🔄

Crypto Arbitrage Calculator

Calculate your exact net profit when buying on one exchange and selling on another, accounting for all fees.

$
Exchange A (Buy)
$
%
Exchange B (Sell)
$
%
$
⚖️

Enter prices and fees across both exchanges to find hidden arbitrage profits.

Arbitrage looks like free money: buy at $50,000 on one exchange, sell at $50,500 on another, pocket the difference. Then you subtract the trading fees, the withdrawal fee, and the network cost — and the $500 spread has become a $47 loss. This calculator does that subtraction before you commit.

Why most arbitrage is not profitable

There are four costs between a visible spread and actual profit. You pay a fee to buy on exchange A. You pay a network fee to move the coins to exchange B. You pay a fee to sell on exchange B. And while all that is happening — often 10 to 60 minutes for a Bitcoin transfer — the price is moving. The spread you saw may not exist when your coins arrive. On a $1,000 trade with a 1% spread, you make $10 gross. Two 0.1% fees take $2. A $5 network fee takes half of what remains. You have made $3 for an hour of work and considerable execution risk. This is why arbitrage is dominated by bots holding balances on both exchanges simultaneously — they never transfer, so they never wait. The opportunity you can see on a chart is usually the one they have already left behind.

The formula

Coins Bought = (Investment ÷ Buy Price) × (1 − Buy Fee%) Gross Revenue = Coins Bought × Sell Price Net Revenue = Gross Revenue × (1 − Sell Fee%) − Transfer Fee Net Profit = Net Revenue − Investment Gross Spread = (Sell Price − Buy Price) ÷ Buy Price × 100

Compare the gross spread against the net ROI. If a 1% spread produces a 0.3% net return, then fees consumed 70% of the opportunity — and that is before any price movement during the transfer.

Arbitrage in practice

  • 1Hold balances on both exchanges. This is the actual technique — sell on the expensive exchange and buy on the cheap one simultaneously, then rebalance later. No transfer, no waiting, no price risk.
  • 2Use fast, cheap networks for transfers if you must move. Bitcoin can take an hour and cost several dollars. USDT on Tron or Solana settles in seconds for pennies.
  • 3A large spread is usually a warning, not an opportunity. Ask why. Often the answer is that the expensive exchange has withdrawals suspended, or the coin is a different wrapped version, or liquidity is so thin the price is meaningless.
  • 4Check withdrawal limits and KYC tiers before you need them. Being unable to move funds when your position is open is how arbitrage becomes an unintended long position.
  • 5Include the deposit confirmation requirement. Some exchanges need 6 confirmations before crediting your Bitcoin. That is an hour where your capital is stranded and the market is moving.

Frequently asked questions

Is crypto arbitrage still profitable?

For bots with capital on multiple exchanges and sub-second execution, yes, on thin margins at high volume. For a person clicking manually, almost never — the spread you spotted is already gone by the time you have logged into the second exchange.

What spread do I need to make it worthwhile?

Depends on your fees, but as a rough guide: with 0.1% fees on both sides and a $5 transfer cost on a $1,000 trade, you need roughly a 1% spread just to break even meaningfully. Use the calculator with your real numbers — the answer is often sobering.

Why do exchanges have different prices at all?

Each exchange is a separate order book with its own supply and demand. Regional demand, deposit restrictions, and liquidity differences all cause divergence. Arbitrageurs are the mechanism that closes those gaps — which is why they close so fast.

What is triangular arbitrage?

Exploiting a price inconsistency between three pairs on a single exchange — BTC/USDT, ETH/BTC, ETH/USDT, for example. It requires no transfers, so it is faster, but the opportunities are tiny and disappear in milliseconds. Bots only.

Does this account for price movement during the transfer?

No, and that is the biggest unmodelled risk here. This assumes both prices hold while you execute. In reality, an hour-long Bitcoin transfer during a volatile period can turn a calculated profit into a real loss. Treat the output as a best case.