Bug: position_size typed float but receives None from kelly sizing — downstream TypeError

[bradsmithmba]

Summary

ScoringEngine._calculate_kelly_sizing explicitly returns None when historical performance data is absent. That None flows into StrategyRecommendation.position_size, which is typed as float. Any downstream code doing arithmetic on recommendation.position_size (e.g., size * capital) raises TypeError.

Evidence

src/models/scoring_engine.py lines 207-208: return None in _calculate_kelly_sizing when history is empty.
src/models/recommendation_engine.py line 329: position_size=scored_strategy.kelly_size
src/models/recommendation_engine.py line 33: position_size: float — not Optional[float].

Fix

Change the type annotation to Optional[float] and document that None means "position size undetermined". Guard all callers that read position_size numerically.