Fees and slippage in crypto backtesting

Why small costs can change a result

Gross profit describes a price move before trading costs. Net profit includes the costs represented in the simulation. A strategy that trades frequently can lose much of its apparent advantage when realistic costs are included.

Commission is a charged fee. Slippage is the difference between a reference price and an assumed execution price. Both should be explicit. Actual costs vary with the exchange, market, order type, size, and liquidity.

A simple round-trip calculation

This is hypothetical arithmetic, not a SeekAlgo backtest or an exchange fee quote. Suppose you buy one unit at a reference price of $100 and sell it at $102. Without costs, the profit is $2.

Assume adverse slippage of 0.10% on each side. The buy fill becomes $100.10 and the sell fill becomes $101.898. The difference is $1.798 before fees.

At an illustrative commission of 0.10% of each executed notional, the entry fee is $0.10010 and the exit fee is $0.101898. Net profit is $1.596002, approximately $1.60. This excludes funding, borrowing, and other costs.

The percentages here explain the arithmetic. Do not enter 0.10 into a platform slippage field without verifying its units and execution behavior. Refer to the current settings documentation and inspect the trade list.

Configure and verify the simulation

SeekAlgo exposes commission, commission type, and slippage in its platform settings. Choose the appropriate commission mode and verify the generated script and the applied run settings agree.

Use one small test trade to check how the engine applies those costs. Keep gross price changes, execution adjustments, and commission separate when reconciling the result. Do not charge the same cost twice.

• Use the fee schedule relevant to your own exchange and account tier.
• Record market type, position size, and assumed order execution.
• For derivatives, check whether funding and other relevant charges are represented rather than assuming they are.

Test a range of cost assumptions

Keep the strategy and data fixed while comparing zero-cost, expected-cost, and higher-cost scenarios. Zero cost is a diagnostic baseline, not a realistic forecast. Observe how turnover and position size affect the difference.

If a small increase in costs removes the apparent advantage, document that sensitivity. Do not tune the strategy only until one cost assumption looks attractive. Test on a separate historical period and observe paper behavior too.

Paper trades remain simulated. A cost model cannot guarantee fills during gaps, outages, or low liquidity.