cash and let the existing portfolio amortize down on its own schedule. Full run-off takes as long as the underlying loan terms, commonly three to five years, with principal and interest trickling back in monthly instalments rather than arriving all at once.

Getting started is quick — fund an account, set allocation criteria, and let automated investing build the portfolio over time. What remains ongoing is lighter but not zero: watching charge-off trends, deciding whether to keep reinvesting or wind down, and reconciling tax paperwork that can list many small interest and loss line items across dozens of underlying loans. Investor eligibility also varies by state and by platform, and some venues restrict participation to accredited investors only.

How it goes wrong

The central risk is that credit losses exceed the platform's own projections, particularly in lower grades during a recession, pulling the net return well below the number that was advertised at the time of investment. Because the loans are unsecured, there is no collateral to seize; recovery depends entirely on collection efforts and on selling defaulted paper to debt buyers, often for a small fraction of face value.

Platform risk sits alongside borrower risk and is a separate failure mode. With payment-dependent notes, the investor is an unsecured creditor of the platform itself, so a platform bankruptcy can impair payments even on loans where the underlying borrower is current. Platforms have also changed the rules or withdrawn retail access outright, stranding investors in a run-off portfolio they can no longer add to or easily exit.

Underwriting models are built on recent data, and a model calibrated through a benign credit stretch tends to misprice the next downturn, understating the loss rates that eventually show up. New portfolios compound this problem: they look strong purely because losses have not had time to season, which tempts investors to add money at precisely the point in the loss curve where risk is least visible. Layered on top of all of it is the tax asymmetry between ordinary income on interest and capital-loss treatment on defaults, which quietly narrows the after-tax return relative to what the headline numbers imply.