FMZ Backtesting Modes, Data Granularity, and Performance Metrics
Summary
This guide explains FMZ’s simulated-tick and real-tick backtesting modes. Simulated ticks are generated from candlestick data, while real-tick mode replays recorded tick data and can include depth and trade records. The guide describes the trade-off between speed, data volume, and fidelity, and notes that the matching model assumes price-taking, full fills, so partial executions cannot be tested. It also warns that larger bar intervals can change trade counts and results, recommending finer data granularity when practical.
The remaining material covers supported strategy languages and markets, parameter optimization, saving backtest configurations, and the platform’s reported performance calculations, including annualized return, Sharpe ratio, volatility, drawdown, and win rate. These features help users configure and interpret historical simulations, but they cannot establish future performance. Results depend on the available data, selected interval, fees and slippage settings, and the system’s fill assumptions; real-tick history is also limited by data availability and size. The guide’s metric algorithm and platform-specific behavior should be understood before comparing results across tools.
Key ideas
- Simulated-tick mode derives intrabar ticks from candlestick prices, while real-tick mode replays recorded market data.
- Real-tick backtests provide richer execution data but use more storage and can run more slowly.
- The matching model assumes full fills, so it cannot represent partial executions.
- Backtest results can vary with data granularity, configuration, and historical data availability.
- Parameter optimization and reported risk metrics describe historical simulations, not guaranteed future outcomes.
Tags
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.