Is Crypto Arbitrage Trading Profitable? The Honest Math

Crypto arbitrage trading is the practice of buying an asset at one price and simultaneously selling it at a higher price somewhere else, pocketing the difference. In theory it is risk-free. In practice it is a latency, balance-sheet, and execution arms race that has industrialised over the last five years. Retail discretionary arbitrage in liquid majors has effectively zero edge after fees. Where retail can still play: cross-protocol gaps, regional rate-limit arbitrage, and triangular paths in less-liquid altcoins.

The four arbitrage families

1. Cross-exchange (geographical): BTC on Exchange A trades $1 above BTC on Exchange B. Buy on B, sell on A, capture $1.
2. Triangular: BTC/USDT, ETH/USDT, ETH/BTC are mispriced relative to each other. Three trades close the loop.
3. Cross-protocol (DeFi vs CeFi): an asset trades $5 cheaper on a DEX than on a centralised venue. Profitable while gas + slippage stays under $5.
4. Funding rate: spot long + perpetual short. If funding is positive (longs pay shorts), the short collects. Net: cash-and-carry that earns the funding rate, market-neutral on price.

Why the edge has compressed

Three forces have squeezed retail arbitrage to near-zero on liquid majors:

• Co-located market makers. The biggest crypto firms run sub-millisecond latency. By the time a retail trader sees a $1 gap on a price feed, the gap has been arbitraged.
• Fee asymmetry. Retail pays 0.05-0.1% per trade. Institutional pays 0.0-0.02%. A 0.1% gap is profit for the institutional desk and a loss for the retail desk.
• Withdrawal bottlenecks. Cross-exchange arbitrage requires moving capital between venues. Withdrawal queues, ID-verification holds, and on-chain confirmations turn 30-second arbitrage into 30-minute risk exposure.

Where retail can still find edge

Two pockets in our experience:

1. Funding rate cash-and-carry. When funding rates spike on a memecoin or a hyped narrative (annualised 30-50%+), the spot-long perp-short pair can earn that funding net of execution. Capital-intensive but mechanical.
2. Cross-protocol DeFi. New L2 launches and DEX-CEX listings create temporary mispricings. Edge requires gas-management discipline and a tolerance for smart-contract risk.

The math of a real arbitrage trade

Suppose BTC bid is $60,005 on Venue A and ask is $60,000 on Venue B. Apparent spread $5 per unit, or 8.3 basis points.

• Round-trip taker fee: 10 bp (5 bp each side).
• Withdrawal fee: 0.0001 BTC = $6 at $60,000.
• Latency risk: a 200ms delay could see the gap close 30% of the time.

Net expected: 8.3 bp gross – 10 bp fees – 1 bp withdrawal – 2 bp latency cost = -4.7 bp. Negative expected value. This is why retail discretionary arbitrage on liquid majors is a learning exercise, not a strategy.

What it actually costs to do this professionally

• Capital. $1m+ across multiple venues to capture meaningful basis points.
• Infrastructure. Co-located servers, websocket feeds, custom matching engines. $50k-150k setup, 10-30k/month run rate.
• Headcount. One quant developer plus one ops person, minimum.
• Fee tier. Either VIP discounts (volume-based) or maker rebates. Plain retail fees kill the trade.

Honest framing

If you are reading a guide on crypto arbitrage with the intent of running it as a strategy, the realistic next step is one of three: (1) accept that arbitrage in 2026 is institutional, and pick a different strategy; (2) focus on funding-rate cash-and-carry, which is mechanical and tolerates retail latency; or (3) treat arbitrage as a tool for understanding market microstructure, then apply that understanding to directional strategies. Volity supports cash-and-carry through CFD exposure on perpetuals plus spot reference pricing on the same account.

Frequently asked questions