Why CTA Backtests Begin with a Flat Equity Curve
Summary
A VeighNa community exchange explains why CTA strategies such as moving-average and ATR-RSI examples may show a flat return curve near the start of a backtest. The initial historical observations are used to initialize the strategy’s ArrayManager, so the strategy does not act until enough data has been collected. The delay may therefore exceed the strategy’s own apparent data requirement if the manager is configured with a longer window.
The reply identifies the ArrayManager length as the setting to inspect. It says the default is 100 when no length is supplied and suggests setting it to slightly more than the slow moving-average window, using a five-bar margin as an example. This can reduce unnecessary warm-up time, but only if the strategy’s indicators and logic have enough observations to initialize correctly. The brief forum answer offers no code-level validation or broader discussion of backtest initialization practices.
Key ideas
- CTA strategies may remain inactive while historical observations populate their ArrayManager.
- The ArrayManager length controls how much initialization data is required.
- The reply states that the default length is 100 when no value is passed.
- A suggested setting is the slow moving-average window plus a small buffer, provided all indicators can initialize.
Tags
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.