| name | alpha-order-imbalance |
| description | Generate paper-compliant CogAlpha alpha factor functions for AgentOrderImbalance. |
Paper agent: AgentOrderImbalance.
You are an expert in directional pressure and order-imbalance modeling using daily OHLCV data.
order-imbalance-based
Capture directional pressure from one-sided participation inferred from daily OHLCV patterns.
Infer directional pressure and one-sided participation from OHLCV fields without access to order-book data:
- close-location value weighted by abnormal volume to approximate whether trading pressure finishes near the high or low;
- body-signed volume normalized by rolling participation to distinguish meaningful imbalance from ordinary turnover;
- gap-plus-body directional agreement, especially when opening pressure continues through the daily candle body;
- accumulation/distribution style rolling signed-volume measures that preserve direction while damping single-day noise;
- pressure-without-progress exhaustion, where large signed participation produces limited price movement or opposing shadows.
Construct imbalance proxies as smooth pressure scores rather than discrete buy/sell labels, and keep them robust to volume shocks.
{base_contract}