malkhut: Flight9 learnings — markout=quality, queue×flow, depth-for-size

Fable's Flight9/BLUE generalizable features incorporated:

1. Slippage model gains trade_flow_intensity parameter:
   - Estimated from book imbalance (proxy for trade arrivals)
   - More flow → better fills (lower slippage)
   - Fable: 'fill = queue position × trade-flow intensity'

2. Markout = quality concept documented:
   - Score fills by post-fill markout, not just fill/no-fill
   - Maker fills are adversely selected

3. Depth-for-size documented:
   - Spread lies; key on depth-within-K-bps vs order notional

4. Measured fees:
   - BingX maker=2.00bp, taker=5.016bp (over 1,455 fills)
   - BingX commission = NEGATIVE (debit)

5. OB study updated with Flight9 learnings

MALKHUT/malkhut/cwm/core.py

@@ -583,9 +583,15 @@ class MinimalCryptoLOBCWM:
             if cumulative_levels > 0:
                 from malkhut.training.slippage_calibration import expected_slippage_bps as _esb
                 total_book_usd = sum(l.price * l.qty for l in (book_depth or ()))
+                bid_vol = sum(l.qty for l in (state.book.bids or ()))
+                ask_vol = sum(l.qty for l in (state.book.asks or ()))
+                total_vol = bid_vol + ask_vol
+                imbalance = abs(bid_vol - ask_vol) / max(total_vol, 1e-12)
+                flow_intensity = min(imbalance * 2.0, 1.0)
                 expected_slippage_bps = _esb(
                     state.venue.symbol, cumulative_levels,
                     new_fill_price * new_fill_qty, total_book_usd,
+                    trade_flow_intensity=flow_intensity,
                 )
         is_maker_fill = (our_action.order_type and our_action.order_type.value == "LIMIT") or our_action.post_only if isinstance(our_action, FulfilmentAction) else False
         price_improvement_bps = 0.0

MALKHUT/malkhut/cwm/hft_cwm.py

@@ -503,12 +503,19 @@ class HftBacktestCWM:
                 if cumulative_usd >= new_fill_price * new_fill_qty:
                     break
             if cumulative_levels > 0:
-                # Calibrated slippage: per-asset model from VST data
+                # Calibrated slippage: per-asset model with flow intensity (Fable Flight9)
                 from malkhut.training.slippage_calibration import expected_slippage_bps as _esb
                 total_book_usd = sum(l.price * l.qty for l in (book_depth or ()))
+                # Estimate flow intensity from book imbalance (proxy for trade arrivals)
+                bid_vol = sum(l.qty for l in (prev_book.bids or ()))
+                ask_vol = sum(l.qty for l in (prev_book.asks or ()))
+                total_vol = bid_vol + ask_vol
+                imbalance = abs(bid_vol - ask_vol) / max(total_vol, 1e-12)
+                flow_intensity = min(imbalance * 2.0, 1.0)  # high imbalance = more flow
                 expected_slippage_bps = _esb(
                     prev_state.venue.symbol, cumulative_levels,
                     new_fill_price * new_fill_qty, total_book_usd,
+                    trade_flow_intensity=flow_intensity,
                 )
 
         # Price improvement: how much better than best bid/ask?

MALKHUT/malkhut/training/slippage_calibration.py

@@ -64,8 +64,18 @@ class SlippageCalibration:
         order_usd: float = 0.0,
         book_depth_usd: float = 1.0,
         is_mainnet: bool = False,
+        trade_flow_intensity: float = 0.0,
     ) -> float:
-        """Predict slippage. Switches model based on book depth."""
+        """Predict slippage. Switches model based on book depth.
+
+        Generalizable features (Fable, Flight9/BLUE):
+          1. Fill = queue position × trade-flow intensity (not just book snapshot)
+          2. Depth-for-size > spread (spread lies — unfillable behind $977)
+          3. Markout = quality (post-fill adverse selection)
+        """
+        # Flow intensity boost: more trade arrivals → higher fill probability
+        flow_boost = 1.0 + trade_flow_intensity * 0.1
+
         if book_depth_usd < self.thin_book_threshold_usd:
             # THIN BOOK: intercept-dominant (alts, meme coins)
             # The fill walks the entire book in 1-2 levels.
@@ -77,6 +87,9 @@ class SlippageCalibration:
             depth_ratio = order_usd / max(book_depth_usd, 1.0)
             base = self.alpha * levels_consumed + self.beta * depth_ratio
 
+        # Adjust for flow intensity: more flow = better fills (lower slippage)
+        base /= max(flow_boost, 0.5)
+
         if is_mainnet:
             base *= self.testnet_to_mainnet
         return base
@@ -123,10 +136,11 @@ class SlippageRegistry:
         order_usd: float = 0.0,
         book_depth_usd: float = 1.0,
         is_mainnet: bool = False,
+        trade_flow_intensity: float = 0.0,
     ) -> float:
         """Predict slippage using the appropriate model."""
         model = self.get(symbol)
-        return model.expected_slippage_bps(levels_consumed, order_usd, book_depth_usd, is_mainnet)
+        return model.expected_slippage_bps(levels_consumed, order_usd, book_depth_usd, is_mainnet, trade_flow_intensity)
 
 
 # ==============================================================================
@@ -219,6 +233,7 @@ def expected_slippage_bps(
     order_usd: float = 0.0,
     book_depth_usd: float = 1.0,
     is_mainnet: bool = False,
+    trade_flow_intensity: float = 0.0,
 ) -> float:
     """Predict slippage using Flight7-calibrated model."""
-    return REGISTRY.expected_slippage_bps(symbol, levels_consumed, order_usd, book_depth_usd, is_mainnet)
+    return REGISTRY.expected_slippage_bps(symbol, levels_consumed, order_usd, book_depth_usd, is_mainnet, trade_flow_intensity)