A realistic route from two funded accounts to a measured first trade — including the costs, failed fills, and numbers that screenshots leave out.
TL;DR: Open two exchanges, put money on both, pick one method, calculate profit from the average fill price (VWAP) — not the best quote — start with $100–$300, and write down every cost. A big % on screen is a tip to check, not income.
In plain terms: arbitrage means buy cheaper on one exchange and sell higher on another (or open mirrored positions) to earn the price gap—not a coin going “up.” In 2026 simple gaps between majors live for seconds: quotes sync faster, fees are public, and bots take the obvious trades. A manual trader wins through less crowded routes and rejecting trades that fail the math—not click speed.
In 2026, preparation creates the advantage: funds are already on both venues, deposit and withdrawal status is verified, size is calculated from the order book, and the response to a partial fill is written down. A scanner reduces search time, but it does not replace checking each leg.
A realistic beginner goal is not a daily percentage; it is positive net profit after 20–30 small trades. With $2,000 of capital, earning $8–$25 on a productive evening already shows that the math and execution work. Only a proven process should be scaled.
An exchange can suspend deposits or withdrawals, one leg can fill partially, funding can change, and price can move during a transfer. Arbitrage reduces directional market risk but adds execution, venue, and operational risk.
Choose one type and make it repeatable. The detailed execution sequence for every leg is covered in How to Trade Each Arbitrage Type.
A token costs $1.000 on Exchange A and $1.025 on Exchange B. You buy $1,000, transfer the tokens, and sell. The method is easy to understand, but the sale price is not locked while the transfer confirms.
Buy $2,000 of spot and simultaneously short $2,000 of futures when the futures trade 1.4% higher. Directional moves offset each other; profit comes from convergence and may be supplemented by funding receipts.
Open a long on the cheaper exchange and an equal short on the expensive one. If the short receives 0.05% per eight hours while the long pays 0.01%, net funding is 0.04% of the position while rates persist.
Before a futures trade, study the funding rate guide and the rules of arbitrage hedging.
Do not open ten accounts. Choose two liquid exchanges available in your country, complete identity verification (KYC), and confirm limits. One evening should produce one working venue pair, not the maximum number of opportunities.
Open accounts early and complete KYC. Links go to signup / invite pages. This is not an endorsement of any venue — verify availability in your country.
18:00 — secure both accounts with unique passwords, authenticator 2FA, withdrawal whitelists, and anti-phishing codes.
18:30 — verify spot and futures fees, minimum orders, withdrawal fees, daily limits, and contract sizes.
19:00 — deposit $500–$1,000 in stablecoins on each venue so both legs can execute without waiting for a transfer.
19:30 — make a $10 test deposit and withdrawal on the exact network you will use; record the fee and confirmation time.
20:00 — set futures to isolated margin, one-way mode, and no more than 2x leverage for the first sessions.
20:30 — open the scanner, both order tickets, D/W status pages, fee pages, and a blank journal before watching signals.
A scanner API key should be read-only. A trading-assistant key may trade, but it should never have withdrawal permission. To compare venues, use the best exchanges for arbitrage in 2026.
For the first trade, choose a liquid pair, a $100–$300 size, and at least $2 of expected net profit. Tiny profit is not the objective; the size caps the cost of mistakes while you validate the sequence.
Stop rule: if any check fails, there is no trade. Opportunities repeat; capital lost to haste does not return automatically.
If the screen says “buy at $1.00,” that is only the first tiny order in the book. At $1,000 you eat several levels — your average buy is higher. VWAP (volume-weighted average price) is that real average. Put it in your math, not the best bid/ask. Details in Slippage Explained.
If sell VWAP is $1.025, the displayed 2.5% spread becomes about 1.41% before fees. After two 0.1% fees and a 0.2% safety buffer, about $10.10 remains on $1,000. At $2,000, deeper book levels may raise slippage to 2.4% and remove the profit entirely.
Start where combined slippage consumes no more than 25–30% of gross spread. Increase size only after five profitable executions in a row and a fresh VWAP calculation.
Most early losses come not from a wrong price forecast, but from a missing cost line or broken execution order. The amount beside each mistake shows the possible direct damage in one small trade.
Before increasing size, study deposit and withdrawal traps and the 5x leverage rule.
Record more than the result. Each trade needs signal and fill times, pair, venues, arbitrage type, size, best bid and ask, both VWAPs, trading fees, withdrawal, funding, and expected versus actual net profit.
A separate variance-reason field matters more than a pretty profit chart. It reveals what degrades results: delayed second-leg execution, a wrong fee, a funding change, or insufficient book depth.
Every ten trades, calculate average net profit, profitable-trade rate, average variance, and worst loss. If actual results consistently trail estimates, stop and fix the cost model before the next trade.
The minimum stack is an order-book-aware scanner, two exchange accounts, a journal spreadsheet, a password manager, and a 2FA app. Alerts help only after filtering for net profit, volume, deposit and withdrawal status, and update time.
Live books for both legs, VWAP at the selected size, trading fees, withdrawal cost, network status, funding rates, and update time. If a tool shows only spread percentage, you still need to do the final calculation manually.
Start with the step-by-step scanner guide, then compare free scanners for 2026.
Keep the first month simple: one exchange pair, one arbitrage type, $100–$300 per trade, mandatory VWAP, and a journal. The goal is to prove net profit with numbers after every cost. Until then, adding capital only raises the cost of mistakes.
Open the VoltArb scanner →VoltArb calculates actual net profit using order book depth, slippage, fees, and funding rates.
20+ exchanges · 5 arbitrage types · Telegram alerts · Fair price screener · Funding scanner · Real-time Tracker
VoltArb is a free real-time crypto arbitrage scanner monitoring 20+ exchanges — with real profit calculations, not just spreads.