How ArbiScreen collects prices, calculates net profit, and what it deliberately does not do.
A scanner is only as trustworthy as its numbers. This page documents exactly where our data comes from, how every figure is calculated, how often it updates, and the honest limitations you should keep in mind. If you ever doubt a number, you can verify it yourself on the exchange — that transparency is the point.
All prices come directly from the public market APIs of 17 exchanges — 11 global venues (Binance, Bybit, OKX, KuCoin, MEXC, Gate.io, HTX, Bitget, Kraken, BingX, CoinEx) and 6 regional ones for geo arbitrage (Luno, VALR, BtcTurk, Paribu, WazirX, CoinDCX). We read spot prices, futures/perpetual prices and funding rates. We never use private, paywalled or scraped data — everything can be checked on the exchange itself.
This is the core of ArbiScreen. A raw price gap is gross; what you keep is net. For every opportunity we subtract:
| Cost | What it covers |
|---|---|
| Trading fee (buy) | The taker fee on the exchange where you buy |
| Trading fee (sell) | The taker fee on the exchange where you sell |
| Withdrawal fee | The flat fee to move the coin off the buy exchange |
| Network gas | On-chain transfer cost for the chosen network |
The result is the net profit — shown per opportunity and per $1,000 — so you only act on gaps that actually pay after all costs.
Funding-rate math
For funding-rate opportunities we annualize the rate correctly by the exchange’s own interval: annual % = rate × (24 ÷ interval hours) × 365 × 100. An 8-hour rate is annualized ×3×365, a 1-hour rate ×24×365 — so cross-exchange funding is comparable, not misleading.
For geo arbitrage we derive each country’s FX rate from Bitcoin itself (local BTC price ÷ global BTC price), then compare local altcoin prices to the global market. That surfaces genuine regional premiums (like the ZAR, TRY or INR premium) rather than currency-conversion noise.
A displayed net figure assumes you can fill your order at the shown price and move the coin promptly. Thin liquidity, order-book depth, slippage, exchange latency and withdrawal locks can all reduce the real outcome. That is why we show the age of each spread and always recommend starting small. We correct data issues quickly — if you spot one, tell us.
