seasonal_direction uses historical data to compute the typical
direction (up or down) for the current day-in-cycle and lets you
filter trades to match it.
Shape
{ "type": "seasonal_direction", "direction": "up" }Params
direction-upordown.
Usage
- Asymmetric filters: only take long entries on historically bullish weekdays for the symbol.
- Combine with other filters - seasonality is a weak edge in isolation; useful as one of several inputs.
Pitfalls
- Sample size matters. Seasonal patterns over short histories are usually noise. The engine uses a minimum history threshold; the condition returns false when not enough data is available.
- Regime change can invalidate seasonality. A pattern that held for ten years can break the eleventh.