fix: backtest replay field mapping and logging

- Map DB 'id' field to 'recommendation_id' for evaluate_recommendation()
- Ensure confidence is cast to float (asyncpg may return Decimal)
- Add per-day logging showing rec count, act/skip, positions, pool balance
- Helps diagnose why backtests produce 0 trades

services/trading/backtest_replay.py

@@ -127,6 +127,12 @@ class BacktestReplay:
                     rec["current_price"] = company_prices.get(ticker, 50.0)
                 if "sector" not in rec or not rec.get("sector"):
                     rec["sector"] = company_sectors.get(ticker, "Unknown")
+                # Map 'id' to 'recommendation_id' for evaluate_recommendation()
+                if "recommendation_id" not in rec and "id" in rec:
+                    rec["recommendation_id"] = str(rec["id"])
+                # Ensure confidence is a float
+                if rec.get("confidence") is not None:
+                    rec["confidence"] = float(rec["confidence"])
 
                 recs_by_date.setdefault(d, []).append(rec)
 
@@ -139,6 +145,8 @@ class BacktestReplay:
                     continue
 
                 day_recs = recs_by_date.get(current_date, [])
+                act_count = 0
+                skip_count = 0
 
                 # Process recommendations for this day
                 for rec in day_recs:
@@ -164,6 +172,7 @@ class BacktestReplay:
                     )
 
                     if decision.decision == "act":
+                        act_count += 1
                         ticker = decision.ticker
                         price = rec.get("current_price", 0.0)
                         qty = decision.computed_share_quantity or 0
@@ -183,6 +192,13 @@ class BacktestReplay:
                                 portfolio_state.active_pool -= cost
                                 portfolio_state.open_position_count += 1
 
+                if day_recs:
+                    logger.info(
+                        "Backtest day %s: %d recs, %d act, %d skip, positions=%d, pool=$%.2f",
+                        current_date, len(day_recs), act_count, len(day_recs) - act_count,
+                        len(simulated_positions), portfolio_state.active_pool,
+                    )
+
                 # Simulate simple exit logic: close positions held > 5 days
                 # (simplified — real engine uses stop-loss/take-profit)
                 tickers_to_close = []