Gunbot Live backtesting simulates how spot or futures strategy logic would have behaved on historical market data. Use it to compare settings, inspect trades, and review drawdown, fees, and exposure—not to predict or guarantee live results.
Backtests do not guarantee future performance. Treat them as research, not proof.
What This Feature Does
Backtesting lets you research a strategy before exposing live capital. When the strategy and data support it, you can test spot or futures-style behavior, compare parameter changes, and inspect wallet change, drawdown, trades, exposure, fees, and ledger-style results.
What Backtesting Cannot Do
Backtesting is useful research, but it cannot guarantee future profit, fully reproduce exchange latency or outages, know future liquidity and slippage, or replace live monitoring and risk limits.
A Practical Backtesting Workflow
Test more than one historical period, change one parameter at a time, and inspect drawdown, fees, exposure, trade count, and ledger entries before focusing on net profit. Review the result with deep strategy statistics, then compare the assumptions with live execution conditions before using capital.
Trading Reality
It gives Gunbot a browser workspace for monitoring, strategy work, backtesting, risk review, history, and exchange-connected operations. These pages explain what the product does. Your account data stays on your local trading bot, behind login.
Operational Limits
The dashboard depends on the connected Gunbot instance, the exchange API, and the network between them. It can make problems easier to notice, but it does not remove market risk or take responsibility away from you: exchange permissions, strategy settings, open orders, and capital allocation still need your review.
Plain Definition
Backtesting is a historical simulation used to ask: "What would this strategy have done under these assumptions?" For crypto automation, it is a research tool for comparing settings, spotting obvious failure patterns, and understanding trade logic before live exposure.
Spot Backtesting Versus Futures Backtesting
| Area | Spot backtesting | Futures backtesting |
|---|---|---|
| Exposure | Usually asset and quote balance exposure. | Can include contract exposure, leverage, margin, and liquidation assumptions. |
| Costs | Trading fees and spread. | Trading fees, spread, funding, leverage effects, and margin rules. |
| Failure mode | Holding an underwater asset or running out of quote/base balance. | Liquidation, margin exhaustion, funding drag, or leveraged drawdown. |
| Review focus | Entries, exits, fees, position sizing, and inventory. | Entries, exits, leverage, funding, liquidation distance, and risk limits. |
How To Read A Backtest
Start a backtest review with drawdown, not profit. Then check trade count, fee drag, ledger rows, different market regimes, underwater periods, and whether the exposure and capital deployed could be tolerated by a real account.
Overfitting Warning Signs
Overfitting usually shows itself when the result only works on one narrow date range, tiny parameter changes destroy performance, profit depends on ignoring costs, the strategy trades too often for the available edge, or the test uses information the live bot could not have known.
FAQ
Is backtesting the same as live trading?
No. Backtesting uses historical data and assumptions. Live trading uses current exchange conditions and real execution constraints.
Should I optimize until a backtest looks perfect?
No. Overfitting on a specific pair and settings can make a strategy look strong historically while becoming fragile live.
Why can a losing backtest still be useful?
A losing test can reveal fee drag, bad entries, too many trades, poor exit logic, or risk settings that need redesign before live capital is used.
Should futures backtesting be treated more conservatively?
Yes. Futures add leverage, funding, margin, and liquidation mechanics. A visually small move can become a large account event.
What should you do after a good backtest?
Review assumptions, run other periods, compare settings, inspect ledger entries, start small if live, and monitor behavior closely.
Last updated: