Fix stats max profit/loss to use win/loss subsets only.

Corrects misleading breakdown rows and summary max amounts after trade log edits.

Co-authored-by: Cursor <cursoragent@cursor.com>

static/js/stats.js

@@ -64,16 +64,26 @@
     }
 
     function cellClass(key, val) {
+        if (key === 'max_loss' || key === 'max_profit') {
+            if (val === 0) return 'text-muted';
+        }
         if (key === 'total_net' || key === 'max_profit' || key === 'avg_profit') {
             if (val > 0) return 'text-profit';
             if (val < 0) return 'text-loss';
         }
         if (key === 'max_loss' || key === 'avg_loss' || key === 'total_fee') {
-            return 'text-loss';
+            if (val < 0) return 'text-loss';
         }
         return '';
     }
 
+    function fmtBreakdownVal(key, val) {
+        if ((key === 'max_loss' || key === 'max_profit') && (val === 0 || val === 0.0)) {
+            return '-';
+        }
+        return fmtNum(val);
+    }
+
     function renderBreakdown(key) {
         if (!cache || !cache.breakdowns) return;
         var block = cache.breakdowns[key];
@@ -110,7 +120,7 @@
                 } else if (col.key === 'label') {
                     td.textContent = val || '-';
                 } else if (typeof val === 'number') {
-                    td.textContent = fmtNum(val);
+                    td.textContent = fmtBreakdownVal(col.key, val);
                     td.className = cellClass(col.key, val);
                 } else {
                     td.textContent = val != null ? val : '-';

stats_engine.py

@@ -102,8 +102,8 @@ def _agg_metrics(label: str, items: list[dict]) -> dict:
     pl_ratio = round(avg_profit / abs(avg_loss), 2) if wins and losses and avg_loss != 0 else 0.0
     total_fee = round(sum(_fee(r) for r in items), 2)
     total_net = round(sum(nets), 2)
-    max_loss = round(min(nets), 2) if nets else 0.0
-    max_profit = round(max(nets), 2) if nets else 0.0
+    max_loss = round(min(losses), 2) if losses else 0.0
+    max_profit = round(max(wins), 2) if wins else 0.0
     win_rate = round(win_cnt / count * 100, 2) if count else 0.0
     return {
         "label": label,
@@ -177,8 +177,8 @@ def compute_summary(trades: list[dict], reviews: list[dict], live_capital: float
     avg_loss = round(sum(losses) / len(losses), 2) if losses else 0.0
     pl_ratio = round(avg_profit / abs(avg_loss), 2) if wins and losses and avg_loss != 0 else 0.0
     total_fee = round(sum(_fee(t) for t in trades) + sum(_fee(r) for r in reviews), 2)
-    max_loss_amt = round(min(nets), 2) if nets else 0.0
-    max_profit_amt = round(max(nets), 2) if nets else 0.0
+    max_loss_amt = round(min(losses), 2) if losses else 0.0
+    max_profit_amt = round(max(wins), 2) if wins else 0.0
 
     margins_loss = [
         _margin_pct(_net_pnl(t), t.get("margin"))