P2P arbitrage is the oldest, most manual face of crypto arbitrage: you buy crypto cheaply from one person for fiat on a peer-to-peer board, then sell it dearer to another person for fiat, and pocket the spread between the two ads. There's no exchange order book in the middle - just two human counterparties, an escrow that holds the coin while fiat moves through a bank or a payment app, and a board that runs KYC on both of you. That makes P2P a fundamentally different class of arbitrage from the crypto-to-crypto routes Finder computes: it's fiat-to-crypto, manual, and counterparty-based, with risks that no spread calculator can price for you. Below we break down what P2P arbitrage actually is, the mechanics of one cycle, where the edge comes from, and why the banking and dispute risk makes it harder than a pure crypto route - not a get-rich scheme.
What P2P arbitrage is
A P2P (peer-to-peer) board is a marketplace where individuals post ads to buy or sell crypto for fiat. The exchange itself isn't your counterparty - another user is. P2P arbitrage exploits the fact that the buy ad price and the sell ad price for the same coin are not the same number, and the gap between them can exceed the fees.
The cycle in one sentence: find a seller advertising USDT cheaply, buy it for fiat, then find a buyer willing to pay more, sell it to them for fiat, and keep the difference. Nothing here touches an exchange's spot order book. Both legs are fiat-on-one-side, crypto-on-the-other, negotiated with a person and settled through an escrow.
This is what separates it from everything else on this blog. A spot route or a cross-exchange route is crypto-to-crypto: capital is a stablecoin start to finish and only moves between exchanges over the blockchain. P2P deliberately re-enters fiat on both ends. That single fact - fiat in the loop, a human on the other side - drives every difference in speed, risk, and how the "edge" behaves.
How one P2P cycle works: escrow and payment methods
The mechanic that makes a stranger-to-stranger trade survivable is escrow. When you agree to buy, the seller's crypto is locked by the board before any fiat moves. You send fiat through the agreed payment method, mark the trade as paid, the seller confirms receipt, and only then does escrow release the coin to you. The board is the referee, not the counterparty.
Two parts of the mechanic carry most of the risk and most of the edge:
- Payment method. Each ad specifies how fiat is accepted - an instant-payment system, transfer by account details, a payment service, cash in person. The method sets the speed (instant vs hours), the limits, and crucially the reversibility: some methods can be charged back days later, which the seller has already released the coin against.
- The buy/sell spread. The seller's ask and the buyer's bid are two independent prices set by two different people reacting to local fiat demand. The arbitrage lives in that gap, after the board's trade fee and any payment-method cost.
Where the edge comes from
A P2P spread isn't magic - it comes from real frictions in the fiat layer that don't exist inside a crypto order book:
- Differences between P2P boards. The same USDT trades at different fiat prices on different boards because each board has its own pool of buyers and sellers. A coin that's cheap where sellers crowd in can be dear where buyers do.
- P2P vs spot. A board's price can drift from the exchange's own spot price. When P2P sellers undercut spot, you can buy on P2P and exit on spot (or vice versa) - though that exit leg is then a normal crypto trade, not P2P.
- Regional fiat premiums. Demand for stablecoins in a given currency - capital controls, weak local banking, a rush to dollars - pushes the local fiat price of USDT above the global one. That premium is the classic P2P edge, and it's exactly where the legal line below gets dangerous.
The honest read: a wide P2P gap usually exists because something is hard - a slow payment rail, a thin board, a jurisdiction people are wary of. The gap is the market paying you for friction and risk, not a free lunch. The bigger the printed premium, the more carefully you should ask what's behind it.
The risks that make P2P harder than crypto routes
This is the section that matters most, because P2P carries categories of risk a crypto route simply does not have. None of these can be computed away by a scanner - they live in the fiat and human layer.
- Chargebacks and disputes. The signature P2P risk. A buyer pays you, you release the coin, then the buyer reverses the payment through their bank or payment provider. The coin is gone irreversibly, the fiat is clawed back, and you eat the loss. Disputes also freeze a trade mid-cycle while the board investigates.
- Account freezes by banks. Frequent, identical, or fast-cycling fiat transfers are exactly the pattern banks flag as suspicious. A bank can freeze the account with your working capital in it, with no warning and a slow appeal. This is the single most common way P2P arbitrageurs get stuck.
- KYC / AML and the legal line. See the warning below - this is where P2P stops being a trading technique and becomes a legal exposure.
- Counterparty fraud. Fake payment confirmations, doctored receipts, pressure to release escrow early, social-engineering to trade off-platform where there's no escrow. The human on the other side may be trying to rob you, not trade with you.
- Manual speed limits. Every leg is hand-done: post or pick an ad, message, transfer fiat, wait for confirmation, release. A cycle takes minutes to hours, and the price can move against you the whole time. You cannot scale this the way you scale a computed crypto route.
"P2P without cards" or "P2P off the exchange" is not anonymous and is not a legal grey zone you can ignore. Every escrow board runs KYC and keeps a full trade history tied to your identity. Switching the payment method is legal. Using P2P to launder funds, to dodge KYC, or to move money you shouldn't is what gets accounts permanently banned and funds frozen - by the board and by the bank. The legal line is not the payment method, it's the intent and the source of funds.
Example: one P2P cycle with numbers
Trace a single P2P round-trip on $5,000 of working capital, in a generic fiat we'll call units (u). The point is to show where the fee and the risk sit, not to promise the spread.
Start: 5000 fiat units (u) sitting in a bank account Step 1 — buy on P2P: seller's ad, price 1.000 u per USDT 5000 u → 5000 USDT, locked in escrow, then released board fee on buy 0 (maker-side often free) Step 2 — hold: 5000 USDT on the board, price can move while you find a buyer Step 3 — sell on P2P: buyer's ad, price 1.018 u per USDT 5000 USDT → 5090 u, you wait for fiat, then release escrow board fee ~0.2% of notional −10 u payment-method cost −10 u ────────────────────────────────────────────────────────────── Gross: 5090 − 5000 = +90 u. After fees ≈ +70 u on 5000 u (~1.4%). BUT: if the buyer charges back the 5090 u after you release the coin, the result is −5000 u, not +70 u. That tail risk has no equivalent in a crypto route, where an on-chain transfer is irreversible.
The arithmetic looks clean, and on a good day it is. What the numbers cannot show is the distribution: most cycles land near +70u, and a rare one lands at -5000u when a chargeback or a frozen account hits. That asymmetry - small frequent gains against a large rare loss you cannot reverse - is the real shape of P2P, and it's why the printed spread overstates the edge.
P2P vs crypto-to-crypto arbitrage: an honest comparison
Both are arbitrage, both monetise a price gap, but they are different instruments with different risk shapes. The contrast is the whole point of this article.
| P2P arbitrage | Crypto-to-crypto | |
|---|---|---|
| What moves | fiat ↔ crypto, both ends | stablecoin, start to finish |
| Counterparty | a human, per leg | the exchange order book |
| Settlement | reversible (chargeback) | irreversible (on-chain) |
| Banking risk | high (freezes, limits) | none |
| Speed | minutes to hours, manual | seconds, automatable |
| Scales by | manual throughput | book depth + network |
| Edge is | negotiated, fragile | computed, transparent |
Neither is strictly "better" - they solve different problems. P2P is how fiat enters and exits crypto in places where banking rails are thin, and that on-ramp role is real. But as a repeatable arbitrage engine it is bounded by hand-speed and exposed to banks and strangers in a way that the math cannot price. A crypto route is the opposite: the spread, the fees, the transfer cost and the deposit/withdrawal status are all numbers you can see before you commit, and the transfer that closes the loop is irreversible.
If what you actually want is a repeatable, computable arbitrage process rather than a fiat on-ramp, the crypto-to-crypto side is where that lives. There the route is numbers - spread minus fees minus network, with an honest deposit/withdrawal check - rather than a negotiation with a stranger over a reversible payment. That's the side Finder is built for. For the crypto-to-crypto side (not P2P), Finder computes net spreads across 20+ exchanges in the spread scanner.
FAQ - P2P arbitrage
What is P2P arbitrage in simple terms?
Buying crypto cheaply from one person for fiat on a peer-to-peer board, then selling it dearer to another person for fiat, and keeping the spread. There's no exchange order book - just two human counterparties and an escrow holding the coin while fiat moves.
Where does the P2P edge actually come from?
Price differences between P2P boards, drift between a board's price and spot, and regional fiat premiums where local demand for stablecoins is high. The gap is the market paying you for friction and risk in the fiat layer, not a free lunch.
Is P2P arbitrage legal?
Switching the payment method - including not using a card - is legal in itself. Using P2P to launder funds or to bypass KYC is not, and it gets accounts banned and funds frozen by both the board and the bank. The legal line is the intent and the source of funds, not the payment method.
Why is P2P harder than crypto-to-crypto arbitrage?
Chargebacks reverse a settled trade, banks freeze accounts over fast-cycling transfers, counterparties can defraud you, and every leg is manual. A crypto route has none of those - the transfer is irreversible and the route is computed up front.
Does Finder scan P2P spreads?
No. Finder has no notion of fiat, banks, or P2P boards. It computes pure crypto-to-crypto routes between exchanges in the spread scanner. This article is informational about P2P as a separate class of arbitrage, not a Finder feature.
Not financial advice. P2P arbitrage carries risks a calculator cannot price for you: a chargeback can turn a +1% cycle into a total loss of the leg, a bank can freeze your working capital without warning, and a counterparty may be defrauding you. The printed spread is the best case, not the expected one. The decision to trade, and its consequences, are yours.
Read on: the closest sibling to this piece is crypto arbitrage without cards, which contrasts P2P-without-a-card against pure crypto routes. The big picture sits in the pillar Crypto arbitrage guide. The crypto-to-crypto counterpart of P2P is spot arbitrage. And the mistakes that catch people - including treating a printed spread as the edge - are in arbitrage mistakes. Live crypto routes, not P2P, are in the web dashboard.