Multi-Day Forex Backtesting with Tick Data and Performance Charts
Summary
The article describes updates to an event-driven forex backtesting system: generating format-compatible simulated tick files, processing daily files sequentially, supporting multiple currency pairs, and plotting equity, returns, and drawdowns. Loading one day per pair at a time limits memory use as the historical test spans more days. The plotted metrics are written to a results file and displayed as three time-series charts.
The example applies a moving-average crossover strategy to randomly generated GBPUSD data. The series uses a random-walk bid and ask with a fixed spread, and the backtest loses money, illustrating that a trend signal on synthetic data can be overwhelmed by transaction costs. This is a software workflow demonstration, not evidence about strategy performance on real markets: the simulated data was designed for format and system testing, not realistic statistical behavior. The article also flags position calculation correctness and richer risk and trade metrics as unfinished areas.
Key ideas
- Daily tick files can be streamed sequentially to support longer backtests with bounded data memory.
- The system supports testing multiple currency pairs across multiple days.
- Simulated ticks use random-walk bid and ask values with a fixed spread for infrastructure testing.
- Equity, returns, and drawdown charts help inspect backtest behavior over time.
- A moving-average crossover loses on the synthetic example, underscoring the impact of spread without establishing real-market performance.
Tags
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.