Enhanced position sizing framework based on confidence score.

We will introduce an Adaptive Position Sizing Module that transforms the current basic confidence-based sizing into a sophisticated risk management system. This new feature would calculate position sizes using a multi-tiered approach: first applying a Kelly-criterion inspired formula where

base_size = account_risk * math.pow(conf_score / adaptive_threshold, 1.5)

to create exponential scaling for high-confidence signals, then adjusting for market conditions with

volatility_multiplier = 1.0 / math.sqrt(vol_score)

to automatically reduce exposure during volatile periods, and finally incorporating a regime-based modifier that increases position sizes by 20% during strong trending markets (when

market_trend >= 2

) while reducing them by 30% in choppy conditions. The system would also feature a mean reversion boost that adds 15% to position size when price reaches extreme Bollinger Band deviations (|z-score| > 2.0), recognizing high-probability reversal opportunities. Additionally, the module would implement protective mechanisms including a maximum position size cap at 2x base allocation, automatic scaling down when the indicator detects regime transitions (reducing size by 50% temporarily), and integration with the existing risk-on/off system to completely halt new positions when

is_risk_on

is false. This enhancement would display the calculated position size in the info table with color coding (green for >1.2x, yellow for 0.8-1.2x, gray for <0.8x), making it easy for traders to quickly assess the indicator's conviction level and adjust their actual position sizes accordingly.