start experiment 80 (exp/80-scheduled-algo-retrain-on-2026-09-22-tac)

finish experiment 79 (exp/79-scheduled-algo-retrain-on-2026-09-21-tac)
exp 79: snapshot custom qlib code
2026-09-23 12:12:46 +00:00 · 2026-09-22 12:50:58 +00:00 · 2026-09-22 12:18:06 +00:00 · 2026-09-22 12:18:00 +00:00 · 2026-09-21 12:02:16 +00:00 · 2026-09-04 12:01:21 +00:00
46 changed files with 2197 additions and 1029 deletions
@@ -1,48 +0,0 @@
-#!/usr/bin/env python3
-"""Precompute the signal-quality gate series and save to pickle.
-
-Usage:
-    python precompute_signal_quality_gate.py <pred_path> <output_path> [topk] [lookback] [threshold]
-
-Example:
-    python precompute_signal_quality_gate.py \
-        /home/data/lake/mlruns/49/34165f27e4a34378ad54843a079a78c0/artifacts/pred.pkl \
-        /app/experiments/book/data/signal_quality_gate/sq_gate_5d_0.50.pkl \
-        10 5 0.5
-"""
-import sys
-import pickle
-from pathlib import Path
-
-# Add tac-qlib to path
-sys.path.insert(0, "/app/tac-qlib")
-
-from tac_qlib.contrib.strategy.signal_quality_gate import compute_signal_quality_gate
-
-if __name__ == "__main__":
-    if len(sys.argv) < 3:
-        print(__doc__)
-        sys.exit(1)
-
-    pred_path = sys.argv[1]
-    output_path = sys.argv[2]
-    topk = int(sys.argv[3]) if len(sys.argv) > 3 else 10
-    lookback = int(sys.argv[4]) if len(sys.argv) > 4 else 5
-    threshold = float(sys.argv[5]) if len(sys.argv) > 5 else 0.5
-
-    lake_root = "/home/data/lake"
-
-    print(f"Computing signal-quality gate: topk={topk}, lookback={lookback}, threshold={threshold}")
-    gate = compute_signal_quality_gate(
-        pred_path,
-        lake_root=lake_root,
-        topk=topk,
-        lookback=lookback,
-        threshold=threshold,
-    )
-
-    print(f"Gate: {gate.sum()}/{len(gate)} days open ({gate.mean():.1%})")
-
-    with open(output_path, "wb") as f:
-        pickle.dump(gate, f)
-    print(f"Saved to {output_path}")
@@ -1,115 +0,0 @@
-# Signal-quality gate backtest for 2021
-{%- set LAKE = TAC_LAKE_DIR %}
-{%- set UNIVERSE = "SPY,QQQ,DIA,IWM,MDY,VTI,VOO,VEA,VWO,VT,EFA,EEM,TLT,IEF,SHY,AGG,BND,LQD,HYG,JNK,EMB,GLD,SLV,USO,UNG,DBA,DBC,XLK,XLF,XLE,XLV,XLI,XLY,XLP,XLU,XLB,XLRE,ARKK,SMH,SOXX,IBB,XBI,ITA,XAR,ICLN,TAN,FDN,IGV,ESPO,REM" %}
-{%- set SP_FIELDS = "sp_ret,sp_ou_zscore,sp_ou_half_life,sp_ou_revert,sp_hmm_p_regime1,sp_hmm_state,sp_jump_ratio,sp_jump_flag,sp_jump_tail,sp_max_move,sp_rv1,sp_rv5,sp_rv22,sp_vol_ratio_5_22,sp_vol_ratio_1_22,sp_trend_slope_5,sp_trend_slope_20,sp_trend_slope_60,sp_logp,sp_hurst_exponent,sp_sig_level1_lead,sp_sig_level1_lag,sp_sig_level2_lead_lag,sp_sig_level2_lag_lead" %}
-{# Gate computed on-the-fly from signal + lake close prices #}
-
-qlib_init:
-    provider_uri: "{{ LAKE }}"
-    region: us
-    expression_cache: null
-    dataset_cache: null
-    calendar_provider:
-        class: tac_qlib.data.providers.LakeCalendarProvider
-        kwargs: { lake_root: "{{ LAKE }}", market: US }
-    instrument_provider:
-        class: tac_qlib.data.providers.LakeInstrumentProvider
-        kwargs: { lake_root: "{{ LAKE }}", market: US, markets: {} }
-    feature_provider:
-        class: tac_qlib.data.providers.LakeFeatureProvider
-        kwargs: { lake_root: "{{ LAKE }}", market: US }
-    exp_manager:
-        class: MLflowExpManager
-        module_path: qlib.workflow.expm
-        kwargs:
-            uri: "sqlite:///{{ LAKE }}/mlruns.db"
-            default_exp_name: "tac-rd-sq-gate-2021"
-
-task:
-    model:
-        class: LGBModel
-        module_path: qlib.contrib.model.gbdt
-        kwargs:
-            loss: mse
-            learning_rate: 0.05
-            num_leaves: 15
-            n_estimators: 200
-            colsample_bytree: 0.8
-            subsample: 0.8
-            subsample_freq: 1
-            reg_alpha: 0.01
-            reg_lambda: 0.01
-
-    dataset:
-        class: DatasetH
-        module_path: qlib.data.dataset
-        kwargs:
-            handler:
-                class: TACHandler
-                module_path: tac_qlib.contrib.data.handler
-                kwargs:
-                    instruments: "{{ UNIVERSE }}"
-                    start_time: 2015-01-03
-                    end_time: 2021-12-31
-                    fit_start_time: 2015-01-03
-                    fit_end_time: 2021-01-03
-                    freq: day
-                    lake_root: "{{ LAKE }}"
-                    market: US
-                    label: "Ref($close,-6)/Ref($close,-1)-1"
-                    feature_fields: "$open,$high,$low,$close,$vwap,$volume,{{ SP_FIELDS }}"
-                    infer_processors:
-                        - class: DropAllNaN
-                          kwargs: {}
-                        - class: ProcessInf
-                          kwargs: {}
-                        - class: CSRankNorm
-                          kwargs: {}
-                        - class: ZScoreNorm
-                          kwargs: {}
-                        - class: Fillna
-                          kwargs: {}
-            segments:
-                train: [2015-01-03, 2020-09-01]
-                valid: [2020-09-03, 2021-01-03]
-                test: [2021-01-04, 2021-12-31]
-
-    record:
-        - class: SignalRecord
-          module_path: qlib.workflow.record_temp
-          kwargs: {}
-        - class: SigAnaRecord
-          module_path: qlib.workflow.record_temp
-          kwargs: { ana_long_short: true, ann_scaler: 252 }
-        - class: PortAnaRecord
-          module_path: qlib.workflow.record_temp
-          kwargs:
-              config:
-                  strategy:
-                      class: SignalQualityGateStrategy
-                      module_path: tac_qlib.contrib.strategy.signal_quality_gate
-                      kwargs:
-                          signal: "<PRED>"
-                          lake_root: "{{ LAKE }}"
-                          gate_topk: 10
-                          gate_lookback: 5
-                          gate_threshold: 0.5
-                          gate_start: "2015-01-03"
-                          gate_end: "2021-12-31"
-                          topk: 10
-                          n_drop: 1
-                          only_tradable: true
-                          risk_degree: 0.95
-                  backtest:
-                      start_time: 2021-01-04
-                      end_time: 2021-12-31
-                      account: 1000000
-                      benchmark: SPY
-                      exchange_kwargs:
-                          codes: "{{ UNIVERSE }}"
-                          deal_price: $close
-                          freq: day
-                          open_cost: 0.0005
-                          close_cost: 0.0015
-                          min_cost: 5.0
-              risk_analysis_freq: 1d
@@ -1,115 +0,0 @@
-# Signal-quality gate backtest for 2024
-{%- set LAKE = TAC_LAKE_DIR %}
-{%- set UNIVERSE = "SPY,QQQ,DIA,IWM,MDY,VTI,VOO,VEA,VWO,VT,EFA,EEM,TLT,IEF,SHY,AGG,BND,LQD,HYG,JNK,EMB,GLD,SLV,USO,UNG,DBA,DBC,XLK,XLF,XLE,XLV,XLI,XLY,XLP,XLU,XLB,XLRE,ARKK,SMH,SOXX,IBB,XBI,ITA,XAR,ICLN,TAN,FDN,IGV,ESPO,REM" %}
-{%- set SP_FIELDS = "sp_ret,sp_ou_zscore,sp_ou_half_life,sp_ou_revert,sp_hmm_p_regime1,sp_hmm_state,sp_jump_ratio,sp_jump_flag,sp_jump_tail,sp_max_move,sp_rv1,sp_rv5,sp_rv22,sp_vol_ratio_5_22,sp_vol_ratio_1_22,sp_trend_slope_5,sp_trend_slope_20,sp_trend_slope_60,sp_logp,sp_hurst_exponent,sp_sig_level1_lead,sp_sig_level1_lag,sp_sig_level2_lead_lag,sp_sig_level2_lag_lead" %}
-{# Gate computed on-the-fly from signal + lake close prices #}
-
-qlib_init:
-    provider_uri: "{{ LAKE }}"
-    region: us
-    expression_cache: null
-    dataset_cache: null
-    calendar_provider:
-        class: tac_qlib.data.providers.LakeCalendarProvider
-        kwargs: { lake_root: "{{ LAKE }}", market: US }
-    instrument_provider:
-        class: tac_qlib.data.providers.LakeInstrumentProvider
-        kwargs: { lake_root: "{{ LAKE }}", market: US, markets: {} }
-    feature_provider:
-        class: tac_qlib.data.providers.LakeFeatureProvider
-        kwargs: { lake_root: "{{ LAKE }}", market: US }
-    exp_manager:
-        class: MLflowExpManager
-        module_path: qlib.workflow.expm
-        kwargs:
-            uri: "sqlite:///{{ LAKE }}/mlruns.db"
-            default_exp_name: "tac-rd-sq-gate-2024"
-
-task:
-    model:
-        class: LGBModel
-        module_path: qlib.contrib.model.gbdt
-        kwargs:
-            loss: mse
-            learning_rate: 0.05
-            num_leaves: 15
-            n_estimators: 200
-            colsample_bytree: 0.8
-            subsample: 0.8
-            subsample_freq: 1
-            reg_alpha: 0.01
-            reg_lambda: 0.01
-
-    dataset:
-        class: DatasetH
-        module_path: qlib.data.dataset
-        kwargs:
-            handler:
-                class: TACHandler
-                module_path: tac_qlib.contrib.data.handler
-                kwargs:
-                    instruments: "{{ UNIVERSE }}"
-                    start_time: 2015-01-03
-                    end_time: 2024-12-31
-                    fit_start_time: 2015-01-03
-                    fit_end_time: 2024-01-03
-                    freq: day
-                    lake_root: "{{ LAKE }}"
-                    market: US
-                    label: "Ref($close,-6)/Ref($close,-1)-1"
-                    feature_fields: "$open,$high,$low,$close,$vwap,$volume,{{ SP_FIELDS }}"
-                    infer_processors:
-                        - class: DropAllNaN
-                          kwargs: {}
-                        - class: ProcessInf
-                          kwargs: {}
-                        - class: CSRankNorm
-                          kwargs: {}
-                        - class: ZScoreNorm
-                          kwargs: {}
-                        - class: Fillna
-                          kwargs: {}
-            segments:
-                train: [2015-01-03, 2023-09-01]
-                valid: [2023-09-03, 2024-01-03]
-                test: [2024-01-02, 2024-12-31]
-
-    record:
-        - class: SignalRecord
-          module_path: qlib.workflow.record_temp
-          kwargs: {}
-        - class: SigAnaRecord
-          module_path: qlib.workflow.record_temp
-          kwargs: { ana_long_short: true, ann_scaler: 252 }
-        - class: PortAnaRecord
-          module_path: qlib.workflow.record_temp
-          kwargs:
-              config:
-                  strategy:
-                      class: SignalQualityGateStrategy
-                      module_path: tac_qlib.contrib.strategy.signal_quality_gate
-                      kwargs:
-                          signal: "<PRED>"
-                          lake_root: "{{ LAKE }}"
-                          gate_topk: 10
-                          gate_lookback: 5
-                          gate_threshold: 0.5
-                          gate_start: "2015-01-03"
-                          gate_end: "2024-12-31"
-                          topk: 10
-                          n_drop: 1
-                          only_tradable: true
-                          risk_degree: 0.95
-                  backtest:
-                      start_time: 2024-01-02
-                      end_time: 2024-12-31
-                      account: 1000000
-                      benchmark: SPY
-                      exchange_kwargs:
-                          codes: "{{ UNIVERSE }}"
-                          deal_price: $close
-                          freq: day
-                          open_cost: 0.0005
-                          close_cost: 0.0015
-                          min_cost: 5.0
-              risk_analysis_freq: 1d
@@ -1,35 +1,34 @@
 # TradeAC custom-qlib-code snapshot (auto-generated)
-# parent repo HEAD : e952feed0a66a20439f4f24ad5524233429cd0c3
+# parent repo HEAD : a37fc0869f99a0eebc34c8eb98af0514c6d7fba8
 #   tac-qlib/tac_qlib/contrib
 #   tac-qlib/tac_qlib/data
 # per-file hashes (git hash-object):
  1b6298c4a5652f2e863cbdc385a1014a570fcd59  tac-qlib/tac_qlib/contrib/__init__.py
-  861592c63edd6a0853a9cb174b5970435b135fc8  tac-qlib/tac_qlib/contrib/__pycache__/__init__.cpython-312.pyc
+  d1a8ec0c9e6c38839e966f687b08ec412f87ec20  tac-qlib/tac_qlib/contrib/__pycache__/__init__.cpython-312.pyc
+  2224424d0ff193be4f55d1b791f8fce89439c5d2  tac-qlib/tac_qlib/contrib/backtest/__init__.py
+  0bf40dee440ddbded357d7bbb4efc67c62c4b084  tac-qlib/tac_qlib/contrib/backtest/tradeac_exchange.py
  c76a9f17f680e74eea766eff27f7624359749ed6  tac-qlib/tac_qlib/contrib/data/__init__.py
-  5c547a2ef92e075e550fe6d01508a2f1d3f536bc  tac-qlib/tac_qlib/contrib/data/__pycache__/__init__.cpython-312.pyc
-  4f656130d167e79dcaaeb7783a121f0b36852374  tac-qlib/tac_qlib/contrib/data/__pycache__/handler.cpython-312.pyc
-  0dd25ef161c6e0f15eafc84886e7e1381deb38c3  tac-qlib/tac_qlib/contrib/data/handler.py
+  c1543fabfc22e07c7e9942015c4462f84acc1e91  tac-qlib/tac_qlib/contrib/data/__pycache__/__init__.cpython-312.pyc
+  89f5cfc15b946b6ea1fb7724f021024967cdb6c5  tac-qlib/tac_qlib/contrib/data/__pycache__/handler.cpython-312.pyc
+  3bba0f1696e4ab4b3deebec3f31f269b2e713899  tac-qlib/tac_qlib/contrib/data/handler.py
  b151d139a0dcde87d74b21e7c4b729176ba5c39b  tac-qlib/tac_qlib/contrib/model/__init__.py
-  b1489f2fc0dee85f0a4f90b2e6ad545ed9c8967b  tac-qlib/tac_qlib/contrib/model/__pycache__/__init__.cpython-312.pyc
-  121ef237da1df1b8e21a561c3ad0db200b901339  tac-qlib/tac_qlib/contrib/model/__pycache__/rank_ensemble.cpython-312.pyc
-  74d0da348cbcc3700c96b6f4fe4391488e61efc5  tac-qlib/tac_qlib/contrib/model/__pycache__/rank_gbdt.cpython-312.pyc
+  f4e9bf0a78d22e256cade0d794a6958aaa7c6721  tac-qlib/tac_qlib/contrib/model/__pycache__/__init__.cpython-312.pyc
+  c35c166a8449c14fa3cd4972c363d3d9d58620ef  tac-qlib/tac_qlib/contrib/model/__pycache__/rank_ensemble.cpython-312.pyc
+  98be60fe19fedb3d9cb21c3f91ef0e176bd93287  tac-qlib/tac_qlib/contrib/model/__pycache__/rank_gbdt.cpython-312.pyc
  d3f051f3a8650c42fedc7b367b966f7c74fb5789  tac-qlib/tac_qlib/contrib/model/rank_ensemble.py
  d03e6611338918d4aac5eea4adf26f85a3763652  tac-qlib/tac_qlib/contrib/model/rank_gbdt.py
-  2c2f167b693f4366a769998e3c9d4804f29e31e0  tac-qlib/tac_qlib/contrib/strategy/__init__.py
-  f9cd9ab729e3248542ccc490af7adc2e51c71914  tac-qlib/tac_qlib/contrib/strategy/__pycache__/__init__.cpython-312.pyc
-  cd3133cfbd2556b25c106e39ae97fb128df0e326  tac-qlib/tac_qlib/contrib/strategy/__pycache__/ic_gate.cpython-312.pyc
-  6dd1c568a2961842793674390d5abffd1a0e71b8  tac-qlib/tac_qlib/contrib/strategy/__pycache__/optimal_stop.cpython-312.pyc
-  03b5e4d80da00800f1b108bee0735d3d18d856d1  tac-qlib/tac_qlib/contrib/strategy/__pycache__/regime_gate.cpython-312.pyc
-  a6a1c21ab71b62080830df45c4784b73c1531036  tac-qlib/tac_qlib/contrib/strategy/__pycache__/signal_quality_gate.cpython-312.pyc
-  755e3b139496a5e22b0328db45c8199c33029fbc  tac-qlib/tac_qlib/contrib/strategy/__pycache__/weekly_rebalance.cpython-312.pyc
-  519a1f4c05dbe0ac018ab8b779eb33d53b4dd545  tac-qlib/tac_qlib/contrib/strategy/ic_gate.py
+  184f80da8edf944bad3c8fb4d4d3d189bf4f082b  tac-qlib/tac_qlib/contrib/strategy/__init__.py
+  69684c9c9624cb16c63f3a18a864c49a7db37fa9  tac-qlib/tac_qlib/contrib/strategy/__pycache__/__init__.cpython-312.pyc
+  4d3a5433e57e65e5dc96c6d2999f9d473b696dc4  tac-qlib/tac_qlib/contrib/strategy/__pycache__/long_short.cpython-312.pyc
+  53e72f512b47207a5292a918efef7fee162deef3  tac-qlib/tac_qlib/contrib/strategy/__pycache__/optimal_stop.cpython-312.pyc
+  896ef74ae47bcd1ed388e1e5d9c8d70c28097fe9  tac-qlib/tac_qlib/contrib/strategy/kelly_dropout.py
+  9090fc6dfbd339f2f4df4b0c9b87f400ecb5c9d5  tac-qlib/tac_qlib/contrib/strategy/long_short.py
  79aaad9e39fcc740a773f4f63c512ce1086cfde0  tac-qlib/tac_qlib/contrib/strategy/optimal_stop.py
-  7bcee5f0b09cfa721440f1354f16f2dd9a112b12  tac-qlib/tac_qlib/contrib/strategy/regime_gate.py
-  16ab80b731aab6d8bb818525615d77c2d1fcb0c8  tac-qlib/tac_qlib/contrib/strategy/signal_quality_gate.py
-  fe60bacdfedd48617863be31f24b7c7daebfac5a  tac-qlib/tac_qlib/contrib/strategy/weekly_rebalance.py
+  5b9acfb4340111b204249add7760bd53c6ae03f1  tac-qlib/tac_qlib/contrib/strategy/regime_gate.py
+  839abb89ad40cd516eabcfd91fe1be626b9f091f  tac-qlib/tac_qlib/contrib/strategy/weekly_rebalance.py
  92e6e90eb0cd0a25142034560f27adb6b705b1a8  tac-qlib/tac_qlib/data/__init__.py
-  316bf4aa160cc8d15929ea648be03f4b4999667d  tac-qlib/tac_qlib/data/__pycache__/__init__.cpython-312.pyc
-  554a3f29d181b64effbf49a8161b32e7f93d8d3e  tac-qlib/tac_qlib/data/__pycache__/config.cpython-312.pyc
-  8b47f6d78ac046b6b7b2fb07bd7f3382773ffb73  tac-qlib/tac_qlib/data/__pycache__/providers.cpython-312.pyc
-  53c9007a928841fd3c3b08450f9a6520ce1ac091  tac-qlib/tac_qlib/data/config.py
+  0f1eb6f41d44bd9064b1f16f529dac6bb9d7c3fd  tac-qlib/tac_qlib/data/__pycache__/__init__.cpython-312.pyc
+  cf99dd8a481f6e2d1f323f8d8d754730eadb357a  tac-qlib/tac_qlib/data/__pycache__/config.cpython-312.pyc
+  ddac5bab6578399c0301390d71bae13511a56cb1  tac-qlib/tac_qlib/data/__pycache__/providers.cpython-312.pyc
+  1953fb2a6371525db7f7b0e1c9dfbf3492d82110  tac-qlib/tac_qlib/data/config.py
  8d0644f6f0d1efb94798ed444cc73e63b643459b  tac-qlib/tac_qlib/data/providers.py
@@ -0,0 +1 @@
+from .tradeac_exchange import TradeACExchange
@@ -0,0 +1,432 @@
+# Copyright (c) Microsoft Corporation.
+# Licensed under the MIT License.
+"""
+TradeACExchange
+
+A short/borrow enabled Exchange implementation built on top of qlib.backtest.exchange.Exchange.
+This exchange adds simple, configurable margin logic (initial/maintenance), borrowing support for
+shorts, a borrow fee, and a lightweight SMA (Special Memorandum Account) concept to emulate
+behaviors similar to brokers such as IBKR and Alpaca for backtesting purposes.
+
+Notes / limitations
+- This implementation is intentionally lightweight and conservative: it implements the
+  key behaviors needed for strategy/backtest experiments (allowing short selling, computing
+  margin requirements, performing margin-call checks, and tracking SMA-like excess equity).
+- It makes some simplifying assumptions compared to real brokers (no per-product house margins,
+  simplified SMA bookkeeping, borrow availability modeled only by a per-symbol boolean/limit).
+- The Position class in qlib.backtest.position was not changed. To support shorts we update the
+  position.position dict directly when necessary. This keeps integration simple but bypasses some
+  internal Position helpers. Use with care.
+
+API additions
+- allow_short: enable short selling (bool)
+- initial_margin_long/short: fraction required to open a position
+- maintenance_margin_long/short: fraction required to keep a position
+- borrow_fee_rate: periodic borrow fee applied on short value (applied at trade time as additional cost)
+- borrowable: dict mapping stock_id -> bool or float (max borrowable shares). Symbols missing from
+  the dict follow `borrow_default` (default True = unlimited; set False for a strict whitelist)
+- get_sma(position): returns SMA-like excess equity available as "buying power credit"
+- check_margin_call(position): returns True if position is below maintenance requirement
+
+"""
+from __future__ import annotations
+
+from typing import Any, Dict, Optional, Tuple
+
+import numpy as np
+
+from qlib.backtest.decision import Order
+from qlib.backtest.exchange import Exchange
+from qlib.backtest.position import BasePosition
+
+
+class TradeACExchange(Exchange):
+    """An exchange that supports short selling / borrowing and basic margin rules.
+
+    The implementation aims to be compatible with the Exchange API used by Account and
+    Position classes in qlib.backtest. It overrides only the minimum methods required to
+    enable short/borrow behavior and margin calculations.
+    """
+
+    def __init__(
+        self,
+        *args: Any,
+        allow_short: bool = True,
+        initial_margin_long: float = 0.5,
+        initial_margin_short: float = 0.5,
+        maintenance_margin_long: float = 0.25,
+        maintenance_margin_short: float = 0.3,
+        borrow_fee_rate: float = 0.0,
+        borrowable: Optional[Dict[str, float]] = None,
+        borrow_default: bool = True,
+        sma_enabled: bool = True,
+        **kwargs: Any,
+    ) -> None:
+        """Create TradeACExchange.
+
+        Parameters mirror Exchange with additional tradeac-specific options.
+        """
+        super().__init__(*args, **kwargs)
+        self.allow_short = allow_short
+        self.initial_margin_long = initial_margin_long
+        self.initial_margin_short = initial_margin_short
+        self.maintenance_margin_long = maintenance_margin_long
+        self.maintenance_margin_short = maintenance_margin_short
+        self.borrow_fee_rate = borrow_fee_rate
+        # borrowable can be a dict with per-symbol max borrowable amount, or None (unlimited)
+        self.borrowable = borrowable or {}
+        # borrow_default: policy for symbols absent from `borrowable`.
+        #   True  -> unlisted symbols are unlimited-borrowable (legacy behavior)
+        #   False -> unlisted symbols are NOT borrowable; only listed ones can be shorted
+        self.borrow_default = bool(borrow_default)
+        # sma_enabled: whether to expose lightweight SMA calculation
+        self.sma_enabled = sma_enabled
+
+    # --------------------------- Helper calculations ---------------------------
+    def _initial_margin_requirement(self, position: BasePosition) -> float:
+        """Compute the initial margin requirement (money) for the given position.
+
+        We treat longs and shorts separately and sum their required initial margins.
+        """
+        im_req = 0.0
+        for sid in position.get_stock_list():
+            amt = position.get_stock_amount(sid)
+            price = position.get_stock_price(sid)
+            val = amt * price
+            if val > 0:
+                im_req += abs(val) * self.initial_margin_long
+            elif val < 0:
+                im_req += abs(val) * self.initial_margin_short
+        return im_req
+
+    def _maintenance_margin_requirement(self, position: BasePosition) -> float:
+        """Compute the maintenance margin requirement (money) for the given position."""
+        mm_req = 0.0
+        for sid in position.get_stock_list():
+            amt = position.get_stock_amount(sid)
+            price = position.get_stock_price(sid)
+            val = amt * price
+            if val > 0:
+                mm_req += abs(val) * self.maintenance_margin_long
+            elif val < 0:
+                mm_req += abs(val) * self.maintenance_margin_short
+        return mm_req
+
+    def get_equity(self, position: BasePosition) -> float:
+        """Return account equity (position value + cash)."""
+        return position.calculate_value()
+
+    def get_sma(self, position: BasePosition) -> float:
+        """Return a simplified SMA: excess equity above initial margin requirement.
+
+        Note: This is a synthetic/Simplified SMA used for strategy/backtest logic. Real-broker
+        SMA accounting (e.g. credits/debits across days) can be more complex.
+        """
+        if not self.sma_enabled:
+            return 0.0
+        equity = self.get_equity(position)
+        im_req = self._initial_margin_requirement(position)
+        return max(0.0, equity - im_req)
+
+    def check_margin_call(self, position: BasePosition) -> bool:
+        """Return True when the account is under maintenance margin (margin call).
+
+        Margin call condition here is simple: equity < maintenance requirement.
+        """
+        equity = self.get_equity(position)
+        mm_req = self._maintenance_margin_requirement(position)
+        return equity < mm_req
+
+    def get_buying_power(self, position: BasePosition) -> float:
+        """Estimate buying power for new long positions assuming opening margin requirement.
+
+        Simplified: the maximum notional long value = equity / initial_margin_long.
+        """
+        equity = self.get_equity(position)
+        if self.initial_margin_long <= 0:
+            return 0.0
+        return equity / self.initial_margin_long
+
+    # --------------------------- Order / execution overrides ---------------------------
+    def _borrow_headroom(self, stock_id: str, current_short: float) -> float:
+        """Remaining borrowable shares for `stock_id` given an already-open short of `current_short` shares.
+
+        borrowable values: bool (True=unlimited, False=not borrowable) or numeric max shares.
+        Missing symbols follow `borrow_default` (True = unlimited when allow_short is enabled).
+        """
+        if not self.allow_short:
+            return 0.0
+        v = self.borrowable.get(stock_id, self.borrow_default)
+        if isinstance(v, bool):
+            return float("inf") if v else 0.0
+        try:
+            limit = float(v)
+        except (TypeError, ValueError):
+            return float("inf")
+        return max(0.0, limit - max(current_short, 0.0))
+
+    def _calc_trade_info_by_order(
+        self,
+        order: Order,
+        position: Optional[BasePosition],
+        dealt_order_amount: Dict[str, float],
+    ) -> Tuple[float, float, float]:
+        """Override to allow (optionally) short selling and to apply borrow fees.
+
+        The original Exchange implementation forbids selling more than you own. Here we allow
+        sell orders to create/expand short positions when allow_short is True. We still rely on
+        most base logic (price discovery, impact, cost calculation) by calling super(), but we
+        adjust the sell-side clipping behavior before delegating to the base implementation.
+        """
+        # When selling and shorts are allowed, temporarily relax the clipping logic in the base
+        # implementation by monkey-patching current position check. Simpler: replicate minimal
+        # parts of logic from Exchange._calc_trade_info_by_order with the key change.
+
+        # Get basic trade price & volume info using Exchange helpers
+        trade_price = float(self.get_deal_price(order.stock_id, order.start_time, order.end_time, direction=order.direction))
+        total_trade_val = float(self.get_volume(order.stock_id, order.start_time, order.end_time) or 0.0) * trade_price
+
+        order.factor = self.get_factor(order.stock_id, order.start_time, order.end_time)
+        order.deal_amount = order.amount  # attempt full
+
+        # volume clipping (same as base)
+        self._clip_amount_by_volume(order, dealt_order_amount)
+
+        # approximate adjusted cost ratio based on liquidity
+        if not total_trade_val or np.isnan(total_trade_val) or total_trade_val <= 0:
+            adj_cost_ratio = self.impact_cost
+        else:
+            trade_val_tmp = order.deal_amount * trade_price
+            adj_cost_ratio = self.impact_cost * (trade_val_tmp / total_trade_val) ** 2
+
+        # Differentiate buy / sell
+        if order.direction == Order.SELL:
+            cost_ratio = self.close_cost + adj_cost_ratio
+            current_amount = (
+                position.get_stock_amount(order.stock_id) if (position is not None and position.check_stock(order.stock_id)) else 0.0
+            )
+            long_held = max(current_amount, 0.0)
+            short_open = max(-current_amount, 0.0)
+
+            if position is not None:
+                if not self.allow_short:
+                    # clip by current holdings only
+                    if not np.isclose(order.deal_amount, current_amount):
+                        order.deal_amount = self.round_amount_by_trade_unit(
+                            min(long_held, order.deal_amount), order.factor
+                        )
+                else:
+                    # allow selling beyond holdings up to the remaining borrow limit;
+                    # later when updating the position we create/expand a short if necessary.
+                    max_sell = long_held + self._borrow_headroom(order.stock_id, short_open)
+                    if order.deal_amount > max_sell and not np.isclose(order.deal_amount, max_sell):
+                        order.deal_amount = self.round_amount_by_trade_unit(max_sell, order.factor)
+
+        elif order.direction == Order.BUY:
+            cost_ratio = self.open_cost + adj_cost_ratio
+            if position is not None:
+                cash = position.get_cash()
+                trade_val = order.deal_amount * trade_price
+                if cash < max(trade_val * cost_ratio, self.min_cost):
+                    order.deal_amount = 0
+                    self.logger.debug(f"Order clipped due to cost higher than cash: {order}")
+                elif cash < trade_val + max(trade_val * cost_ratio, self.min_cost):
+                    max_buy_amount = self._get_buy_amount_by_cash_limit(trade_price, cash, cost_ratio)
+                    order.deal_amount = self.round_amount_by_trade_unit(min(max_buy_amount, order.deal_amount), order.factor)
+                    self.logger.debug(f"Order clipped due to cash limitation: {order}")
+                else:
+                    order.deal_amount = self.round_amount_by_trade_unit(order.deal_amount, order.factor)
+            else:
+                order.deal_amount = self.round_amount_by_trade_unit(order.deal_amount, order.factor)
+        else:
+            raise NotImplementedError("order direction {} error".format(order.direction))
+
+        # compute final trade_val & trade_cost
+        trade_val = order.deal_amount * trade_price
+        # base trade_cost
+        trade_cost = max(trade_val * cost_ratio, self.min_cost)
+        # apply borrow fee only on the net-new short portion of the sell
+        if order.direction == Order.SELL and self.allow_short:
+            new_short = max(0.0, order.deal_amount - long_held)
+            trade_cost += new_short * trade_price * self.borrow_fee_rate
+
+        if trade_val <= 1e-5:
+            trade_cost = 0
+
+        return trade_price, trade_val, trade_cost
+
+    def deal_order(
+        self,
+        order: Order,
+        trade_account: Optional[Any] = None,
+        position: Optional[BasePosition] = None,
+        dealt_order_amount: Dict[str, float] = None,
+    ) -> Tuple[float, float, float]:
+        """Deal order and handle short position bookkeeping.
+
+        This method mirrors Exchange.deal_order but when a position is provided and shorts are
+        allowed it will update the Position.position dict directly to support negative amounts.
+        """
+        if dealt_order_amount is None:
+            dealt_order_amount = {}
+
+        if not self.check_order(order):
+            order.deal_amount = 0.0
+            self.logger.debug(f"Order failed due to trading limitation: {order}")
+            return 0.0, 0.0, np.nan
+
+        if trade_account is not None and position is not None:
+            raise ValueError("trade_account and position can only choose one")
+
+        pos = position or (trade_account.current_position if trade_account is not None else None)
+        trade_price, trade_val, trade_cost = self._calc_trade_info_by_order(order, pos, dealt_order_amount)
+
+        if trade_val > 1e-5:
+            if trade_account is not None:
+                cp = trade_account.current_position
+                if not cp.skip_update():
+                    held = cp.check_stock(order.stock_id)
+                    # Account-level bookkeeping (turnover/cost/returns). Mirrors
+                    # Account._update_state_from_order except for fresh short sales,
+                    # where no prior price exists to compute order profit from.
+                    if order.direction == Order.SELL and not held:
+                        trade_account.accum_info.add_turnover(trade_val)
+                        trade_account.accum_info.add_cost(trade_cost)
+                        trade_account.accum_info.add_return_value(0.0)
+                    if order.direction == Order.SELL:
+                        # sell: update account state first (stock entry may be deleted)
+                        if held:
+                            trade_account._update_state_from_order(order, trade_val, trade_cost, trade_price)
+                        self._position_sell(cp, order, trade_val, trade_cost, trade_price)
+                    else:
+                        # buy: update position first (entry may be created), then account state
+                        # A buy that covers a short to exactly flat deletes the entry inside
+                        # _position_buy; re-seed a transient zero-amount stub so the
+                        # account's order-profit lookup still finds the trade price,
+                        # then drop it (_update_state_from_order never mutates entries).
+                        sid = order.stock_id
+                        had_entry = isinstance(cp.position.get(sid), dict)
+                        self._position_buy(cp, order, trade_val, trade_cost, trade_price)
+                        covered_to_flat = had_entry and not isinstance(cp.position.get(sid), dict)
+                        if covered_to_flat:
+                            cp.position[sid] = {"amount": 0.0, "price": trade_price, "weight": 0}
+                        trade_account._update_state_from_order(order, trade_val, trade_cost, trade_price)
+                        if covered_to_flat:
+                            cp.position.pop(sid, None)
+            elif position is not None:
+                if order.direction == Order.BUY:
+                    self._position_buy(position, order, trade_val, trade_cost, trade_price)
+                else:
+                    self._position_sell(position, order, trade_val, trade_cost, trade_price)
+        return trade_val, trade_cost, trade_price
+
+    # --------------------------- Position mutation helpers ---------------------------
+    def _position_buy(self, position: BasePosition, order: Order, trade_val: float, cost: float, trade_price: float) -> None:
+        """Handle buy order bookkeeping against a BasePosition while supporting shorts.
+
+        Rules implemented (simplified):
+        - If there is an existing short (amount < 0), the buy will first cover the short.
+        - If covering closes the short completely, the remaining buy becomes a long
+        - Cash updates mimic Position._buy_stock/_sell_stock (cash decreases by trade_val+cost for buys)
+        """
+        trade_amount = trade_val / trade_price
+        sid = order.stock_id
+        current_amount = position.get_stock_amount(sid) if position.check_stock(sid) else 0.0
+
+        # covering existing short
+        if current_amount < -1e-12:
+            # amount is negative -> we are short. Buying reduces the short.
+            new_amount = current_amount + trade_amount
+            if abs(new_amount) <= 1e-8:
+                # short fully covered exactly -> remove entry
+                if sid in position.position:
+                    del position.position[sid]
+            elif new_amount > 0:
+                # short fully covered with leftover buy amount -> leftover becomes a long position
+                position.position[sid] = {"amount": new_amount, "price": trade_price, "weight": 0}
+            else:
+                # partially cover
+                position.position[sid]["amount"] = new_amount
+                position.position[sid]["price"] = trade_price
+        else:
+            # normal or increasing long
+            if sid not in position.position or not isinstance(position.position[sid], dict):
+                # initialize stock
+                position.position[sid] = {"amount": trade_amount, "price": trade_price, "weight": 0}
+            else:
+                position.position[sid]["amount"] = position.position[sid].get("amount", 0.0) + trade_amount
+                position.position[sid]["price"] = trade_price
+
+        # cash effect same as Position._buy_stock
+        position.position["cash"] -= trade_val + cost
+
+    def _position_sell(self, position: BasePosition, order: Order, trade_val: float, cost: float, trade_price: float) -> None:
+        """Handle sell order bookkeeping against a BasePosition while supporting shorts.
+
+        Rules implemented (simplified):
+        - If holding enough long shares, sell will reduce/close the long position normally.
+        - If not holding enough long shares and shorts are allowed, the remaining sold amount will create/expand a short position.
+        - Cash update for sells follows Position._sell_stock logic (cash increases by trade_val - cost)
+        """
+        trade_amount = trade_val / trade_price
+        sid = order.stock_id
+        current_amount = position.get_stock_amount(sid) if position.check_stock(sid) else 0.0
+
+        if current_amount > 1e-12:
+            # we have long shares; sell from them first
+            if trade_amount >= current_amount - 1e-8:
+                # selling all or more than holdings
+                # remove long position
+                if sid in position.position:
+                    del position.position[sid]
+                # remaining sold amount becomes short if allowed
+                remain = trade_amount - current_amount
+                if remain > 1e-8:
+                    if not self.allow_short:
+                        # should not happen due to clipping earlier, but guard anyway
+                        raise ValueError(f"Attempt to short {sid} while shorting disabled")
+                    # create short entry
+                    position.position[sid] = {"amount": -remain, "price": trade_price, "weight": 0}
+            else:
+                # partial sell
+                position.position[sid]["amount"] = current_amount - trade_amount
+                position.position[sid]["price"] = trade_price
+        else:
+            # currently flat or already short
+            if not self.allow_short:
+                raise ValueError(f"Attempt to short {sid} while shorting disabled")
+            # expand short
+            new_amount = current_amount - trade_amount
+            if sid not in position.position or not isinstance(position.position[sid], dict):
+                position.position[sid] = {"amount": new_amount, "price": trade_price, "weight": 0}
+            else:
+                position.position[sid]["amount"] = new_amount
+                position.position[sid]["price"] = trade_price
+
+        # cash effect same as Position._sell_stock
+        new_cash = trade_val - cost
+        if getattr(position, "_settle_type", None) == position.ST_CASH:
+            position.position["cash_delay"] = position.position.get("cash_delay", 0.0) + new_cash
+        else:
+            position.position["cash"] = position.position.get("cash", 0.0) + new_cash
+
+    # --------------------------- Borrow availability helpers ---------------------------
+    def is_borrowable(self, stock_id: str, amount: float) -> bool:
+        """Check whether the requested amount is borrowable for the given stock.
+
+        If a borrowable dict is provided, it may contain either booleans or numeric limits (maximum borrowable shares).
+        Symbols absent from the dict follow `borrow_default`.
+        """
+        if not self.allow_short:
+            return False
+        if stock_id not in self.borrowable:
+            return self.borrow_default
+        v = self.borrowable[stock_id]
+        if isinstance(v, bool):
+            return v
+        try:
+            limit = float(v)
+            return amount <= limit
+        except Exception:
+            return True
+
@@ -15,6 +15,9 @@ import os
 from inspect import getfullargspec
 from typing import List, Optional, Tuple, Union

+import numpy as np
+import pandas as pd
+
 from qlib.data.dataset import processor as processor_module
 from qlib.data.dataset.handler import DataHandlerLP
 from qlib.utils import get_callable_kwargs
@@ -22,6 +25,7 @@ from qlib.utils import get_callable_kwargs
 from ...data.config import (
    LakeConfig,
    timeframe_for_freq,
+    FEATURE_FAMILIES,
    NON_FEATURE_COLUMNS,
 )

@@ -92,7 +96,7 @@ def get_common_feature_fields(lake_root=None, market="US", timeframe="1d") -> Li

    common: set = set()
    # family tier: features/market=*/timeframe=*/family=*/symbol=*.parquet
-    for fam in ("ta", "sp"):
+    for fam in FEATURE_FAMILIES:
        fam_dir = feat_dir / f"family={fam}"
        if fam_dir.is_dir():
            common |= _family_common(fam_dir)
@@ -143,6 +147,174 @@ class DropAllNaN(processor_module.Processor):
        return df


+class BenchResidual(processor_module.Processor):
+    """Subtract a benchmark instrument's forward return from the label, per datetime.
+
+    Turns the training target from an absolute-return rank into a *residual* rank:
+    ``r_i - r_bench`` is ranked cross-sectionally by the downstream ``CSRankNorm`` /
+    ``CSZScoreNorm`` processors instead of ``r_i`` alone. Must be inserted BEFORE any
+    per-date normalization so the ranking itself is computed on residual returns
+    (ordering flips exactly where the benchmark trends).
+
+    Stateless: ``fit`` is a no-op and the benchmark forward return is recomputed from
+    the lake parquet on first ``__call__``. Rows whose benchmark value is missing are
+    left untouched. Accepts ``fit_start_time``/``fit_end_time`` (ignored) so
+    ``check_transform_proc`` can inject the fit window uniformly.
+
+    NOTE: under any cross-sectional normalization downstream (``CSRankNorm`` /
+    ``CSZScoreNorm``) this processor is a mathematical no-op: subtracting the same
+    per-date constant preserves ranks, and z-scoring absorbs constant shifts. Use
+    ``BenchBetaResidual`` for a target that actually reorders.
+    """
+
+    def __init__(
+        self,
+        benchmark="SPY",
+        fields_group="label",
+        lake_root=None,
+        market="US",
+        timeframe=None,
+        freq="day",
+        fit_start_time=None,
+        fit_end_time=None,
+    ):
+        self.benchmark = benchmark
+        self.fields_group = fields_group
+        self.lake_root = lake_root
+        self.market = market
+        self.timeframe = timeframe or timeframe_for_freq(freq)
+        self.fit_start_time = fit_start_time
+        self.fit_end_time = fit_end_time
+        self._bench_label = None
+
+    def _load_bench_label(self):
+        if self._bench_label is not None:
+            return self._bench_label
+        cfg = LakeConfig(self.lake_root, self.market)
+        p = cfg.bar_path(self.timeframe, self.benchmark)
+        if not p.exists():
+            raise FileNotFoundError(f"BenchResidual: benchmark bar file not found: {p}")
+        df = pd.read_parquet(p)
+        s = pd.Series(df["c"].astype(float).values, index=pd.to_datetime(df["t"])).sort_index()
+        s.index = s.index.normalize()
+        # mirror Ref($close,-6)/Ref($close,-1)-1 on the benchmark's own calendar
+        bench_label = s.shift(-6) / s.shift(-1) - 1
+        self._bench_label = bench_label[~bench_label.index.duplicated(keep="last")]
+        return self._bench_label
+
+    def fit(self, df=None):
+        return self
+
+    def __call__(self, df):
+        bl = self._load_bench_label()
+        cols = processor_module.get_group_columns(df, self.fields_group)
+        dt = df.index.get_level_values("datetime")
+        aligned = bl.reindex(pd.DatetimeIndex(dt.unique())).reindex(dt)
+        mask = aligned.notna().values
+        out = df.copy()
+        for c in cols:
+            vals = df[c].values
+            res = vals.copy()
+            res[mask] = np.asarray(vals[mask], dtype=float) - aligned[mask].values
+            out[c] = res
+        return out
+
+
+class BenchBetaResidual(processor_module.Processor):
+    """Residualize the label against a beta-scaled benchmark move: ``r_i - b_i * r_bench``.
+
+    Unlike a plain constant subtraction (see ``BenchResidual``), the name-specific rolling
+    beta ``b_i`` makes this survive cross-sectional normalization: in up-weeks high-beta
+    names lose rank, in down-weeks they gain — exactly the relative structure an absolute-
+    return ranking hides.
+
+    Beta is estimated from *past* data only (rolling ``window`` trading days of daily close
+    returns of each instrument vs the benchmark, both read up to and including ``t``), so
+    no lookahead enters the target. The benchmark leg uses the same horizon as the label
+    expression (``Ref($close,-6)/Ref($close,-1)-1`` by default via ``horizon``/``base``,
+    matching the yaml's 6-day label). Rows with missing beta or benchmark values keep
+    their raw label.
+
+    Requires ``$close`` to be present in the feature group (it always is for TACHandler).
+    Stateless; accepts ``fit_start_time``/``fit_end_time`` (ignored) for uniform kwargs
+    injection. Must be inserted BEFORE any per-date normalization processor.
+    """
+
+    def __init__(
+        self,
+        benchmark="SPY",
+        fields_group="label",
+        lake_root=None,
+        market="US",
+        timeframe=None,
+        freq="day",
+        window=63,
+        horizon=6,
+        base=1,
+        feature_field="$close",
+        fit_start_time=None,
+        fit_end_time=None,
+    ):
+        self.benchmark = benchmark
+        self.fields_group = fields_group
+        self.lake_root = lake_root
+        self.market = market
+        self.timeframe = timeframe or timeframe_for_freq(freq)
+        self.window = int(window)
+        self.horizon = int(horizon)
+        self.base = int(base)
+        self.feature_field = feature_field
+        self.fit_start_time = fit_start_time
+        self.fit_end_time = fit_end_time
+        self._bench = None
+
+    def _load_bench_close(self):
+        if self._bench is not None:
+            return self._bench
+        cfg = LakeConfig(self.lake_root, self.market)
+        p = cfg.bar_path(self.timeframe, self.benchmark)
+        if not p.exists():
+            raise FileNotFoundError(f"BenchBetaResidual: benchmark bar file not found: {p}")
+        df = pd.read_parquet(p)
+        s = pd.Series(df["c"].astype(float).values, index=pd.to_datetime(df["t"])).sort_index()
+        s.index = s.index.normalize()
+        self._bench = s[~s.index.duplicated(keep="last")]
+        return self._bench
+
+    def fit(self, df=None):
+        return self
+
+    def __call__(self, df):
+        bench = self._load_bench_close()
+
+        # benchmark forward return over the same horizon as the label expression
+        fwd = bench.shift(-(self.base + self.horizon - 1)) / bench.shift(-self.base) - 1
+
+        px_col = ("feature", self.feature_field)
+        if px_col not in df.columns:
+            raise KeyError(f"BenchBetaResidual: {self.feature_field} not found in features")
+        px = df[px_col].unstack("instrument").sort_index()
+        rets = px / px.shift(1) - 1
+        bret = bench.reindex(px.index).pct_change()
+
+        # rolling beta per instrument using data <= t (no lookahead)
+        cov = rets.rolling(self.window, min_periods=max(10, self.window // 2)).cov(bret)
+        var = bret.rolling(self.window, min_periods=max(10, self.window // 2)).var()
+        beta = cov.div(var, axis=0)
+
+        contrib = beta.mul(fwd.reindex(px.index), axis=0)
+        cols = list(processor_module.get_group_columns(df, self.fields_group))
+        out = df.copy()
+        for c in cols:
+            lab = df[c].unstack("instrument").reindex(px.index)
+            resid = lab - contrib.where(contrib.notna() & lab.notna(), 0.0)
+            new_vals = resid.stack()
+            new_vals.index.names = df.index.names
+            # residual where available, raw label otherwise (e.g. beta warm-up rows)
+            out[c] = new_vals.reindex(out.index).fillna(df[c])
+        return out
+
+
 class TACHandler(DataHandlerLP):
    """DataHandlerLP backed by the TradeAC parquet lake.

@@ -245,10 +417,12 @@ class TACHandler(DataHandlerLP):
        return get_common_feature_fields(lake_root, market, timeframe_for_freq(freq))


-__all__ = ["TACHandler", "DropAllNaN", "get_common_feature_fields"]
+__all__ = ["TACHandler", "DropAllNaN", "BenchResidual", "BenchBetaResidual", "get_common_feature_fields"]


-# Make `DropAllNaN` resolvable by bare name from processor configs (e.g. the default
-# ``infer_processors`` and workflow yamls that reference it without a ``module_path``),
-# mirroring how qlib registers its own processors in ``qlib.data.dataset.processor``.
+# Make `DropAllNaN`/`BenchResidual`/`BenchBetaResidual` resolvable by bare name from processor
+# configs (e.g. the default ``infer_processors`` and workflow yamls that reference them without a
+# ``module_path``), mirroring how qlib registers its own processors in ``qlib.data.dataset.processor``.
 processor_module.DropAllNaN = DropAllNaN
+processor_module.BenchResidual = BenchResidual
+processor_module.BenchBetaResidual = BenchBetaResidual
@@ -1,11 +1,4 @@
-from .ic_gate import ICGateTopkDropoutStrategy  # noqa: F401
 from .optimal_stop import OptimalStopControl  # noqa: F401
-from .regime_gate import RegimeGateTopkDropoutStrategy  # noqa: F401
-from .weekly_rebalance import WeeklyRebalanceDropoutStrategy  # noqa: F401
+from .long_short import LongShortTopkStrategy  # noqa: F401

-__all__ = [
-    "ICGateTopkDropoutStrategy",
-    "OptimalStopControl",
-    "RegimeGateTopkDropoutStrategy",
-    "WeeklyRebalanceDropoutStrategy",
-]
+__all__ = ["OptimalStopControl", "LongShortTopkStrategy"]
@@ -1,117 +0,0 @@
-"""Realized-IC circuit breaker TopkDropout strategy.
-
-Subclass of ``qlib.contrib.strategy.signal_strategy.TopkDropoutStrategy`` that
-holds the book (issues NO orders) while the streaming realized RankIC of the
-deployed signal is below threshold — i.e. the model's cross-sectional
-predictions are no longer earning against realized forward returns. When the
-gate is open it behaves exactly like the reference TopkDropoutStrategy.
-
-The gate is evaluated per trade step on the trailing mean realized RankIC of
-the signal over the last ``ic_window`` trading days whose label is fully
-realized as of the decision date (no lookahead — a 5d fwd label ``close[t+6]/
-close[t+1]-1`` is only known at ``t+6``).
-
-Two wiring modes:
-
-* ``ic_gate``: a precomputed ``pd.Series`` indexed by datetime of booleans
-  (True = gate open / trade allowed). Computed once by the caller (e.g.
-  ``rd_backtest``) and looked up per step. Missing dates default to open.
-* realized-IC self-computation: when ``ic_min_rankic`` is given but no
-  ``ic_gate``, the strategy computes the per-date realized RankIC itself from
-  ``self.signal`` (the pred scores) and the lake 1d bars via
-  ``tac_qlib.risk_limits.realized_rankic_series``, then applies the same
-  trailing-window comparison. Works when instantiated from a workflow YAML
-  PortAnaRecord config (``lake_root`` / ``market`` must be provided).
-"""
-
-from __future__ import annotations
-
-import pandas as pd
-
-from qlib.backtest.decision import TradeDecisionWO
-from qlib.contrib.strategy.signal_strategy import TopkDropoutStrategy
-
-from tac_qlib.risk_limits import ic_circuit_breaker, realized_rankic_series
-
-__all__ = ["ICGateTopkDropoutStrategy"]
-
-
-class ICGateTopkDropoutStrategy(TopkDropoutStrategy):
-    """TopkDropout with a streaming realized-IC circuit breaker.
-
-    Parameters
-    ----------
-    topk, n_drop, method_sell, method_buy, hold_thresh, only_tradable,
-    forbid_all_trade_at_limit : same as ``TopkDropoutStrategy``.
-    ic_min_rankic : float — pause new trading while trailing realized RankIC is
-        below this threshold (0 disables the gate).
-    ic_window : int — trailing window for the realized RankIC mean (default 22).
-    ic_label_horizon : int — label horizon in trading days (default 6).
-    ic_min_obs : int — min realized labels before the gate arms (default 10).
-    ic_gate : pd.Series, optional — precomputed per-date gate (bool indexed by
-        datetime). When provided, it overrides self-computation.
-    lake_root, market : str — lake location for self-computed realized IC.
-    """
-
-    def __init__(
-        self,
-        *,
-        topk,
-        n_drop,
-        ic_min_rankic: float = 0.0,
-        ic_window: int = 22,
-        ic_label_horizon: int = 6,
-        ic_min_obs: int = 10,
-        ic_gate=None,
-        lake_root: str = "",
-        market: str = "US",
-        **kwargs,
-    ):
-        super().__init__(topk=topk, n_drop=n_drop, **kwargs)
-        self.ic_min_rankic = float(ic_min_rankic or 0.0)
-        self.ic_window = int(ic_window or 22)
-        self.ic_label_horizon = int(ic_label_horizon or 6)
-        self.ic_min_obs = int(ic_min_obs or 10)
-        self._ic_gate = ic_gate
-        self._realized_ic = None
-        self.lake_root = lake_root or ""
-        self.market = market or "US"
-
-    def _load_realized_ic(self):
-        if self._realized_ic is None:
-            pred_start_time, pred_end_time = self.trade_calendar.get_step_time(
-                self.trade_calendar.get_trade_step(), shift=-self.ic_label_horizon
-            )
-            pred = self.signal.get_signal(start_time=pred_start_time, end_time=pred_end_time)
-            if isinstance(pred, pd.DataFrame):
-                pred = pred.iloc[:, 0]
-            self._realized_ic = realized_rankic_series(
-                pred, self.lake_root, self.market, label_horizon=self.ic_label_horizon
-            )
-        return self._realized_ic
-
-    def _gate_open(self, trade_start_time) -> bool:
-        ts = pd.Timestamp(trade_start_time)
-        if self._ic_gate is not None:
-            # precomputed gate series: look up the latest known decision date <= ts
-            known = self._ic_gate[self._ic_gate.index <= ts]
-            if len(known):
-                return bool(known.iloc[-1])
-            return True
-        if self.ic_min_rankic <= 0:
-            return True
-        realized = self._load_realized_ic()
-        limits = {
-            "ic_min_rankic": self.ic_min_rankic,
-            "ic_window": self.ic_window,
-            "ic_min_obs": self.ic_min_obs,
-        }
-        tripped, _reason, _trail = ic_circuit_breaker(realized, ts, limits)
-        return not tripped
-
-    def generate_trade_decision(self, execute_result=None):
-        trade_step = self.trade_calendar.get_trade_step()
-        trade_start_time, _ = self.trade_calendar.get_step_time(trade_step)
-        if not self._gate_open(trade_start_time):
-            return TradeDecisionWO([], self)
-        return super().generate_trade_decision(execute_result)
@@ -0,0 +1,201 @@
+"""Fractional-Kelly dropout strategy for cross-sectional signals.
+
+Sizing rule variant of ``qlib.contrib.strategy.signal_strategy.TopkDropoutStrategy``:
+the topk/n_drop SELECTION is identical to the reference, but the buy size is
+proportional to the score MAGNITUDE (edge) instead of equal-weight, capped at a
+fraction ``cap_frac`` of the equal-weight notional so a single name cannot
+over-concentrate the book.
+
+``cap_frac`` is the fraction of the equal-weight per-name notional that a top
+signal can deploy at most (e.g. 0.5 = at most half the equal-weight size).
+Names whose score is below the median of the buy set get a proportionally
+smaller slice; the residual stays in cash (that is the point of the rule:
+throw away less edge per name, deploy less capital when conviction is low).
+"""
+
+from __future__ import annotations
+
+from typing import List
+
+import numpy as np
+import pandas as pd
+
+from qlib.backtest import Order
+from qlib.backtest.decision import OrderDir, TradeDecisionWO
+from qlib.contrib.strategy.signal_strategy import TopkDropoutStrategy
+
+__all__ = ["FractionalKellyDropoutStrategy"]
+
+DEFAULT_CAP_FRAC = 0.5
+
+
+class FractionalKellyDropoutStrategy(TopkDropoutStrategy):
+    """TopkDropout selection with score-magnitude (fractional-Kelly) sizing.
+
+    Parameters
+    ----------
+    topk, n_drop, method_sell, method_buy, hold_thresh, only_tradable,
+    forbid_all_trade_at_limit : same as ``TopkDropoutStrategy``.
+    cap_frac : max buy notional as a fraction of the equal-weight notional.
+    """
+
+    def __init__(self, *, topk, n_drop, cap_frac: float = DEFAULT_CAP_FRAC, **kwargs):
+        super().__init__(topk=topk, n_drop=n_drop, **kwargs)
+        self.cap_frac = cap_frac
+
+    def generate_trade_decision(self, execute_result=None):
+        import copy
+
+        trade_step = self.trade_calendar.get_trade_step()
+        trade_start_time, trade_end_time = self.trade_calendar.get_step_time(trade_step)
+        pred_start_time, pred_end_time = self.trade_calendar.get_step_time(trade_step, shift=1)
+        pred_score = self.signal.get_signal(start_time=pred_start_time, end_time=pred_end_time)
+        if isinstance(pred_score, pd.DataFrame):
+            pred_score = pred_score.iloc[:, 0]
+        if pred_score is None:
+            return TradeDecisionWO([], self)
+
+        if self.only_tradable:
+
+            def get_first_n(li, n, reverse=False):
+                cur_n = 0
+                res = []
+                for si in reversed(li) if reverse else li:
+                    if self.trade_exchange.is_stock_tradable(
+                        stock_id=si, start_time=trade_start_time, end_time=trade_end_time
+                    ):
+                        res.append(si)
+                        cur_n += 1
+                        if cur_n >= n:
+                            break
+                return res[::-1] if reverse else res
+
+            def get_last_n(li, n):
+                return get_first_n(li, n, reverse=True)
+
+            def filter_stock(li):
+                return [
+                    si
+                    for si in li
+                    if self.trade_exchange.is_stock_tradable(
+                        stock_id=si, start_time=trade_start_time, end_time=trade_end_time
+                    )
+                ]
+
+        else:
+
+            def get_first_n(li, n):
+                return list(li)[:n]
+
+            def get_last_n(li, n):
+                return list(li)[-n:]
+
+            def filter_stock(li):
+                return li
+
+        current_temp: "object" = copy.deepcopy(self.trade_position)
+        sell_order_list: List[Order] = []
+        buy_order_list: List[Order] = []
+        cash = current_temp.get_cash()
+        current_stock_list = current_temp.get_stock_list()
+        last = pred_score.reindex(current_stock_list).sort_values(ascending=False).index
+
+        if self.method_buy == "top":
+            today = get_first_n(
+                pred_score[~pred_score.index.isin(last)].sort_values(ascending=False).index,
+                self.n_drop + self.topk - len(last),
+            )
+        elif self.method_buy == "random":
+            topk_candi = get_first_n(pred_score.sort_values(ascending=False).index, self.topk)
+            candi = list(filter(lambda x: x not in last, topk_candi))
+            n = self.n_drop + self.topk - len(last)
+            try:
+                today = np.random.choice(candi, n, replace=False)
+            except ValueError:
+                today = candi
+        else:
+            raise NotImplementedError(f"This type of input is not supported")
+
+        comb = pred_score.reindex(last.union(pd.Index(today))).sort_values(ascending=False).index
+
+        if self.method_sell == "bottom":
+            sell = last[last.isin(get_last_n(comb, self.n_drop))]
+        elif self.method_sell == "random":
+            candi = filter_stock(last)
+            try:
+                sell = pd.Index(np.random.choice(candi, self.n_drop, replace=False) if len(last) else [])
+            except ValueError:
+                sell = candi
+        else:
+            raise NotImplementedError(f"This type of input is not supported")
+
+        buy = today[: len(sell) + self.topk - len(last)]
+        for code in current_stock_list:
+            if not self.trade_exchange.is_stock_tradable(
+                stock_id=code,
+                start_time=trade_start_time,
+                end_time=trade_end_time,
+                direction=None if self.forbid_all_trade_at_limit else OrderDir.SELL,
+            ):
+                continue
+            if code in sell:
+                time_per_step = self.trade_calendar.get_freq()
+                if current_temp.get_stock_count(code, bar=time_per_step) < self.hold_thresh:
+                    continue
+                sell_amount = current_temp.get_stock_amount(code=code)
+                sell_order = Order(
+                    stock_id=code,
+                    amount=sell_amount,
+                    start_time=trade_start_time,
+                    end_time=trade_end_time,
+                    direction=Order.SELL,
+                )
+                if self.trade_exchange.check_order(sell_order):
+                    sell_order_list.append(sell_order)
+                    trade_val, trade_cost, trade_price = self.trade_exchange.deal_order(
+                        sell_order, position=current_temp
+                    )
+                    cash += trade_val - trade_cost
+
+        if len(buy) == 0:
+            return TradeDecisionWO(sell_order_list, self)
+
+        # ---- fractional-Kelly sizing --------------------------------------
+        # equal-weight notional (reference baseline)
+        eq_notional = cash * self.risk_degree / len(buy)
+        buy_scores = pred_score.reindex(buy).astype(float)
+        lo, hi = buy_scores.min(), buy_scores.max()
+        if hi == lo:
+            w = pd.Series(1.0, index=buy_scores.index)
+        else:
+            w = (buy_scores - lo) / (hi - lo)  # [0,1] edge magnitude
+            w = w.clip(lower=0.0)
+            w_max = w.max()
+            w = w / w_max if w_max > 0 else w  # max == 1.0
+        for code in buy:
+            if not self.trade_exchange.is_stock_tradable(
+                stock_id=code,
+                start_time=trade_start_time,
+                end_time=trade_end_time,
+                direction=None if self.forbid_all_trade_at_limit else OrderDir.BUY,
+            ):
+                continue
+            buy_price = self.trade_exchange.get_deal_price(
+                stock_id=code, start_time=trade_start_time, end_time=trade_end_time, direction=OrderDir.BUY
+            )
+            notional = eq_notional * min(self.cap_frac, float(w.get(code, 0.0)))
+            buy_amount = notional / buy_price
+            factor = self.trade_exchange.get_factor(
+                stock_id=code, start_time=trade_start_time, end_time=trade_end_time
+            )
+            buy_amount = self.trade_exchange.round_amount_by_trade_unit(buy_amount, factor)
+            buy_order = Order(
+                stock_id=code,
+                amount=buy_amount,
+                start_time=trade_start_time,
+                end_time=trade_end_time,
+                direction=Order.BUY,
+            )
+            buy_order_list.append(buy_order)
+
+        return TradeDecisionWO(sell_order_list + buy_order_list, self)
@@ -0,0 +1,361 @@
+"""Long-short Top-K strategy for cross-sectional signals.
+
+Each day the strategy ranks the cross-section by prediction score and rebalances
+to an equal-weight two-sided book: the ``topk`` highest-ranked names go long and
+the ``k_short`` lowest-ranked names go short. Net-new shorts are opened by
+selling beyond current holdings, which requires a short-aware exchange such as
+``tac_qlib.contrib.backtest.tradeac_exchange.TradeACExchange`` with
+``allow_short=True`` (borrow limits, margin requirements and borrow fees are
+enforced there, not here).
+
+Sizing deploys ``equity * risk_degree`` as gross notional split evenly across
+all long and short legs, so the book is approximately market neutral.
+``allow_short=False`` disables the short side entirely (long-only ``topk``).
+
+Short eligibility can be restricted further, with static or dynamic gates:
+``short_whitelist`` limits shorts to an explicit symbol set; ``short_vol_top_pct``
+requires a candidate's trailing realized volatility to rank in the top fraction
+of that day's cross-section; ``short_max_mom`` (falling-knife filter) only
+allows shorting names whose own trailing momentum is at/below a threshold;
+``short_regime_sma`` disables shorts entirely while the benchmark trades above
+its moving average (risk-on). Borrow availability itself is enforced by the
+exchange (``borrowable`` whitelist / per-symbol caps via ``TradeACExchange``).
+
+Wired into qrun workflows like any ``BaseStrategy`` (see ``PortAnaRecord``
+config). Mirrors the API usage of qlib's ``TopkDropoutStrategy``: ``Order`` /
+``OrderDir`` from ``qlib.backtest.decision``, ``trade_calendar`` /
+``trade_exchange`` / ``trade_position`` injected by the backtest executor.
+"""
+
+from __future__ import annotations
+
+import copy
+from typing import Dict, List, Optional
+
+import pandas as pd
+
+from qlib.backtest import Order
+from qlib.backtest.decision import OrderDir, TradeDecisionWO
+from qlib.contrib.strategy.signal_strategy import BaseSignalStrategy
+from qlib.log import get_module_logger
+
+__all__ = ["LongShortTopkStrategy"]
+
+
+class LongShortTopkStrategy(BaseSignalStrategy):
+    """Equal-weight long-short Top-K strategy over a cross-sectional signal.
+
+    Parameters
+    ----------
+    topk : number of long legs (highest-ranked names).
+    k_short : number of short legs (lowest-ranked names).
+    hold_thresh : minimum holding days before a leg may be closed/reduced.
+    only_tradable : only select candidates tradable on the trade date.
+    rebalance_tol : skip rebalances smaller than this fraction of a leg's
+        target notional (turnover control).
+    allow_short : enable/disable the short side. With ``False`` the bottom-ranked
+        legs are dropped and the book is long-only ``topk``; pair with
+        ``allow_short=False`` on the exchange for a fully borrow-free run.
+        Legacy alias ``enable_short`` is accepted.
+    short_whitelist : optional list of symbols eligible for shorting; candidates
+        outside the list are skipped (``None`` = all names eligible).
+    short_vol_window : trailing window (trading days) for realized-vol estimation.
+    short_vol_top_pct : if set, a short candidate's trailing realized volatility
+        must rank at or above this percentile of that day's cross-section
+        (e.g. ``0.5`` = only the more volatile half may be shorted). Candidates
+        without measurable vol are never shorted.
+    short_mom_window : trailing window (trading days) for the candidate momentum
+        used by the falling-knife gate.
+    short_max_mom : if set, a candidate's trailing ``short_mom_window``-day return
+        must be <= this value to be shortable (e.g. ``0.0`` = only short names
+        that are actually falling). Candidates without measurable momentum are
+        never shorted.
+    short_regime_symbol : benchmark symbol for the regime gate (default SPY).
+    short_regime_sma : if set, shorts are only allowed on days where the regime
+        symbol's last close (strictly before the execution bar) is BELOW its
+        ``short_regime_sma``-day moving average — i.e. shorts are disabled in
+        risk-on regimes and enabled in drawdowns.
+    """
+
+    def __init__(
+        self,
+        *,
+        signal=None,
+        topk: int = 4,
+        k_short: int = 2,
+        hold_thresh: int = 1,
+        only_tradable: bool = True,
+        rebalance_tol: float = 0.05,
+        allow_short: Optional[bool] = None,
+        enable_short: Optional[bool] = None,
+        short_whitelist: Optional[List[str]] = None,
+        short_vol_window: int = 20,
+        short_vol_top_pct: Optional[float] = None,
+        short_mom_window: int = 20,
+        short_max_mom: Optional[float] = None,
+        short_regime_symbol: str = "SPY",
+        short_regime_sma: Optional[int] = None,
+        risk_degree: float = 0.95,
+        trade_exchange=None,
+        level_infra=None,
+        common_infra=None,
+        **kwargs,
+    ):
+        super().__init__(
+            signal=signal,
+            risk_degree=risk_degree,
+            trade_exchange=trade_exchange,
+            level_infra=level_infra,
+            common_infra=common_infra,
+            **kwargs,
+        )
+        if allow_short is None:
+            allow_short = True if enable_short is None else bool(enable_short)
+        self.allow_short = bool(allow_short)
+        self.topk = topk
+        self.k_short = k_short
+        self.hold_thresh = hold_thresh
+        self.only_tradable = only_tradable
+        self.rebalance_tol = rebalance_tol
+        self.short_whitelist = set(short_whitelist) if short_whitelist is not None else None
+        if not 0 < float(short_vol_window) <= 1000:
+            raise ValueError(f"short_vol_window must be in (0, 1000], got {short_vol_window}")
+        self.short_vol_window = int(short_vol_window)
+        if short_vol_top_pct is not None and not 0.0 < float(short_vol_top_pct) <= 1.0:
+            raise ValueError(f"short_vol_top_pct must be in (0, 1], got {short_vol_top_pct}")
+        self.short_vol_top_pct = None if short_vol_top_pct is None else float(short_vol_top_pct)
+        if not 0 < float(short_mom_window) <= 1000:
+            raise ValueError(f"short_mom_window must be in (0, 1000], got {short_mom_window}")
+        self.short_mom_window = int(short_mom_window)
+        self.short_max_mom = None if short_max_mom is None else float(short_max_mom)
+        self.short_regime_symbol = str(short_regime_symbol)
+        if short_regime_sma is not None and not 1 < int(short_regime_sma) <= 1000:
+            raise ValueError(f"short_regime_sma must be in (1, 1000], got {short_regime_sma}")
+        self.short_regime_sma = None if short_regime_sma is None else int(short_regime_sma)
+        # per-day caches (keyed by trade date)
+        self._vol_cache_key: Optional[str] = None
+        self._vol_cache_val: Dict[str, Dict[str, float]] = {}
+        self._regime_cache: Dict[str, bool] = {}
+
+    # ------------------------------------------------------------------ utils
+    def _is_tradable(self, code, start, end) -> bool:
+        if not self.only_tradable:
+            return True
+        try:
+            return self.trade_exchange.is_stock_tradable(stock_id=code, start_time=start, end_time=end)
+        except TypeError:
+            return True
+
+    def _mark_price(self, code, start, end) -> Optional[float]:
+        try:
+            px = self.trade_exchange.get_deal_price(
+                stock_id=code, start_time=start, end_time=end, direction=OrderDir.BUY
+            )
+        except (KeyError, ValueError):
+            return None
+        if px is None or px != px or px <= 0:
+            return None
+        return float(px)
+
+    def _day_stats(self, codes: List[str], trade_start) -> Dict[str, Dict[str, float]]:
+        """Per-day cross-sectional stats used by the dynamic short gates.
+
+        For each code, returns ``{"vol_rank": r}`` (percentile of trailing
+        realized vol over ``short_vol_window`` bars across that day's
+        cross-section) when the vol gate is on, and ``{"mom": m}`` (trailing
+        ``short_mom_window``-bar return) when the falling-knife gate is on.
+        All series end on the last bar strictly BEFORE the execution bar (no
+        lookahead). Codes without measurable data are simply absent — such
+        candidates are never shorted (fail-closed).
+        """
+        if self.short_vol_top_pct is None and self.short_max_mom is None:
+            return {}
+        key = str(pd.Timestamp(trade_start))
+        if self._vol_cache_key == key:
+            return self._vol_cache_val
+        out: Dict[str, Dict[str, float]] = {}
+        try:
+            from qlib.data import D
+
+            end = pd.Timestamp(trade_start)
+            buf = max(self.short_vol_window, self.short_mom_window) * 3 + 30
+            df = D.features(
+                sorted(codes),
+                ["$close"],
+                start_time=end - pd.Timedelta(days=buf),
+                end_time=end - pd.Timedelta(days=1),
+            )
+            close = df["$close"].unstack(level="instrument") if isinstance(df.index, pd.MultiIndex) else df["$close"]
+            if self.short_vol_top_pct is not None:
+                vol = close.pct_change().rolling(self.short_vol_window).std().iloc[-1]
+                for code, rank in vol.rank(pct=True).dropna().items():
+                    out.setdefault(str(code), {})["vol_rank"] = float(rank)
+            if self.short_max_mom is not None:
+                w = min(self.short_mom_window, len(close) - 1)
+                mom = close.iloc[-1] / close.iloc[-(w + 1)] - 1
+                for code, m in mom.items():
+                    if m == m:
+                        out.setdefault(str(code), {})["mom"] = float(m)
+        except Exception as e:  # noqa: BLE001 - degrade to fail-closed (no shorts)
+            get_module_logger(self.__class__.__name__).warning(
+                f"short gates unavailable ({type(e).__name__}: {e}); no shorts this step"
+            )
+        self._vol_cache_key, self._vol_cache_val = key, out
+        return out
+
+    def _regime_ok(self, trade_start) -> bool:
+        """True when shorting is allowed by the benchmark-regime gate.
+
+        With ``short_regime_sma`` set, shorts are permitted only while the
+        regime symbol's last close strictly before the execution bar sits below
+        its moving average (risk-off). Data failure fails closed (no shorts).
+        """
+        if self.short_regime_sma is None:
+            return True
+        key = str(pd.Timestamp(trade_start))
+        cached = self._regime_cache.get(key)
+        if cached is not None:
+            return cached
+        ok = False
+        try:
+            from qlib.data import D
+
+            end = pd.Timestamp(trade_start)
+            df = D.features(
+                [self.short_regime_symbol],
+                ["$close"],
+                start_time=end - pd.Timedelta(days=int(self.short_regime_sma * 3 + 30)),
+                end_time=end - pd.Timedelta(days=1),
+            )
+            s = df["$close"]
+            if isinstance(s.index, pd.MultiIndex):
+                s = s.droplevel("instrument")
+            sma = s.rolling(self.short_regime_sma).mean().iloc[-1]
+            px = s.iloc[-1]
+            ok = bool(px < sma)
+        except Exception as e:  # noqa: BLE001 - degrade to fail-closed (no shorts)
+            get_module_logger(self.__class__.__name__).warning(
+                f"regime gate unavailable ({type(e).__name__}: {e}); no shorts this step"
+            )
+        self._regime_cache[key] = ok
+        return ok
+
+    # ------------------------------------------------------------- decision
+    def generate_trade_decision(self, execute_result=None):
+        trade_step = self.trade_calendar.get_trade_step()
+        trade_start, trade_end = self.trade_calendar.get_step_time(trade_step)
+        pred_start, pred_end = self.trade_calendar.get_step_time(trade_step, shift=1)
+        pred_score = self.signal.get_signal(start_time=pred_start, end_time=pred_end)
+        if isinstance(pred_score, pd.DataFrame):
+            pred_score = pred_score.iloc[:, 0]
+        if pred_score is None or len(pred_score) == 0:
+            return TradeDecisionWO([], self)
+        pred_score = pred_score.dropna()
+        if pred_score.empty:
+            return TradeDecisionWO([], self)
+
+        time_per_step = self.trade_calendar.get_freq()
+        current_temp = copy.deepcopy(self.trade_position)
+
+        # ---- signed current holdings ---------------------------------------
+        cur_amount: Dict[str, float] = {}
+        for code in current_temp.get_stock_list():
+            amt = float(current_temp.get_stock_amount(code))
+            if abs(amt) > 1e-6:
+                cur_amount[code] = amt
+
+        # ---- targets: top-k long, bottom-k_short short ----------------------
+        ranked = list(pred_score.sort_values(ascending=False).index)
+        longs: List[str] = []
+        for code in ranked:
+            if len(longs) >= self.topk:
+                break
+            if self._is_tradable(code, trade_start, trade_end):
+                longs.append(code)
+        shorts: List[str] = []
+        if self.allow_short and self._regime_ok(trade_start):
+            stats = self._day_stats(list(ranked), trade_start)
+            for code in reversed(ranked):
+                if len(shorts) >= self.k_short:
+                    break
+                if code in longs:
+                    continue
+                if not self._is_tradable(code, trade_start, trade_end):
+                    continue
+                if self.short_whitelist is not None and code not in self.short_whitelist:
+                    continue
+                st = stats.get(code)
+                if self.short_vol_top_pct is not None:
+                    rank = None if st is None else st.get("vol_rank")
+                    if rank is None or rank < self.short_vol_top_pct:
+                        continue
+                if self.short_max_mom is not None:
+                    mom = None if st is None else st.get("mom")
+                    if mom is None or mom > self.short_max_mom:
+                        continue
+                shorts.append(code)
+
+        # ---- marks & equity --------------------------------------------------
+        marks: Dict[str, float] = {}
+        for code in set(cur_amount) | set(longs) | set(shorts):
+            px = self._mark_price(code, trade_start, trade_end)
+            if px is not None:
+                marks[code] = px
+
+        equity = current_temp.get_cash()
+        for code, amt in cur_amount.items():
+            if code in marks:
+                equity += amt * marks[code]
+        if equity <= 0:
+            return TradeDecisionWO([], self)
+
+        n_legs = len([c for c in longs if c in marks]) + len([c for c in shorts if c in marks])
+        if n_legs == 0:
+            return TradeDecisionWO([], self)
+        per_leg = equity * self.risk_degree / n_legs
+
+        target_signed: Dict[str, float] = {}
+        for code in longs:
+            if code in marks:
+                target_signed[code] = per_leg / marks[code]
+        for code in shorts:
+            if code in marks:
+                target_signed[code] = -(per_leg / marks[code])
+
+        # ---- order generation -------------------------------------------------
+        sell_orders: List[Order] = []
+        buy_orders: List[Order] = []
+
+        def submit(code: str, amount: float, direction: int) -> None:
+            factor = self.trade_exchange.get_factor(stock_id=code, start_time=trade_start, end_time=trade_end)
+            amount = self.trade_exchange.round_amount_by_trade_unit(amount, factor)
+            if amount <= 1e-6:
+                return
+            o = Order(stock_id=code, amount=amount, start_time=trade_start, end_time=trade_end, direction=direction)
+            if self.trade_exchange.check_order(o):
+                (buy_orders if direction == Order.BUY else sell_orders).append(o)
+
+        # close holdings that are no longer targeted (frees cash / unwinds shorts)
+        for code, amt in cur_amount.items():
+            if code in target_signed:
+                continue
+            if marks.get(code) is None:
+                continue
+            if current_temp.get_stock_count(code, bar=time_per_step) < self.hold_thresh:
+                continue
+            submit(code, abs(amt), Order.SELL if amt > 0 else Order.BUY)
+
+        # rebalance targeted legs toward their signed target quantity
+        for code, tgt in target_signed.items():
+            cur = cur_amount.get(code, 0.0)
+            delta = tgt - cur
+            if abs(delta * marks[code]) < max(self.rebalance_tol * per_leg, 1.0):
+                continue
+            if delta > 0:
+                submit(code, delta, Order.BUY)
+            else:
+                if cur > 0 and current_temp.get_stock_count(code, bar=time_per_step) < self.hold_thresh:
+                    continue
+                submit(code, -delta, Order.SELL)
+
+        return TradeDecisionWO(sell_orders + buy_orders, self)
@@ -1,215 +1,231 @@
-"""Regime-gate TopkDropout strategy.
+"""HMM-regime overlay TopkDropout strategy.

-Subclass of ``qlib.contrib.strategy.signal_strategy.TopkDropoutStrategy`` that
-holds the book (issues NO orders) while a regime detector says the market is in
-an unfavorable state.  When the gate is open it behaves exactly like the
-reference TopkDropoutStrategy.
+Regime-gate overlay on ``qlib.contrib.strategy.signal_strategy.TopkDropoutStrategy``:
+selection and sizing are identical to the reference, but a name is only BOUGHT
+(entry gate) when its per-symbol HMM regime posterior ``sp_hmm_p_regime1`` on
+the signal date is >= ``regime_threshold``; otherwise it is held in cash instead
+of being opened.

-Three detector types are supported (all causal — no lookahead):
+The regime posterior is read from the lake feature provider on the fly via
+``qlib.data.D.features`` (field ``$sp_hmm_p_regime1``) for the signal window, so
+no regime column needs to enter the model's ``feature_fields`` — the gate is a
+pure overlay (book ch.01: regime flags regressed as model features, survived
+only as an overlay). The HMM itself was fit with ``fit_end=<train end>`` when
+the lake features were backfilled, so there is no lookahead.

-* ``dispersion``: cross-sectional standard deviation of 22-day rolling returns
-  across the universe.  Gate closes when CS dispersion < threshold (low
-  dispersion means the spread between winners and losers is too narrow for
-  TopkDropout to exploit).
-* ``vol``: cross-sectional mean of 22-day rolling realized volatility.  Gate
-  closes when avg vol is outside a band ``[vol_low, vol_high]`` (strategy
-  needs moderate vol — too calm or too turbulent both hurt).
-* ``hmm``: pre-computed HMM posterior for regime 1 (``sp_hmm_p_regime1``).
-  Gate closes when posterior < threshold (model is not confident the calm
-  regime is active).
-
-The gate is provided as a precomputed ``pd.Series`` of booleans indexed by
-datetime (True = trade allowed).  The companion ``compute_regime_gate``
-function builds this series from lake bars; call it once before backtesting
-and pass the result as the ``regime_gate`` parameter.
+Names already held are NOT force-sold when the regime turns unfavourable
+(entry gate only, matching the queue-10 design).
 """

 from __future__ import annotations

+from typing import List
+
+import numpy as np
 import pandas as pd

-from qlib.backtest.decision import TradeDecisionWO
+from qlib.backtest import Order
+from qlib.backtest.decision import OrderDir, TradeDecisionWO
 from qlib.contrib.strategy.signal_strategy import TopkDropoutStrategy

-__all__ = ["RegimeGateTopkDropoutStrategy", "compute_regime_gate"]
+try:
+    from qlib.data import D
+except ImportError:  # pragma: no cover - qlib always present in this stack
+    D = None
+
+__all__ = ["RegimeGateDropoutStrategy"]
+
+DEFAULT_REGIME_THRESHOLD = 0.5
+REGIME_FIELD = "$sp_hmm_p_regime1"


-class RegimeGateTopkDropoutStrategy(TopkDropoutStrategy):
-    """TopkDropout with a regime-gate circuit breaker.
+class RegimeGateDropoutStrategy(TopkDropoutStrategy):
+    """TopkDropout with an HMM-regime entry gate on buy candidates.

    Parameters
    ----------
    topk, n_drop, method_sell, method_buy, hold_thresh, only_tradable,
    forbid_all_trade_at_limit : same as ``TopkDropoutStrategy``.
-    regime_gate : pd.Series — precomputed per-date gate (bool indexed by
-        datetime).  True = trade allowed, False = no orders.  Missing dates
-        default to open (trade allowed).
+    regime_threshold : minimum ``sp_hmm_p_regime1`` posterior required to open a
+        new position (default 0.5).
    """

-    def __init__(self, *, regime_gate=None, **kwargs):
-        super().__init__(**kwargs)
-        self._regime_gate = regime_gate
+    def __init__(self, *, topk, n_drop, regime_threshold: float = DEFAULT_REGIME_THRESHOLD, **kwargs):
+        super().__init__(topk=topk, n_drop=n_drop, **kwargs)
+        self.regime_threshold = regime_threshold

-    def _gate_open(self, trade_start_time) -> bool:
-        if self._regime_gate is None:
-            return True
-        ts = pd.Timestamp(trade_start_time)
-        known = self._regime_gate[self._regime_gate.index <= ts]
-        if len(known):
-            return bool(known.iloc[-1])
-        return True  # default open if no history yet
+    def _regime_for(self, codes, pred_start, pred_end) -> pd.Series:
+        """Return {code: sp_hmm_p_regime1} for the signal window (last day)."""
+        if D is None:
+            return pd.Series(dtype=float)
+        try:
+            df = D.features(list(codes), [REGIME_FIELD], start_time=pred_start, end_time=pred_end, freq="day")
+        except Exception:  # noqa: BLE001 - a regime read failure should gate open, not crash
+            return pd.Series(dtype=float)
+        if df is None or len(df) == 0:
+            return pd.Series(dtype=float)
+        # df index is MultiIndex (datetime, instrument); take the last day's values
+        df = df.reset_index()
+        ts_col = "datetime" if "datetime" in df.columns else df.columns[0]
+        sym_col = "instrument" if "instrument" in df.columns else df.columns[1]
+        last_ts = df[ts_col].max()
+        last = df[df[ts_col] == last_ts]
+        out = {}
+        for _, row in last.iterrows():
+            sym = str(row[sym_col]).split("/")[-1].upper()
+            val = row.iloc[-1]
+            out[sym] = float(val) if val == val else np.nan
+        return pd.Series(out)

    def generate_trade_decision(self, execute_result=None):
+        import copy
+
        trade_step = self.trade_calendar.get_trade_step()
-        trade_start_time, _ = self.trade_calendar.get_step_time(trade_step)
-        if not self._gate_open(trade_start_time):
+        trade_start_time, trade_end_time = self.trade_calendar.get_step_time(trade_step)
+        pred_start_time, pred_end_time = self.trade_calendar.get_step_time(trade_step, shift=1)
+        pred_score = self.signal.get_signal(start_time=pred_start_time, end_time=pred_end_time)
+        if isinstance(pred_score, pd.DataFrame):
+            pred_score = pred_score.iloc[:, 0]
+        if pred_score is None:
            return TradeDecisionWO([], self)
-        return super().generate_trade_decision(execute_result)

+        if self.only_tradable:

-# ---------------------------------------------------------------------------
-# Precomputation helper
-# ---------------------------------------------------------------------------
+            def get_first_n(li, n, reverse=False):
+                cur_n = 0
+                res = []
+                for si in reversed(li) if reverse else li:
+                    if self.trade_exchange.is_stock_tradable(
+                        stock_id=si, start_time=trade_start_time, end_time=trade_end_time
+                    ):
+                        res.append(si)
+                        cur_n += 1
+                        if cur_n >= n:
+                            break
+                return res[::-1] if reverse else res

-def compute_regime_gate(
-    detector: str,
-    threshold: float = 0.0,
-    *,
-    lake_root: str = "",
-    market: str = "US",
-    start: str = "2015-01-03",
-    end: str = "2026-08-19",
-    vol_low: float = 0.0,
-    vol_high: float = 999.0,
-    hmm_field: str = "sp_hmm_p_regime1",
-) -> pd.Series:
-    """Build a per-date regime gate series from lake bars.
+            def get_last_n(li, n):
+                return get_first_n(li, n, reverse=True)

-    Parameters
-    ----------
-    detector : str — ``"dispersion"``, ``"vol"``, or ``"hmm"``.
-    threshold : float — for ``dispersion``: min CS dispersion to allow trading.
-        For ``hmm``: min HMM posterior to allow trading.
-        Ignored for ``vol`` (uses ``vol_low``/``vol_high`` band instead).
-    lake_root, market : str — lake location.
-    start, end : str — date window.
-    vol_low, vol_high : float — annualized vol band for the ``vol`` detector.
-    hmm_field : str — HMM feature column name for the ``hmm`` detector.
+            def filter_stock(li):
+                return [
+                    si
+                    for si in li
+                    if self.trade_exchange.is_stock_tradable(
+                        stock_id=si, start_time=trade_start_time, end_time=trade_end_time
+                    )
+                ]

-    Returns
-    -------
-    pd.Series — bool, indexed by datetime.  True = trade allowed.
-    """
-    from tac_qlib.data.config import LakeConfig, resolve_lake_root
+        else:

-    cfg = LakeConfig(resolve_lake_root(lake_root or None), market)
-    symbols = _universe_symbols(cfg)
-    close_df, vol_df = _load_daily_bars(symbols, cfg, start, end)
-    if close_df.empty:
-        return pd.Series(dtype=bool)
+            def get_first_n(li, n):
+                return list(li)[:n]

-    if detector == "dispersion":
-        return _dispersion_gate(close_df, threshold)
-    elif detector == "vol":
-        return _vol_gate(close_df, vol_low, vol_high)
-    elif detector == "hmm":
-        return _hmm_gate(cfg, symbols, threshold, start, end, hmm_field)
-    else:
-        raise ValueError(f"Unknown detector: {detector!r}")
+            def get_last_n(li, n):
+                return list(li)[-n:]

+            def filter_stock(li):
+                return li

-def _universe_symbols(cfg) -> list:
-    """Read symbols from the lake symbols.parquet."""
-    import pathlib
+        current_temp: "object" = copy.deepcopy(self.trade_position)
+        sell_order_list: List[Order] = []
+        buy_order_list: List[Order] = []
+        cash = current_temp.get_cash()
+        current_stock_list = current_temp.get_stock_list()
+        last = pred_score.reindex(current_stock_list).sort_values(ascending=False).index

-    sp = cfg.lake_root / "symbols.parquet"
-    if sp.exists():
-        df = pd.read_parquet(sp)
-        col = "symbol" if "symbol" in df.columns else df.columns[0]
-        return sorted(df[col].astype(str).str.upper().tolist())
-    return []
-
-
-def _load_daily_bars(symbols, cfg, start, end):
-    """Load daily close prices for all symbols into a wide DataFrame."""
-    closes = {}
-    vols = {}
-    for sym in symbols:
-        p = cfg.bar_path("1d", sym)
-        if not p.exists():
-            continue
-        try:
-            df = pd.read_parquet(p)
-        except Exception:
-            continue
-        if not len(df):
-            continue
-        tcol = df["t"] if "t" in df.columns else df["date"]
-        ts = pd.to_datetime(tcol)
-        df = df.assign(_t=ts).set_index("_t").sort_index()
-        df = df.loc[start:end]
-        if len(df) < 22:
-            continue
-        closes[sym] = df["c"]
-        if "v" in df.columns:
-            vols[sym] = df["v"]
-    close_df = pd.DataFrame(closes)
-    vol_df = pd.DataFrame(vols) if vols else None
-    return close_df, vol_df
-
-
-def _dispersion_gate(close_df, threshold):
-    """Cross-sectional dispersion of 22-day rolling returns."""
-    if close_df.empty or close_df.shape[1] < 2:
-        return pd.Series(dtype=bool)
-    ret = close_df.pct_change(22)
-    cs_disp = ret.std(axis=1)
-    gate = cs_disp >= threshold
-    gate.iloc[:22] = True  # warmup: allow trading
-    return gate
-
-
-def _vol_gate(close_df, vol_low, vol_high):
-    """Cross-sectional mean of 22-day rolling realized vol."""
-    if close_df.empty or close_df.shape[1] < 2:
-        return pd.Series(dtype=bool)
-    import numpy as np
-    log_ret = np.log(close_df / close_df.shift(1))
-    rv22 = log_ret.rolling(22).std() * (252 ** 0.5)
-    cs_mean_vol = rv22.mean(axis=1)
-    gate = (cs_mean_vol >= vol_low) & (cs_mean_vol <= vol_high)
-    gate.iloc[:22] = True  # warmup
-    return gate
-
-
-def _hmm_gate(cfg, symbols, threshold, start, end, hmm_field):
-    """HMM regime posterior gate from persisted SP features."""
-    feat_root = cfg.lake_root / "features"
-    all_posteriors = {}
-    for sym in symbols:
-        # check both ta and sp family paths
-        for family in ("sp", "ta"):
-            p = feat_root / f"market=US" / f"timeframe=1d" / f"family={family}" / f"symbol={sym}.parquet"
-            if not p.exists():
-                continue
+        if self.method_buy == "top":
+            today = get_first_n(
+                pred_score[~pred_score.index.isin(last)].sort_values(ascending=False).index,
+                self.n_drop + self.topk - len(last),
+            )
+        elif self.method_buy == "random":
+            topk_candi = get_first_n(pred_score.sort_values(ascending=False).index, self.topk)
+            candi = list(filter(lambda x: x not in last, topk_candi))
+            n = self.n_drop + self.topk - len(last)
            try:
-                df = pd.read_parquet(p)
-            except Exception:
+                today = np.random.choice(candi, n, replace=False)
+            except ValueError:
+                today = candi
+        else:
+            raise NotImplementedError(f"This type of input is not supported")
+
+        comb = pred_score.reindex(last.union(pd.Index(today))).sort_values(ascending=False).index
+
+        if self.method_sell == "bottom":
+            sell = last[last.isin(get_last_n(comb, self.n_drop))]
+        elif self.method_sell == "random":
+            candi = filter_stock(last)
+            try:
+                sell = pd.Index(np.random.choice(candi, self.n_drop, replace=False) if len(last) else [])
+            except ValueError:
+                sell = candi
+        else:
+            raise NotImplementedError(f"This type of input is not supported")
+
+        buy = today[: len(sell) + self.topk - len(last)]
+
+        # ---- regime gate -----------------------------------------------------
+        if buy:
+            regime = self._regime_for(buy, pred_start_time, pred_end_time)
+            gated = [c for c in buy if regime.get(c, np.nan) >= self.regime_threshold]
+        else:
+            gated = []
+
+        for code in current_stock_list:
+            if not self.trade_exchange.is_stock_tradable(
+                stock_id=code,
+                start_time=trade_start_time,
+                end_time=trade_end_time,
+                direction=None if self.forbid_all_trade_at_limit else OrderDir.SELL,
+            ):
                continue
-            if hmm_field not in df.columns:
+            if code in sell:
+                time_per_step = self.trade_calendar.get_freq()
+                if current_temp.get_stock_count(code, bar=time_per_step) < self.hold_thresh:
+                    continue
+                sell_amount = current_temp.get_stock_amount(code=code)
+                sell_order = Order(
+                    stock_id=code,
+                    amount=sell_amount,
+                    start_time=trade_start_time,
+                    end_time=trade_end_time,
+                    direction=Order.SELL,
+                )
+                if self.trade_exchange.check_order(sell_order):
+                    sell_order_list.append(sell_order)
+                    trade_val, trade_cost, trade_price = self.trade_exchange.deal_order(
+                        sell_order, position=current_temp
+                    )
+                    cash += trade_val - trade_cost
+
+        if len(gated) == 0:
+            return TradeDecisionWO(sell_order_list, self)
+
+        value = cash * self.risk_degree / len(gated)
+        for code in gated:
+            if not self.trade_exchange.is_stock_tradable(
+                stock_id=code,
+                start_time=trade_start_time,
+                end_time=trade_end_time,
+                direction=None if self.forbid_all_trade_at_limit else OrderDir.BUY,
+            ):
                continue
-            tcol = df["t"] if "t" in df.columns else df["date"]
-            ts = pd.to_datetime(tcol)
-            s = pd.Series(df[hmm_field].values, index=ts, name=sym)
-            s = s.loc[start:end].dropna()
-            if len(s) > 0:
-                all_posteriors[sym] = s
-                break
-    if not all_posteriors:
-        # no HMM features found — default open
-        idx = pd.date_range(start, end, freq="B")
-        return pd.Series(True, index=idx)
-    post_df = pd.DataFrame(all_posteriors)
-    cs_mean = post_df.mean(axis=1)
-    gate = cs_mean >= threshold
-    return gate
+            buy_price = self.trade_exchange.get_deal_price(
+                stock_id=code, start_time=trade_start_time, end_time=trade_end_time, direction=OrderDir.BUY
+            )
+            buy_amount = value / buy_price
+            factor = self.trade_exchange.get_factor(
+                stock_id=code, start_time=trade_start_time, end_time=trade_end_time
+            )
+            buy_amount = self.trade_exchange.round_amount_by_trade_unit(buy_amount, factor)
+            buy_order = Order(
+                stock_id=code,
+                amount=buy_amount,
+                start_time=trade_start_time,
+                end_time=trade_end_time,
+                direction=Order.BUY,
+            )
+            buy_order_list.append(buy_order)
+
+        return TradeDecisionWO(sell_order_list + buy_order_list, self)
@@ -1,301 +0,0 @@
-"""Signal-quality gate TopkDropout strategy.
-
-Subclass of ``qlib.contrib.strategy.signal_strategy.TopkDropoutStrategy`` that
-holds the book (issues NO orders) when the model's recent prediction accuracy
-is below a threshold.  When the gate is open it behaves exactly like the
-reference TopkDropoutStrategy.
-
-Unlike the regime gate (which asks "is the market calm?"), the signal-quality
-gate asks "are my predictions accurate?" — and works across ALL years.
-
-The gate is provided as a precomputed ``pd.Series`` of booleans indexed by
-datetime (True = trade allowed).  The companion ``compute_signal_quality_gate``
-function builds this series from a pred.pkl and lake bars; call it once before
-backtesting and pass the result as the ``signal_quality_gate`` parameter.
-"""
-
-from __future__ import annotations
-
-import numpy as np
-import pandas as pd
-
-from qlib.backtest.decision import TradeDecisionWO
-from qlib.contrib.strategy.signal_strategy import TopkDropoutStrategy
-
-__all__ = ["SignalQualityGateStrategy", "compute_signal_quality_gate"]
-
-
-class SignalQualityGateStrategy(TopkDropoutStrategy):
-    """TopkDropout with signal-quality gate overlay.
-
-    When ``lake_root`` is provided the gate is computed on-the-fly from the
-    signal (``<PRED>``) and close prices — no precomputed gate file needed.
-    This ensures the gate matches the model that is actually generating the
-    predictions (critical when the model is retrained each year).
-
-    Parameters
-    ----------
-    topk, n_drop, method_sell, method_buy, hold_thresh, only_tradable,
-    forbid_all_trade_at_limit : same as ``TopkDropoutStrategy``.
-    signal_quality_gate : pd.Series — precomputed per-date gate (bool).
-    signal_quality_gate_path : str — path to pickled gate Series.
-    lake_root : str — lake root for on-the-fly gate computation (preferred).
-    gate_topk, gate_lookback, gate_threshold : int/float — gate params.
-    gate_start, gate_end : str — date window for loading close prices.
-    """
-
-    def __init__(self, *, signal_quality_gate=None, signal_quality_gate_path=None,
-                 lake_root=None, gate_topk=10, gate_lookback=5, gate_threshold=0.5,
-                 gate_start="2015-01-03", gate_end="2026-08-19", **kwargs):
-        super().__init__(**kwargs)
-        self._sq_gate_computed = False
-        if signal_quality_gate is not None:
-            self._sq_gate = signal_quality_gate
-            self._sq_gate_computed = True
-        elif signal_quality_gate_path is not None:
-            import pickle
-            with open(signal_quality_gate_path, "rb") as f:
-                self._sq_gate = pickle.load(f)
-            self._sq_gate_computed = True
-        elif lake_root is not None:
-            self._sq_gate = None
-            self._lake_root = lake_root
-            self._gate_topk = gate_topk
-            self._gate_lookback = gate_lookback
-            self._gate_threshold = gate_threshold
-            self._gate_start = gate_start
-            self._gate_end = gate_end
-        else:
-            self._sq_gate = None
-
-    def _compute_gate_on_fly(self):
-        """Compute gate from the signal (pred.pkl) and lake close prices."""
-        import pickle
-        from pathlib import Path
-
-        signal_path = self._signal
-        if not Path(signal_path).exists():
-            return
-
-        with open(signal_path, "rb") as f:
-            pred = pickle.load(f)
-
-        # Handle MultiIndex DataFrame -> unstack to wide
-        if isinstance(pred, pd.DataFrame) and isinstance(pred.index, pd.MultiIndex):
-            pred = pred.iloc[:, 0]
-            pred.index = pd.MultiIndex.from_arrays([
-                pd.to_datetime(pred.index.get_level_values(0)).normalize(),
-                pred.index.get_level_values(1)
-            ])
-            pred = pred.unstack(level=1)
-        elif isinstance(pred, pd.DataFrame):
-            pred = pred.iloc[:, 0] if pred.shape[1] >= 1 else pred.squeeze()
-            pred.index = pd.to_datetime(pred.index).normalize()
-
-        # Load close prices from lake
-        close_df = _load_close_prices(self._lake_root, "US", self._gate_start, self._gate_end)
-        if close_df.empty:
-            return
-
-        ret_df = close_df.pct_change()
-        ret_df.index = pd.to_datetime(ret_df.index).normalize()
-
-        pred_dates = sorted(pred.index.unique())
-        if len(pred_dates) < 2:
-            self._sq_gate = pd.Series(True, index=pd.DatetimeIndex(pred_dates))
-            self._sq_gate_computed = True
-            return
-
-        hit_rates = {}
-        for i in range(1, len(pred_dates)):
-            day = pred_dates[i]
-            prev_day = pred_dates[i - 1]
-            try:
-                prev_scores = pred.loc[prev_day]
-            except KeyError:
-                continue
-            if isinstance(prev_scores, pd.DataFrame):
-                prev_scores = prev_scores.iloc[:, 0]
-            prev_scores = prev_scores.dropna().sort_values(ascending=False)
-            topk_syms = list(prev_scores.index[:self._gate_topk])
-
-            if day not in ret_df.index:
-                continue
-            today_ret = ret_df.loc[day]
-            topk_rets = today_ret.reindex(topk_syms).dropna()
-            if len(topk_rets) == 0:
-                continue
-
-            hit_rates[day] = (topk_rets > 0).sum() / len(topk_rets)
-
-        if not hit_rates:
-            self._sq_gate_computed = True
-            return
-
-        hr_series = pd.Series(hit_rates).sort_index()
-        rolling_hr = hr_series.rolling(self._gate_lookback, min_periods=1).mean()
-        gate = rolling_hr >= self._gate_threshold
-        gate.iloc[:self._gate_lookback] = True
-
-        self._sq_gate = gate
-        self._sq_gate_computed = True
-
-    def _gate_open(self, trade_start_time) -> bool:
-        if self._sq_gate is None:
-            return True
-        ts = pd.Timestamp(trade_start_time)
-        known = self._sq_gate[self._sq_gate.index <= ts]
-        if len(known):
-            return bool(known.iloc[-1])
-        return True  # default open if no history yet
-
-    def generate_trade_decision(self, execute_result=None):
-        if not self._sq_gate_computed:
-            self._compute_gate_on_fly()
-        trade_step = self.trade_calendar.get_trade_step()
-        trade_start_time, _ = self.trade_calendar.get_step_time(trade_step)
-        if not self._gate_open(trade_start_time):
-            return TradeDecisionWO([], self)
-        return super().generate_trade_decision(execute_result)
-
-
-# ---------------------------------------------------------------------------
-# Precomputation helper
-# ---------------------------------------------------------------------------
-
-def compute_signal_quality_gate(
-    pred_path: str,
-    *,
-    lake_root: str = "",
-    market: str = "US",
-    topk: int = 10,
-    lookback: int = 5,
-    threshold: float = 0.5,
-    start: str = "2015-01-03",
-    end: str = "2026-08-19",
-) -> pd.Series:
-    """Build a per-date signal-quality gate series from a pred.pkl and lake bars.
-
-    For each day, checks whether the model's topk picks from the previous day
-    had positive returns.  Computes a rolling hit rate over ``lookback`` days
-    and opens the gate when hit rate >= ``threshold``.
-
-    Parameters
-    ----------
-    pred_path : str — path to pred.pkl (from rd_train / rd_predict).
-    lake_root, market : str — lake location (for loading close prices).
-    topk : int — number of top picks to track for hit rate.
-    lookback : int — rolling window for hit rate computation.
-    threshold : float — hit rate threshold to keep trading.
-    start, end : str — date window for loading prices.
-
-    Returns
-    -------
-    pd.Series — bool, indexed by datetime.  True = trade allowed.
-    """
-    from pathlib import Path
-    import pickle
-
-    # Load pred.pkl
-    with open(pred_path, "rb") as f:
-        pred = pickle.load(f)
-
-    # Handle MultiIndex DataFrame (datetime, instrument) -> unstack to wide
-    if isinstance(pred, pd.DataFrame) and isinstance(pred.index, pd.MultiIndex):
-        pred = pred.iloc[:, 0]  # take score column as Series
-        pred.index = pd.MultiIndex.from_arrays([
-            pd.to_datetime(pred.index.get_level_values(0)).normalize(),
-            pred.index.get_level_values(1)
-        ])
-        # Unstack to wide: dates x instruments
-        pred = pred.unstack(level=1)
-    elif isinstance(pred, pd.DataFrame):
-        pred = pred.iloc[:, 0] if pred.shape[1] >= 1 else pred.squeeze()
-        pred.index = pd.to_datetime(pred.index).normalize()
-
-    # Load close prices from lake
-    close_df = _load_close_prices(lake_root, market, start, end)
-    if close_df.empty:
-        return pd.Series(dtype=bool)
-
-    ret_df = close_df.pct_change()
-    ret_df.index = pd.to_datetime(ret_df.index).normalize()
-
-    # Get sorted unique prediction dates
-    pred_dates = sorted(pred.index.unique())
-    if len(pred_dates) < 2:
-        return pd.Series(True, index=pd.DatetimeIndex(pred_dates))
-
-    # Compute hit rates
-    hit_rates = {}
-    for i in range(1, len(pred_dates)):
-        day = pred_dates[i]
-        # Get yesterday's topk
-        prev_day = pred_dates[i - 1]
-        try:
-            prev_scores = pred.loc[prev_day]
-        except KeyError:
-            continue
-        if isinstance(prev_scores, pd.DataFrame):
-            prev_scores = prev_scores.iloc[:, 0]
-        prev_scores = prev_scores.dropna().sort_values(ascending=False)
-        topk_syms = list(prev_scores.index[:topk])
-
-        # Get today's returns
-        if day not in ret_df.index:
-            continue
-        today_ret = ret_df.loc[day]
-        topk_rets = today_ret.reindex(topk_syms).dropna()
-        if len(topk_rets) == 0:
-            continue
-
-        hit_rates[day] = (topk_rets > 0).sum() / len(topk_rets)
-
-    if not hit_rates:
-        return pd.Series(dtype=bool)
-
-    hr_series = pd.Series(hit_rates).sort_index()
-
-    # Rolling hit rate
-    rolling_hr = hr_series.rolling(lookback, min_periods=1).mean()
-
-    # Gate is open when rolling hit rate >= threshold
-    gate = rolling_hr >= threshold
-    gate.iloc[:lookback] = True  # warmup: allow trading
-
-    return gate
-
-
-def _load_close_prices(lake_root, market, start, end):
-    """Load daily close prices for all symbols into a wide DataFrame."""
-    from pathlib import Path
-
-    lake = Path(lake_root)
-    symbols_parquet = lake / "symbols.parquet"
-    if not symbols_parquet.exists():
-        return pd.DataFrame()
-
-    df = pd.read_parquet(symbols_parquet)
-    col = "symbol" if "symbol" in df.columns else df.columns[0]
-    symbols = sorted(df[col].astype(str).str.upper().tolist())
-
-    closes = {}
-    for sym in symbols:
-        p = lake / "market=US" / "timeframe=1d" / f"symbol={sym}.parquet"
-        if not p.exists():
-            continue
-        try:
-            bar = pd.read_parquet(p)
-        except Exception:
-            continue
-        if not len(bar):
-            continue
-        tcol = "t" if "t" in bar.columns else "date"
-        ts = pd.to_datetime(bar[tcol])
-        bar = bar.assign(_t=ts).set_index("_t").sort_index()
-        bar = bar.loc[start:end]
-        if len(bar) < 10:
-            continue
-        closes[sym] = bar["c"] if "c" in bar.columns else bar["close"]
-
-    return pd.DataFrame(closes)
@@ -37,11 +37,20 @@ class WeeklyRebalanceDropoutStrategy(TopkDropoutStrategy):
    forbid_all_trade_at_limit : same as ``TopkDropoutStrategy``.
    hold_band_pct : skip order for a name whose deviation from target weight is
        below this fraction of the target (no-trade buffer band).
+    rebalance_every_n_weeks : rebalance every N ISO weeks instead of every week
+        (default 1 = weekly; 2 = biweekly).  Ignored when
+        ``rebalance_every_n_days`` is set.
+    rebalance_every_n_days : rebalance every N trading days (daily when N=1).
+        When set, overrides the weekly gating logic entirely.
    """

-    def __init__(self, *, topk, n_drop, hold_band_pct: float = DEFAULT_HOLD_BAND_PCT, **kwargs):
+    def __init__(self, *, topk, n_drop, hold_band_pct: float = DEFAULT_HOLD_BAND_PCT,
+                 rebalance_every_n_weeks: int = 1,
+                 rebalance_every_n_days: int = 0, **kwargs):
        super().__init__(topk=topk, n_drop=n_drop, **kwargs)
        self.hold_band_pct = hold_band_pct
+        self.rebalance_every_n_weeks = rebalance_every_n_weeks
+        self.rebalance_every_n_days = rebalance_every_n_days

    @staticmethod
    def _iso_week(ts) -> tuple:
@@ -53,13 +62,25 @@ class WeeklyRebalanceDropoutStrategy(TopkDropoutStrategy):
        trade_step = self.trade_calendar.get_trade_step()
        trade_start_time, trade_end_time = self.trade_calendar.get_step_time(trade_step)

-        cur_week = self._iso_week(trade_start_time)
-        prev_week = getattr(self, "_last_week", None)
-        self._last_week = cur_week
+        if self.rebalance_every_n_days > 0:
+            # daily gating: count trading steps since last rebalance
+            step_num = trade_step
+            if hasattr(self, "_last_rebal_step"):
+                if (step_num - self._last_rebal_step) < self.rebalance_every_n_days:
+                    return TradeDecisionWO([], self)
+            self._last_rebal_step = step_num
+        else:
+            cur_week = self._iso_week(trade_start_time)
+            prev_week = getattr(self, "_last_week", None)
+            self._last_week = cur_week

-        if prev_week is not None and prev_week == cur_week:
-            # not the first trading day of this ISO week -> hold
-            return TradeDecisionWO([], self)
+            if prev_week is not None and prev_week == cur_week:
+                return TradeDecisionWO([], self)
+
+            if self.rebalance_every_n_weeks > 1:
+                week_num = cur_week[1]
+                if prev_week is not None and (week_num % self.rebalance_every_n_weeks) != 1:
+                    return TradeDecisionWO([], self)

        pred_start_time, pred_end_time = self.trade_calendar.get_step_time(trade_step, shift=1)
        pred_score = self.signal.get_signal(start_time=pred_start_time, end_time=pred_end_time)
@@ -61,6 +61,11 @@ UNKNOWN_FIELD_NAMES = ("factor", "change", "trade_unit", "suspend_flag")
 #: columns in the parquet files that are not features
 NON_FEATURE_COLUMNS = ("t", "date", "market", "timeframe", "symbol")

+#: Feature-family partitions merged by ``LakeConfig.load_features`` and scanned
+#: by the handler's field discovery. ``macro`` holds broadcast market-state
+#: columns (see skills/tac-qlib-custom/examples/persist_macro_broadcast.py).
+FEATURE_FAMILIES = ("ta", "sp", "macro")
+

 def timeframe_for_freq(freq: str) -> str:
    """Map a qlib frequency (e.g. ``day``, ``1min``) to a lake timeframe (e.g. ``1d``)."""
@@ -113,12 +118,12 @@ class LakeConfig:
        return self.features_dir(timeframe) / f"symbol={str(symbol).upper()}.parquet"

    def load_features(self, timeframe: str, symbol: str) -> pd.DataFrame:
-        """All feature columns for a symbol, merging the `family=ta` and
-        `family=sp` partitions by timestamp. Returns an empty frame when no
+        """All feature columns for a symbol, merging the `family=ta|sp|macro`
+        partitions by timestamp. Returns an empty frame when no
        feature files exist (legacy flat layout falls back transparently)."""
        sym = str(symbol).upper()
        frames = []
-        for family in ("ta", "sp"):
+        for family in FEATURE_FAMILIES:
            p = self.features_dir(timeframe) / f"family={family}" / f"symbol={sym}.parquet"
            if p.exists():
                frames.append(pd.read_parquet(p))
@@ -1,44 +1,73 @@
-# Signal-quality gate backtest for 2023
+# -----------------------------------------------------------------------------
+# QUEUE-01 — M2 reproduction: risk-adjusted 22d Sharpe drift (sp_sharpe_22).
+#
+# Hypothesis (book ch.01/ch.07, EVIDENCE#018 -> exp 30): adding the
+# risk-adjusted 22d Sharpe drift feature (sp_sharpe_22) to the compact
+# stochastic reference IMPROVES net portfolio performance (exp 30: net +6.53%
+# IR 0.62 vs reference +2.13% IR 0.21) while rank metrics dip (RankIC 0.0576 vs
+# 0.0663). exp 30 is a SINGLE clean-lake run, unreproduced -> HYPOTHESIS.
+#
+# Change vs exp-26 reference (EVIDENCE#015, run 21afc6af...): ONE feature added,
+# feature_fields = compact set + sp_sharpe_22. Everything else byte-identical.
+#
+# Acceptance: net_ann_return > +2.13% AND net_IR > 0.21 (else HYPOTHESIS -> REFUTED).
+# Run: rd_run_workflow config_path=<repo>/experiments/queue/workflows/q01_m2_sharpe22_repro.yaml \
+#        experiment_name=tac-rd-q01-m2-sharpe22-repro
+# -----------------------------------------------------------------------------
 {%- set LAKE = TAC_LAKE_DIR %}
 {%- set UNIVERSE = "SPY,QQQ,DIA,IWM,MDY,VTI,VOO,VEA,VWO,VT,EFA,EEM,TLT,IEF,SHY,AGG,BND,LQD,HYG,JNK,EMB,GLD,SLV,USO,UNG,DBA,DBC,XLK,XLF,XLE,XLV,XLI,XLY,XLP,XLU,XLB,XLRE,ARKK,SMH,SOXX,IBB,XBI,ITA,XAR,ICLN,TAN,FDN,IGV,ESPO,REM" %}
-{%- set SP_FIELDS = "sp_ret,sp_ou_zscore,sp_ou_half_life,sp_ou_revert,sp_hmm_p_regime1,sp_hmm_state,sp_jump_ratio,sp_jump_flag,sp_jump_tail,sp_max_move,sp_rv1,sp_rv5,sp_rv22,sp_vol_ratio_5_22,sp_vol_ratio_1_22,sp_trend_slope_5,sp_trend_slope_20,sp_trend_slope_60,sp_logp,sp_hurst_exponent,sp_sig_level1_lead,sp_sig_level1_lag,sp_sig_level2_lead_lag,sp_sig_level2_lag_lead" %}
-{# Gate computed on-the-fly from signal + lake close prices #}
+{%- set FEATURES = "$open,$high,$low,$close,$vwap,$volume,sp_ret,sp_jump_ratio,sp_jump_flag,sp_jump_tail,sp_max_move,sp_rv1,sp_rv5,sp_rv22,sp_vol_ratio_5_22,sp_vol_ratio_1_22,sp_trend_slope_5,sp_trend_slope_20,sp_trend_slope_60,sp_logp,sp_hurst_exponent,sp_sig_level1_lead,sp_sig_level1_lag,sp_sig_level2_lead_lag,sp_sig_level2_lag_lead,sp_sharpe_22" %}

 qlib_init:
    provider_uri: "{{ LAKE }}"
    region: us
    expression_cache: null
    dataset_cache: null
+
    calendar_provider:
        class: tac_qlib.data.providers.LakeCalendarProvider
-        kwargs: { lake_root: "{{ LAKE }}", market: US }
+        kwargs:
+            lake_root: "{{ LAKE }}"
+            market: US
    instrument_provider:
        class: tac_qlib.data.providers.LakeInstrumentProvider
-        kwargs: { lake_root: "{{ LAKE }}", market: US, markets: {} }
+        kwargs:
+            lake_root: "{{ LAKE }}"
+            market: US
+            markets: {}
    feature_provider:
        class: tac_qlib.data.providers.LakeFeatureProvider
-        kwargs: { lake_root: "{{ LAKE }}", market: US }
+        kwargs:
+            lake_root: "{{ LAKE }}"
+            market: US
+
    exp_manager:
        class: MLflowExpManager
        module_path: qlib.workflow.expm
        kwargs:
            uri: "sqlite:///{{ LAKE }}/mlruns.db"
-            default_exp_name: "tac-rd-sq-gate-2023"
+            default_exp_name: "tac-rd-q01-m2-sharpe22-repro"

 task:
    model:
-        class: LGBModel
-        module_path: qlib.contrib.model.gbdt
+        class: RankICEnsembleLGBModel
+        module_path: tac_qlib.contrib.model.rank_ensemble
        kwargs:
            loss: mse
-            learning_rate: 0.05
-            num_leaves: 15
-            n_estimators: 200
+            learning_rate: 0.02
+            num_leaves: 31
+            n_estimators: 3000
+            num_boost_round: 3000
+            early_stopping_rounds: 200
+            min_data_in_leaf: 20
+            lambda_l2: 0.5
            colsample_bytree: 0.8
            subsample: 0.8
            subsample_freq: 1
-            reg_alpha: 0.01
-            reg_lambda: 0.01
+            reg_alpha: 0.1
+            reg_lambda: 1.0
+            seeds: "42,7,2026,99,123"
+            parallel: 5

    dataset:
        class: DatasetH
@@ -50,14 +79,14 @@ task:
                kwargs:
                    instruments: "{{ UNIVERSE }}"
                    start_time: 2015-01-03
-                    end_time: 2023-12-29
-                    fit_start_time: 2015-01-03
-                    fit_end_time: 2023-01-03
+                    end_time: 2026-08-10
+                    fit_start_time: 2016-01-04
+                    fit_end_time: 2025-09-01
                    freq: day
                    lake_root: "{{ LAKE }}"
                    market: US
                    label: "Ref($close,-6)/Ref($close,-1)-1"
-                    feature_fields: "$open,$high,$low,$close,$vwap,$volume,{{ SP_FIELDS }}"
+                    feature_fields: "{{ FEATURES }}"
                    infer_processors:
                        - class: DropAllNaN
                          kwargs: {}
@@ -70,9 +99,9 @@ task:
                        - class: Fillna
                          kwargs: {}
            segments:
-                train: [2015-01-03, 2022-09-01]
-                valid: [2022-09-03, 2023-01-03]
-                test: [2023-01-03, 2023-12-29]
+                train: [2016-01-04, 2025-09-01]
+                valid: [2025-09-03, 2026-01-03]
+                test: [2026-01-04, 2026-08-10]

    record:
        - class: SignalRecord
@@ -80,29 +109,25 @@ task:
          kwargs: {}
        - class: SigAnaRecord
          module_path: qlib.workflow.record_temp
-          kwargs: { ana_long_short: true, ann_scaler: 252 }
+          kwargs:
+              ana_long_short: true
+              ann_scaler: 252
        - class: PortAnaRecord
          module_path: qlib.workflow.record_temp
          kwargs:
              config:
                  strategy:
-                      class: SignalQualityGateStrategy
-                      module_path: tac_qlib.contrib.strategy.signal_quality_gate
+                      class: TopkDropoutStrategy
+                      module_path: qlib.contrib.strategy
                      kwargs:
                          signal: "<PRED>"
-                          lake_root: "{{ LAKE }}"
-                          gate_topk: 10
-                          gate_lookback: 5
-                          gate_threshold: 0.5
-                          gate_start: "2015-01-03"
-                          gate_end: "2023-12-29"
                          topk: 10
                          n_drop: 1
                          only_tradable: true
                          risk_degree: 0.95
                  backtest:
-                      start_time: 2023-01-03
-                      end_time: 2023-12-29
+                      start_time: 2026-01-04
+                      end_time: 2026-08-10
                      account: 1000000
                      benchmark: SPY
                      exchange_kwargs:
@@ -112,4 +137,4 @@ task:
                          open_cost: 0.0005
                          close_cost: 0.0015
                          min_cost: 5.0
-              risk_analysis_freq: 1d
+              risk_analysis_freq: 1d
@@ -1,16 +1,16 @@
 # -----------------------------------------------------------------------------
-# Signal-quality gate: TopkDropout gated by rolling hit-rate of topk picks.
+# ABLATION A (baseline): LightGBM with RankIC early-stopping on the 50-ETF SP-5d
+# panel, using ALL 24 sp_* feature columns (ou,hmm,jump,har,trend,hurst,
+# signature). Copy of the canonical workflow_lgb_sp5d_rankic.yaml with a
+# distinct experiment name so the ablation runs are isolated.
 #
-# 1. Compute the gate: python precompute_signal_quality_gate.py <pred.pkl> <gate.pkl>
-# 2. Run this workflow: rd_run_workflow config_path=<this yaml> experiment_name=<exp>
-#
-# The strategy loads the precomputed gate from signal_quality_gate_path.
-# When hit rate >= threshold, trade; otherwise, go to cash.
+# Run:
+#   rd_run_workflow config_path=tac-qlib/workflows/ablate_baseline_all_sp_fields.yaml \
+#       experiment_name=tac-rd-rank-ablate
 # -----------------------------------------------------------------------------
 {%- set LAKE = TAC_LAKE_DIR %}
 {%- set UNIVERSE = "SPY,QQQ,DIA,IWM,MDY,VTI,VOO,VEA,VWO,VT,EFA,EEM,TLT,IEF,SHY,AGG,BND,LQD,HYG,JNK,EMB,GLD,SLV,USO,UNG,DBA,DBC,XLK,XLF,XLE,XLV,XLI,XLY,XLP,XLU,XLB,XLRE,ARKK,SMH,SOXX,IBB,XBI,ITA,XAR,ICLN,TAN,FDN,IGV,ESPO,REM" %}
 {%- set SP_FIELDS = "sp_ret,sp_ou_zscore,sp_ou_half_life,sp_ou_revert,sp_hmm_p_regime1,sp_hmm_state,sp_jump_ratio,sp_jump_flag,sp_jump_tail,sp_max_move,sp_rv1,sp_rv5,sp_rv22,sp_vol_ratio_5_22,sp_vol_ratio_1_22,sp_trend_slope_5,sp_trend_slope_20,sp_trend_slope_60,sp_logp,sp_hurst_exponent,sp_sig_level1_lead,sp_sig_level1_lag,sp_sig_level2_lead_lag,sp_sig_level2_lag_lead" %}
-{%- set GATE_PATH = "/app/experiments/book/data/signal_quality_gate/sq_gate_5d_0.50.pkl" %}

 qlib_init:
    provider_uri: "{{ LAKE }}"
@@ -40,22 +40,27 @@ qlib_init:
        module_path: qlib.workflow.expm
        kwargs:
            uri: "sqlite:///{{ LAKE }}/mlruns.db"
-            default_exp_name: "tac-rd-sq-gate"
+            default_exp_name: "tac-rd-rank-ablate"

 task:
    model:
-        class: LGBModel
-        module_path: qlib.contrib.model.gbdt
+        class: RankICLGBModel
+        module_path: tac_qlib.contrib.model.rank_gbdt
        kwargs:
            loss: mse
-            learning_rate: 0.05
-            num_leaves: 15
-            n_estimators: 200
+            learning_rate: 0.02
+            num_leaves: 31
+            n_estimators: 3000
+            num_boost_round: 3000
+            early_stopping_rounds: 200
+            min_data_in_leaf: 20
+            lambda_l2: 0.5
            colsample_bytree: 0.8
            subsample: 0.8
            subsample_freq: 1
-            reg_alpha: 0.01
-            reg_lambda: 0.01
+            reg_alpha: 0.1
+            reg_lambda: 1.0
+            seed: 42

    dataset:
        class: DatasetH
@@ -105,13 +110,12 @@ task:
          kwargs:
              config:
                  strategy:
-                      class: SignalQualityGateStrategy
-                      module_path: tac_qlib.contrib.strategy.signal_quality_gate
+                      class: TopkDropoutStrategy
+                      module_path: qlib.contrib.strategy
                      kwargs:
                          signal: "<PRED>"
-                          signal_quality_gate_path: "{{ GATE_PATH }}"
                          topk: 10
-                          n_drop: 1
+                          n_drop: 2
                          only_tradable: true
                          risk_degree: 0.95
                  backtest:
@@ -1,44 +1,67 @@
-# Signal-quality gate backtest for 2026
+# -----------------------------------------------------------------------------
+# ABLATION B (generic-only): same panel/model as the baseline, but feature
+# fields restricted to the model-free / generic stochastic-process families
+# (jump,har,trend,hurst,signature). Drops the model-specific ou (OU/AR-1
+# half-life) and hmm (2-state regime) families to test whether the generic
+# families alone dominate the rank dimension.
+#
+# Run:
+#   rd_run_workflow config_path=tac-qlib/workflows/ablate_generic_only_sp_fields.yaml \
+#       experiment_name=tac-rd-rank-ablate
+# -----------------------------------------------------------------------------
 {%- set LAKE = TAC_LAKE_DIR %}
 {%- set UNIVERSE = "SPY,QQQ,DIA,IWM,MDY,VTI,VOO,VEA,VWO,VT,EFA,EEM,TLT,IEF,SHY,AGG,BND,LQD,HYG,JNK,EMB,GLD,SLV,USO,UNG,DBA,DBC,XLK,XLF,XLE,XLV,XLI,XLY,XLP,XLU,XLB,XLRE,ARKK,SMH,SOXX,IBB,XBI,ITA,XAR,ICLN,TAN,FDN,IGV,ESPO,REM" %}
-{%- set SP_FIELDS = "sp_ret,sp_ou_zscore,sp_ou_half_life,sp_ou_revert,sp_hmm_p_regime1,sp_hmm_state,sp_jump_ratio,sp_jump_flag,sp_jump_tail,sp_max_move,sp_rv1,sp_rv5,sp_rv22,sp_vol_ratio_5_22,sp_vol_ratio_1_22,sp_trend_slope_5,sp_trend_slope_20,sp_trend_slope_60,sp_logp,sp_hurst_exponent,sp_sig_level1_lead,sp_sig_level1_lag,sp_sig_level2_lead_lag,sp_sig_level2_lag_lead" %}
-{# Gate computed on-the-fly from signal + lake close prices #}
+{%- set SP_FIELDS = "sp_ret,sp_jump_ratio,sp_jump_flag,sp_jump_tail,sp_max_move,sp_rv1,sp_rv5,sp_rv22,sp_vol_ratio_5_22,sp_vol_ratio_1_22,sp_trend_slope_5,sp_trend_slope_20,sp_trend_slope_60,sp_logp,sp_hurst_exponent,sp_sig_level1_lead,sp_sig_level1_lag,sp_sig_level2_lead_lag,sp_sig_level2_lag_lead" %}

 qlib_init:
    provider_uri: "{{ LAKE }}"
    region: us
    expression_cache: null
    dataset_cache: null
+
    calendar_provider:
        class: tac_qlib.data.providers.LakeCalendarProvider
-        kwargs: { lake_root: "{{ LAKE }}", market: US }
+        kwargs:
+            lake_root: "{{ LAKE }}"
+            market: US
    instrument_provider:
        class: tac_qlib.data.providers.LakeInstrumentProvider
-        kwargs: { lake_root: "{{ LAKE }}", market: US, markets: {} }
+        kwargs:
+            lake_root: "{{ LAKE }}"
+            market: US
+            markets: {}
    feature_provider:
        class: tac_qlib.data.providers.LakeFeatureProvider
-        kwargs: { lake_root: "{{ LAKE }}", market: US }
+        kwargs:
+            lake_root: "{{ LAKE }}"
+            market: US
+
    exp_manager:
        class: MLflowExpManager
        module_path: qlib.workflow.expm
        kwargs:
            uri: "sqlite:///{{ LAKE }}/mlruns.db"
-            default_exp_name: "tac-rd-sq-gate-2026"
+            default_exp_name: "tac-rd-rank-ablate"

 task:
    model:
-        class: LGBModel
-        module_path: qlib.contrib.model.gbdt
+        class: RankICLGBModel
+        module_path: tac_qlib.contrib.model.rank_gbdt
        kwargs:
            loss: mse
-            learning_rate: 0.05
-            num_leaves: 15
-            n_estimators: 200
+            learning_rate: 0.02
+            num_leaves: 31
+            n_estimators: 3000
+            num_boost_round: 3000
+            early_stopping_rounds: 200
+            min_data_in_leaf: 20
+            lambda_l2: 0.5
            colsample_bytree: 0.8
            subsample: 0.8
            subsample_freq: 1
-            reg_alpha: 0.01
-            reg_lambda: 0.01
+            reg_alpha: 0.1
+            reg_lambda: 1.0
+            seed: 42

    dataset:
        class: DatasetH
@@ -52,7 +75,7 @@ task:
                    start_time: 2015-01-03
                    end_time: 2026-08-10
                    fit_start_time: 2015-01-03
-                    fit_end_time: 2026-01-03
+                    fit_end_time: 2025-09-01
                    freq: day
                    lake_root: "{{ LAKE }}"
                    market: US
@@ -80,24 +103,20 @@ task:
          kwargs: {}
        - class: SigAnaRecord
          module_path: qlib.workflow.record_temp
-          kwargs: { ana_long_short: true, ann_scaler: 252 }
+          kwargs:
+              ana_long_short: true
+              ann_scaler: 252
        - class: PortAnaRecord
          module_path: qlib.workflow.record_temp
          kwargs:
              config:
                  strategy:
-                      class: SignalQualityGateStrategy
-                      module_path: tac_qlib.contrib.strategy.signal_quality_gate
+                      class: TopkDropoutStrategy
+                      module_path: qlib.contrib.strategy
                      kwargs:
                          signal: "<PRED>"
-                          lake_root: "{{ LAKE }}"
-                          gate_topk: 10
-                          gate_lookback: 5
-                          gate_threshold: 0.5
-                          gate_start: "2015-01-03"
-                          gate_end: "2026-08-19"
                          topk: 10
-                          n_drop: 1
+                          n_drop: 2
                          only_tradable: true
                          risk_degree: 0.95
                  backtest:
@@ -1,44 +1,74 @@
-# Signal-quality gate backtest for 2025
+# -----------------------------------------------------------------------------
+# ISOLATION: multi-seed RankIC ensemble, ablate-B generic-only feature set.
+#
+# Isolates the ensemble effect on the SP-5d rank signal. Same panel, segments,
+# history (full backfilled 2016+) and feature set as the exp-9 ablate-B winner
+# (generic-only sp_* families: jump,har,trend,hurst,signature), but replaces the
+# single RankICLGBModel with a 5-seed RankICEnsembleLGBModel (42,7,2026,99,123)
+# that averages per-day predictions.
+#
+# Differs from exp-15 (tac-rd-rank-ensemble, mlflow exp 15) ONLY by dropping the
+# TA subset (rsi_14,roc_10,macd_hist,willr_14,atr_14) and the inter-asset xr_*
+# features, so any change vs exp-15 is attributable to the feature set alone,
+# and any change vs exp-9 is attributable to the ensemble + full history alone.
+#
+# Run:
+#   rd_run_workflow config_path=experiments/workflows/exp12_isolation_ensemble.yaml \
+#       experiment_name=tac-rd-rank-ensemble-isolated
+# -----------------------------------------------------------------------------
 {%- set LAKE = TAC_LAKE_DIR %}
 {%- set UNIVERSE = "SPY,QQQ,DIA,IWM,MDY,VTI,VOO,VEA,VWO,VT,EFA,EEM,TLT,IEF,SHY,AGG,BND,LQD,HYG,JNK,EMB,GLD,SLV,USO,UNG,DBA,DBC,XLK,XLF,XLE,XLV,XLI,XLY,XLP,XLU,XLB,XLRE,ARKK,SMH,SOXX,IBB,XBI,ITA,XAR,ICLN,TAN,FDN,IGV,ESPO,REM" %}
-{%- set SP_FIELDS = "sp_ret,sp_ou_zscore,sp_ou_half_life,sp_ou_revert,sp_hmm_p_regime1,sp_hmm_state,sp_jump_ratio,sp_jump_flag,sp_jump_tail,sp_max_move,sp_rv1,sp_rv5,sp_rv22,sp_vol_ratio_5_22,sp_vol_ratio_1_22,sp_trend_slope_5,sp_trend_slope_20,sp_trend_slope_60,sp_logp,sp_hurst_exponent,sp_sig_level1_lead,sp_sig_level1_lag,sp_sig_level2_lead_lag,sp_sig_level2_lag_lead" %}
-{# Gate computed on-the-fly from signal + lake close prices #}
+{%- set SP_FIELDS = "sp_ret,sp_jump_ratio,sp_jump_flag,sp_jump_tail,sp_max_move,sp_rv1,sp_rv5,sp_rv22,sp_vol_ratio_5_22,sp_vol_ratio_1_22,sp_trend_slope_5,sp_trend_slope_20,sp_trend_slope_60,sp_logp,sp_hurst_exponent,sp_sig_level1_lead,sp_sig_level1_lag,sp_sig_level2_lead_lag,sp_sig_level2_lag_lead" %}

 qlib_init:
    provider_uri: "{{ LAKE }}"
    region: us
    expression_cache: null
    dataset_cache: null
+
    calendar_provider:
        class: tac_qlib.data.providers.LakeCalendarProvider
-        kwargs: { lake_root: "{{ LAKE }}", market: US }
+        kwargs:
+            lake_root: "{{ LAKE }}"
+            market: US
    instrument_provider:
        class: tac_qlib.data.providers.LakeInstrumentProvider
-        kwargs: { lake_root: "{{ LAKE }}", market: US, markets: {} }
+        kwargs:
+            lake_root: "{{ LAKE }}"
+            market: US
+            markets: {}
    feature_provider:
        class: tac_qlib.data.providers.LakeFeatureProvider
-        kwargs: { lake_root: "{{ LAKE }}", market: US }
+        kwargs:
+            lake_root: "{{ LAKE }}"
+            market: US
+
    exp_manager:
        class: MLflowExpManager
        module_path: qlib.workflow.expm
        kwargs:
-            uri: "sqlite:///{{ LAKE }}/mlruns.db"
-            default_exp_name: "tac-rd-sq-gate-2025"
+            uri: "sqlite:///mlruns.db"
+            default_exp_name: "tac-rd-rank-ensemble-isolated"

 task:
    model:
-        class: LGBModel
-        module_path: qlib.contrib.model.gbdt
+        class: RankICEnsembleLGBModel
+        module_path: tac_qlib.contrib.model.rank_ensemble
        kwargs:
            loss: mse
-            learning_rate: 0.05
-            num_leaves: 15
-            n_estimators: 200
+            learning_rate: 0.02
+            num_leaves: 31
+            n_estimators: 3000
+            num_boost_round: 3000
+            early_stopping_rounds: 200
+            min_data_in_leaf: 20
+            lambda_l2: 0.5
            colsample_bytree: 0.8
            subsample: 0.8
            subsample_freq: 1
-            reg_alpha: 0.01
-            reg_lambda: 0.01
+            reg_alpha: 0.1
+            reg_lambda: 1.0
+            seeds: "42,7,2026,99,123"

    dataset:
        class: DatasetH
@@ -50,9 +80,9 @@ task:
                kwargs:
                    instruments: "{{ UNIVERSE }}"
                    start_time: 2015-01-03
-                    end_time: 2025-12-31
-                    fit_start_time: 2015-01-03
-                    fit_end_time: 2026-01-03
+                    end_time: 2026-08-14
+                    fit_start_time: 2016-01-04
+                    fit_end_time: 2025-09-01
                    freq: day
                    lake_root: "{{ LAKE }}"
                    market: US
@@ -70,9 +100,9 @@ task:
                        - class: Fillna
                          kwargs: {}
            segments:
-                train: [2015-01-03, 2025-09-01]
+                train: [2016-01-04, 2025-09-01]
                valid: [2025-09-03, 2026-01-03]
-                test: [2025-01-02, 2025-12-31]
+                test: [2026-01-04, 2026-08-10]

    record:
        - class: SignalRecord
@@ -80,29 +110,25 @@ task:
          kwargs: {}
        - class: SigAnaRecord
          module_path: qlib.workflow.record_temp
-          kwargs: { ana_long_short: true, ann_scaler: 252 }
+          kwargs:
+              ana_long_short: true
+              ann_scaler: 252
        - class: PortAnaRecord
          module_path: qlib.workflow.record_temp
          kwargs:
              config:
                  strategy:
-                      class: SignalQualityGateStrategy
-                      module_path: tac_qlib.contrib.strategy.signal_quality_gate
+                      class: TopkDropoutStrategy
+                      module_path: qlib.contrib.strategy
                      kwargs:
                          signal: "<PRED>"
-                          lake_root: "{{ LAKE }}"
-                          gate_topk: 10
-                          gate_lookback: 5
-                          gate_threshold: 0.5
-                          gate_start: "2015-01-03"
-                          gate_end: "2025-12-31"
                          topk: 10
-                          n_drop: 1
+                          n_drop: 2
                          only_tradable: true
                          risk_degree: 0.95
                  backtest:
-                      start_time: 2025-01-02
-                      end_time: 2025-12-31
+                      start_time: 2026-01-04
+                      end_time: 2026-08-10
                      account: 1000000
                      benchmark: SPY
                      exchange_kwargs:
@@ -0,0 +1,97 @@
+# Re-run of experiment 16 with validated family=ta and family=sp lake features.
+{%- set LAKE = TAC_LAKE_DIR %}
+{%- set UNIVERSE = "SPY,QQQ,DIA,IWM,MDY,VTI,VOO,VEA,VWO,VT,EFA,EEM,TLT,IEF,SHY,AGG,BND,LQD,HYG,JNK,EMB,GLD,SLV,USO,UNG,DBA,DBC,XLK,XLF,XLE,XLV,XLI,XLY,XLP,XLU,XLB,XLRE,ARKK,SMH,SOXX,IBB,XBI,ITA,XAR,ICLN,TAN,FDN,IGV,ESPO,REM" %}
+{%- set FEATURES = "$open,$high,$low,$close,$vwap,$volume,sma_5,sma_20,ema_12,ema_26,rsi_14,macd,macd_signal,macd_hist,bb_upper,bb_middle,bb_lower,atr_14,adx_14,sp_ret,sp_ou_half_life,sp_ou_revert,sp_ou_zscore,sp_hmm_p_regime1,sp_hmm_state,sp_jump_ratio,sp_jump_flag,sp_jump_tail,sp_max_down,sp_max_move,sp_max_up,sp_rv1,sp_rv5,sp_rv22,sp_rv_ac1,sp_rv_cv_22,sp_vol_ratio_1_22,sp_vol_ratio_5_22,sp_trend_slope_5,sp_trend_slope_20,sp_trend_slope_60,sp_logp,sp_hurst_exponent,sp_rskew_5,sp_rskew_22,sp_rkurt_5,sp_rkurt_22,sp_dsv_1,sp_dsv_5,sp_dsv_22,sp_dsv_ratio_1,sp_dsv_ratio_5,sp_dsv_ratio_22,sp_sig_level1_lead,sp_sig_level1_lag,sp_sig_level2_lead_lag,sp_sig_level2_lag_lead,sp_sig_level2_lead_lag_5,sp_sig_level2_lag_lead_5" %}
+
+qlib_init:
+    provider_uri: "{{ LAKE }}"
+    region: us
+    expression_cache: null
+    dataset_cache: null
+    calendar_provider:
+        class: tac_qlib.data.providers.LakeCalendarProvider
+        kwargs: { lake_root: "{{ LAKE }}", market: US }
+    instrument_provider:
+        class: tac_qlib.data.providers.LakeInstrumentProvider
+        kwargs: { lake_root: "{{ LAKE }}", market: US, markets: {} }
+    feature_provider:
+        class: tac_qlib.data.providers.LakeFeatureProvider
+        kwargs: { lake_root: "{{ LAKE }}", market: US }
+    exp_manager:
+        class: MLflowExpManager
+        module_path: qlib.workflow.expm
+        kwargs: { uri: "sqlite:///mlruns.db", default_exp_name: "tac-rd-exp16-db-ta-sp" }
+
+task:
+    model:
+        class: RankICEnsembleLGBModel
+        module_path: tac_qlib.contrib.model.rank_ensemble
+        kwargs:
+            loss: mse
+            learning_rate: 0.02
+            num_leaves: 31
+            n_estimators: 3000
+            num_boost_round: 3000
+            early_stopping_rounds: 200
+            min_data_in_leaf: 20
+            lambda_l2: 0.5
+            colsample_bytree: 0.8
+            subsample: 0.8
+            subsample_freq: 1
+            reg_alpha: 0.1
+            reg_lambda: 1.0
+            seeds: "42,7,2026,99,123"
+
+    dataset:
+        class: DatasetH
+        module_path: qlib.data.dataset
+        kwargs:
+            handler:
+                class: TACHandler
+                module_path: tac_qlib.contrib.data.handler
+                kwargs:
+                    instruments: "{{ UNIVERSE }}"
+                    start_time: 2015-01-03
+                    end_time: 2026-08-10
+                    fit_start_time: 2016-01-04
+                    fit_end_time: 2025-09-01
+                    freq: day
+                    lake_root: "{{ LAKE }}"
+                    market: US
+                    label: "Ref($close,-6)/Ref($close,-1)-1"
+                    feature_fields: "{{ FEATURES }}"
+                    infer_processors:
+                        - { class: DropAllNaN, kwargs: { fit_start_time: "2016-01-04", fit_end_time: "2025-09-01" } }
+                        - { class: ProcessInf, kwargs: {} }
+                        - { class: CSRankNorm, kwargs: {} }
+                        - { class: ZScoreNorm, kwargs: { fit_start_time: "2016-01-04", fit_end_time: "2025-09-01" } }
+                        - { class: Fillna, kwargs: {} }
+            segments:
+                train: [2016-01-04, 2025-09-01]
+                valid: [2025-09-03, 2026-01-03]
+                test: [2026-01-04, 2026-08-10]
+
+    record:
+        - { class: SignalRecord, module_path: qlib.workflow.record_temp, kwargs: {} }
+        - { class: SigAnaRecord, module_path: qlib.workflow.record_temp, kwargs: { ana_long_short: true, ann_scaler: 252 } }
+        - class: PortAnaRecord
+          module_path: qlib.workflow.record_temp
+          kwargs:
+              config:
+                  strategy:
+                      class: TopkDropoutStrategy
+                      module_path: qlib.contrib.strategy
+                      kwargs: { signal: "<PRED>", topk: 10, n_drop: 2, only_tradable: true, risk_degree: 0.95 }
+                  backtest:
+                      start_time: 2026-01-04
+                      end_time: 2026-08-10
+                      account: 1000000
+                      benchmark: SPY
+                      exchange_kwargs:
+                          codes: "{{ UNIVERSE }}"
+                          deal_price: $close
+                          freq: day
+                          open_cost: 0.0005
+                          close_cost: 0.0015
+                          min_cost: 5.0
+              risk_analysis_freq: 1d
@@ -0,0 +1,97 @@
+# General stochastic-process feature ablation: no TA, HMM, or OU fields.
+{%- set LAKE = TAC_LAKE_DIR %}
+{%- set UNIVERSE = "SPY,QQQ,DIA,IWM,MDY,VTI,VOO,VEA,VWO,VT,EFA,EEM,TLT,IEF,SHY,AGG,BND,LQD,HYG,JNK,EMB,GLD,SLV,USO,UNG,DBA,DBC,XLK,XLF,XLE,XLV,XLI,XLY,XLP,XLU,XLB,XLRE,ARKK,SMH,SOXX,IBB,XBI,ITA,XAR,ICLN,TAN,FDN,IGV,ESPO,REM" %}
+{%- set FEATURES = "$open,$high,$low,$close,$vwap,$volume,sp_ret,sp_jump_ratio,sp_jump_flag,sp_jump_tail,sp_max_down,sp_max_move,sp_max_up,sp_rv1,sp_rv5,sp_rv22,sp_rv_ac1,sp_rv_cv_22,sp_vol_ratio_1_22,sp_vol_ratio_5_22,sp_trend_slope_5,sp_trend_slope_20,sp_trend_slope_60,sp_logp,sp_hurst_exponent,sp_rskew_5,sp_rskew_22,sp_rkurt_5,sp_rkurt_22,sp_dsv_1,sp_dsv_5,sp_dsv_22,sp_dsv_ratio_1,sp_dsv_ratio_5,sp_dsv_ratio_22,sp_sig_level1_lead,sp_sig_level1_lag,sp_sig_level2_lead_lag,sp_sig_level2_lag_lead,sp_sig_level2_lead_lag_5,sp_sig_level2_lag_lead_5" %}
+
+qlib_init:
+    provider_uri: "{{ LAKE }}"
+    region: us
+    expression_cache: null
+    dataset_cache: null
+    calendar_provider:
+        class: tac_qlib.data.providers.LakeCalendarProvider
+        kwargs: { lake_root: "{{ LAKE }}", market: US }
+    instrument_provider:
+        class: tac_qlib.data.providers.LakeInstrumentProvider
+        kwargs: { lake_root: "{{ LAKE }}", market: US, markets: {} }
+    feature_provider:
+        class: tac_qlib.data.providers.LakeFeatureProvider
+        kwargs: { lake_root: "{{ LAKE }}", market: US }
+    exp_manager:
+        class: MLflowExpManager
+        module_path: qlib.workflow.expm
+        kwargs: { uri: "sqlite:///mlruns.db", default_exp_name: "tac-rd-exp22-stochastic-general" }
+
+task:
+    model:
+        class: RankICEnsembleLGBModel
+        module_path: tac_qlib.contrib.model.rank_ensemble
+        kwargs:
+            loss: mse
+            learning_rate: 0.02
+            num_leaves: 31
+            n_estimators: 3000
+            num_boost_round: 3000
+            early_stopping_rounds: 200
+            min_data_in_leaf: 20
+            lambda_l2: 0.5
+            colsample_bytree: 0.8
+            subsample: 0.8
+            subsample_freq: 1
+            reg_alpha: 0.1
+            reg_lambda: 1.0
+            seeds: "42,7,2026,99,123"
+
+    dataset:
+        class: DatasetH
+        module_path: qlib.data.dataset
+        kwargs:
+            handler:
+                class: TACHandler
+                module_path: tac_qlib.contrib.data.handler
+                kwargs:
+                    instruments: "{{ UNIVERSE }}"
+                    start_time: 2015-01-03
+                    end_time: 2026-08-10
+                    fit_start_time: 2016-01-04
+                    fit_end_time: 2025-09-01
+                    freq: day
+                    lake_root: "{{ LAKE }}"
+                    market: US
+                    label: "Ref($close,-6)/Ref($close,-1)-1"
+                    feature_fields: "{{ FEATURES }}"
+                    infer_processors:
+                        - { class: DropAllNaN, kwargs: { fit_start_time: "2016-01-04", fit_end_time: "2025-09-01" } }
+                        - { class: ProcessInf, kwargs: {} }
+                        - { class: CSRankNorm, kwargs: {} }
+                        - { class: ZScoreNorm, kwargs: { fit_start_time: "2016-01-04", fit_end_time: "2025-09-01" } }
+                        - { class: Fillna, kwargs: {} }
+            segments:
+                train: [2016-01-04, 2025-09-01]
+                valid: [2025-09-03, 2026-01-03]
+                test: [2026-01-04, 2026-08-10]
+
+    record:
+        - { class: SignalRecord, module_path: qlib.workflow.record_temp, kwargs: {} }
+        - { class: SigAnaRecord, module_path: qlib.workflow.record_temp, kwargs: { ana_long_short: true, ann_scaler: 252 } }
+        - class: PortAnaRecord
+          module_path: qlib.workflow.record_temp
+          kwargs:
+              config:
+                  strategy:
+                      class: TopkDropoutStrategy
+                      module_path: qlib.contrib.strategy
+                      kwargs: { signal: "<PRED>", topk: 10, n_drop: 2, only_tradable: true, risk_degree: 0.95 }
+                  backtest:
+                      start_time: 2026-01-04
+                      end_time: 2026-08-10
+                      account: 1000000
+                      benchmark: SPY
+                      exchange_kwargs:
+                          codes: "{{ UNIVERSE }}"
+                          deal_price: $close
+                          freq: day
+                          open_cost: 0.0005
+                          close_cost: 0.0015
+                          min_cost: 5.0
+              risk_analysis_freq: 1d
@@ -0,0 +1,97 @@
+# Exact compact stochastic feature set requested for a new run in MLflow exp 25.
+{%- set LAKE = TAC_LAKE_DIR %}
+{%- set UNIVERSE = "SPY,QQQ,DIA,IWM,MDY,VTI,VOO,VEA,VWO,VT,EFA,EEM,TLT,IEF,SHY,AGG,BND,LQD,HYG,JNK,EMB,GLD,SLV,USO,UNG,DBA,DBC,XLK,XLF,XLE,XLV,XLI,XLY,XLP,XLU,XLB,XLRE,ARKK,SMH,SOXX,IBB,XBI,ITA,XAR,ICLN,TAN,FDN,IGV,ESPO,REM" %}
+{%- set FEATURES = "$open,$high,$low,$close,$vwap,$volume,sp_ret,sp_jump_ratio,sp_jump_flag,sp_jump_tail,sp_max_move,sp_rv1,sp_rv5,sp_rv22,sp_vol_ratio_5_22,sp_vol_ratio_1_22,sp_trend_slope_5,sp_trend_slope_20,sp_trend_slope_60,sp_logp,sp_hurst_exponent,sp_sig_level1_lead,sp_sig_level1_lag,sp_sig_level2_lead_lag,sp_sig_level2_lag_lead" %}
+
+qlib_init:
+    provider_uri: "{{ LAKE }}"
+    region: us
+    expression_cache: null
+    dataset_cache: null
+    calendar_provider:
+        class: tac_qlib.data.providers.LakeCalendarProvider
+        kwargs: { lake_root: "{{ LAKE }}", market: US }
+    instrument_provider:
+        class: tac_qlib.data.providers.LakeInstrumentProvider
+        kwargs: { lake_root: "{{ LAKE }}", market: US, markets: {} }
+    feature_provider:
+        class: tac_qlib.data.providers.LakeFeatureProvider
+        kwargs: { lake_root: "{{ LAKE }}", market: US }
+    exp_manager:
+        class: MLflowExpManager
+        module_path: qlib.workflow.expm
+        kwargs: { uri: "sqlite:///mlruns.db", default_exp_name: "tac-rd-exp22-stochastic-general" }
+
+task:
+    model:
+        class: RankICEnsembleLGBModel
+        module_path: tac_qlib.contrib.model.rank_ensemble
+        kwargs:
+            loss: mse
+            learning_rate: 0.02
+            num_leaves: 31
+            n_estimators: 3000
+            num_boost_round: 3000
+            early_stopping_rounds: 200
+            min_data_in_leaf: 20
+            lambda_l2: 0.5
+            colsample_bytree: 0.8
+            subsample: 0.8
+            subsample_freq: 1
+            reg_alpha: 0.1
+            reg_lambda: 1.0
+            seeds: "42,7,2026,99,123"
+
+    dataset:
+        class: DatasetH
+        module_path: qlib.data.dataset
+        kwargs:
+            handler:
+                class: TACHandler
+                module_path: tac_qlib.contrib.data.handler
+                kwargs:
+                    instruments: "{{ UNIVERSE }}"
+                    start_time: 2015-01-03
+                    end_time: 2026-08-10
+                    fit_start_time: 2016-01-04
+                    fit_end_time: 2025-09-01
+                    freq: day
+                    lake_root: "{{ LAKE }}"
+                    market: US
+                    label: "Ref($close,-6)/Ref($close,-1)-1"
+                    feature_fields: "{{ FEATURES }}"
+                    infer_processors:
+                        - { class: DropAllNaN, kwargs: { fit_start_time: "2016-01-04", fit_end_time: "2025-09-01" } }
+                        - { class: ProcessInf, kwargs: {} }
+                        - { class: CSRankNorm, kwargs: {} }
+                        - { class: ZScoreNorm, kwargs: { fit_start_time: "2016-01-04", fit_end_time: "2025-09-01" } }
+                        - { class: Fillna, kwargs: {} }
+            segments:
+                train: [2016-01-04, 2025-09-01]
+                valid: [2025-09-03, 2026-01-03]
+                test: [2026-01-04, 2026-08-10]
+
+    record:
+        - { class: SignalRecord, module_path: qlib.workflow.record_temp, kwargs: {} }
+        - { class: SigAnaRecord, module_path: qlib.workflow.record_temp, kwargs: { ana_long_short: true, ann_scaler: 252 } }
+        - class: PortAnaRecord
+          module_path: qlib.workflow.record_temp
+          kwargs:
+              config:
+                  strategy:
+                      class: TopkDropoutStrategy
+                      module_path: qlib.contrib.strategy
+                      kwargs: { signal: "<PRED>", topk: 10, n_drop: 2, only_tradable: true, risk_degree: 0.95 }
+                  backtest:
+                      start_time: 2026-01-04
+                      end_time: 2026-08-10
+                      account: 1000000
+                      benchmark: SPY
+                      exchange_kwargs:
+                          codes: "{{ UNIVERSE }}"
+                          deal_price: $close
+                          freq: day
+                          open_cost: 0.0005
+                          close_cost: 0.0015
+                          min_cost: 5.0
+              risk_analysis_freq: 1d
@@ -0,0 +1,98 @@
+# Compact stochastic feature set with reduced turnover: n_drop=1 instead of 2.
+# Same setup as exp24 (compact baseline) but replacing the TopkDropout n_drop 2 with 1.
+{%- set LAKE = TAC_LAKE_DIR %}
+{%- set UNIVERSE = "SPY,QQQ,DIA,IWM,MDY,VTI,VOO,VEA,VWO,VT,EFA,EEM,TLT,IEF,SHY,AGG,BND,LQD,HYG,JNK,EMB,GLD,SLV,USO,UNG,DBA,DBC,XLK,XLF,XLE,XLV,XLI,XLY,XLP,XLU,XLB,XLRE,ARKK,SMH,SOXX,IBB,XBI,ITA,XAR,ICLN,TAN,FDN,IGV,ESPO,REM" %}
+{%- set FEATURES = "$open,$high,$low,$close,$vwap,$volume,sp_ret,sp_jump_ratio,sp_jump_flag,sp_jump_tail,sp_max_move,sp_rv1,sp_rv5,sp_rv22,sp_vol_ratio_5_22,sp_vol_ratio_1_22,sp_trend_slope_5,sp_trend_slope_20,sp_trend_slope_60,sp_logp,sp_hurst_exponent,sp_sig_level1_lead,sp_sig_level1_lag,sp_sig_level2_lead_lag,sp_sig_level2_lag_lead" %}
+
+qlib_init:
+    provider_uri: "{{ LAKE }}"
+    region: us
+    expression_cache: null
+    dataset_cache: null
+    calendar_provider:
+        class: tac_qlib.data.providers.LakeCalendarProvider
+        kwargs: { lake_root: "{{ LAKE }}", market: US }
+    instrument_provider:
+        class: tac_qlib.data.providers.LakeInstrumentProvider
+        kwargs: { lake_root: "{{ LAKE }}", market: US, markets: {} }
+    feature_provider:
+        class: tac_qlib.data.providers.LakeFeatureProvider
+        kwargs: { lake_root: "{{ LAKE }}", market: US }
+    exp_manager:
+        class: MLflowExpManager
+        module_path: qlib.workflow.expm
+        kwargs: { uri: "sqlite:///mlruns.db", default_exp_name: "tac-rd-exp22-stochastic-general" }
+
+task:
+    model:
+        class: RankICEnsembleLGBModel
+        module_path: tac_qlib.contrib.model.rank_ensemble
+        kwargs:
+            loss: mse
+            learning_rate: 0.02
+            num_leaves: 31
+            n_estimators: 3000
+            num_boost_round: 3000
+            early_stopping_rounds: 200
+            min_data_in_leaf: 20
+            lambda_l2: 0.5
+            colsample_bytree: 0.8
+            subsample: 0.8
+            subsample_freq: 1
+            reg_alpha: 0.1
+            reg_lambda: 1.0
+            seeds: "42,7,2026,99,123"
+
+    dataset:
+        class: DatasetH
+        module_path: qlib.data.dataset
+        kwargs:
+            handler:
+                class: TACHandler
+                module_path: tac_qlib.contrib.data.handler
+                kwargs:
+                    instruments: "{{ UNIVERSE }}"
+                    start_time: 2015-01-03
+                    end_time: 2026-08-10
+                    fit_start_time: 2016-01-04
+                    fit_end_time: 2025-09-01
+                    freq: day
+                    lake_root: "{{ LAKE }}"
+                    market: US
+                    label: "Ref($close,-6)/Ref($close,-1)-1"
+                    feature_fields: "{{ FEATURES }}"
+                    infer_processors:
+                        - { class: DropAllNaN, kwargs: { fit_start_time: "2016-01-04", fit_end_time: "2025-09-01" } }
+                        - { class: ProcessInf, kwargs: {} }
+                        - { class: CSRankNorm, kwargs: {} }
+                        - { class: ZScoreNorm, kwargs: { fit_start_time: "2016-01-04", fit_end_time: "2025-09-01" } }
+                        - { class: Fillna, kwargs: {} }
+            segments:
+                train: [2016-01-04, 2025-09-01]
+                valid: [2025-09-03, 2026-01-03]
+                test: [2026-01-04, 2026-08-10]
+
+    record:
+        - { class: SignalRecord, module_path: qlib.workflow.record_temp, kwargs: {} }
+        - { class: SigAnaRecord, module_path: qlib.workflow.record_temp, kwargs: { ana_long_short: true, ann_scaler: 252 } }
+        - class: PortAnaRecord
+          module_path: qlib.workflow.record_temp
+          kwargs:
+              config:
+                  strategy:
+                      class: TopkDropoutStrategy
+                      module_path: qlib.contrib.strategy
+                      kwargs: { signal: "<PRED>", topk: 10, n_drop: 1, only_tradable: true, risk_degree: 0.95 }
+                  backtest:
+                      start_time: 2026-01-04
+                      end_time: 2026-08-10
+                      account: 1000000
+                      benchmark: SPY
+                      exchange_kwargs:
+                          codes: "{{ UNIVERSE }}"
+                          deal_price: $close
+                          freq: day
+                          open_cost: 0.0005
+                          close_cost: 0.0015
+                          min_cost: 5.0
+              risk_analysis_freq: 1d
@@ -0,0 +1,99 @@
+# M2 isolation run: base compact set + risk-adjusted drift sp_sharpe_22.
+# Exact copy of exp26 (reference: expId=25 run=21afc6afdb674a399b59dd76c97628ce)
+# except feature_fields. 5-seed ensemble.
+{%- set LAKE = TAC_LAKE_DIR %}
+{%- set UNIVERSE = "SPY,QQQ,DIA,IWM,MDY,VTI,VOO,VEA,VWO,VT,EFA,EEM,TLT,IEF,SHY,AGG,BND,LQD,HYG,JNK,EMB,GLD,SLV,USO,UNG,DBA,DBC,XLK,XLF,XLE,XLV,XLI,XLY,XLP,XLU,XLB,XLRE,ARKK,SMH,SOXX,IBB,XBI,ITA,XAR,ICLN,TAN,FDN,IGV,ESPO,REM" %}
+{%- set FEATURES = "$open,$high,$low,$close,$vwap,$volume,sp_ret,sp_jump_ratio,sp_jump_flag,sp_jump_tail,sp_max_move,sp_rv1,sp_rv5,sp_rv22,sp_vol_ratio_5_22,sp_vol_ratio_1_22,sp_trend_slope_5,sp_trend_slope_20,sp_trend_slope_60,sp_logp,sp_hurst_exponent,sp_sig_level1_lead,sp_sig_level1_lag,sp_sig_level2_lead_lag,sp_sig_level2_lag_lead,sp_sharpe_22" %}
+
+qlib_init:
+    provider_uri: "{{ LAKE }}"
+    region: us
+    expression_cache: null
+    dataset_cache: null
+    calendar_provider:
+        class: tac_qlib.data.providers.LakeCalendarProvider
+        kwargs: { lake_root: "{{ LAKE }}", market: US }
+    instrument_provider:
+        class: tac_qlib.data.providers.LakeInstrumentProvider
+        kwargs: { lake_root: "{{ LAKE }}", market: US, markets: {} }
+    feature_provider:
+        class: tac_qlib.data.providers.LakeFeatureProvider
+        kwargs: { lake_root: "{{ LAKE }}", market: US }
+    exp_manager:
+        class: MLflowExpManager
+        module_path: qlib.workflow.expm
+        kwargs: { uri: "sqlite:///mlruns.db", default_exp_name: "tac-rd-exp30-m2-sharpe" }
+
+task:
+    model:
+        class: RankICEnsembleLGBModel
+        module_path: tac_qlib.contrib.model.rank_ensemble
+        kwargs:
+            loss: mse
+            learning_rate: 0.02
+            num_leaves: 31
+            n_estimators: 3000
+            num_boost_round: 3000
+            early_stopping_rounds: 200
+            min_data_in_leaf: 20
+            lambda_l2: 0.5
+            colsample_bytree: 0.8
+            subsample: 0.8
+            subsample_freq: 1
+            reg_alpha: 0.1
+            reg_lambda: 1.0
+            seeds: "42,7,2026,99,123"
+
+    dataset:
+        class: DatasetH
+        module_path: qlib.data.dataset
+        kwargs:
+            handler:
+                class: TACHandler
+                module_path: tac_qlib.contrib.data.handler
+                kwargs:
+                    instruments: "{{ UNIVERSE }}"
+                    start_time: 2015-01-03
+                    end_time: 2026-08-10
+                    fit_start_time: 2016-01-04
+                    fit_end_time: 2025-09-01
+                    freq: day
+                    lake_root: "{{ LAKE }}"
+                    market: US
+                    label: "Ref($close,-6)/Ref($close,-1)-1"
+                    feature_fields: "{{ FEATURES }}"
+                    infer_processors:
+                        - { class: DropAllNaN, kwargs: { fit_start_time: "2016-01-04", fit_end_time: "2025-09-01" } }
+                        - { class: ProcessInf, kwargs: {} }
+                        - { class: CSRankNorm, kwargs: {} }
+                        - { class: ZScoreNorm, kwargs: { fit_start_time: "2016-01-04", fit_end_time: "2025-09-01" } }
+                        - { class: Fillna, kwargs: {} }
+            segments:
+                train: [2016-01-04, 2025-09-01]
+                valid: [2025-09-03, 2026-01-03]
+                test: [2026-01-04, 2026-08-10]
+
+    record:
+        - { class: SignalRecord, module_path: qlib.workflow.record_temp, kwargs: {} }
+        - { class: SigAnaRecord, module_path: qlib.workflow.record_temp, kwargs: { ana_long_short: true, ann_scaler: 252 } }
+        - class: PortAnaRecord
+          module_path: qlib.workflow.record_temp
+          kwargs:
+              config:
+                  strategy:
+                      class: TopkDropoutStrategy
+                      module_path: qlib.contrib.strategy
+                      kwargs: { signal: "<PRED>", topk: 10, n_drop: 1, only_tradable: true, risk_degree: 0.95 }
+                  backtest:
+                      start_time: 2026-01-04
+                      end_time: 2026-08-10
+                      account: 1000000
+                      benchmark: SPY
+                      exchange_kwargs:
+                          codes: "{{ UNIVERSE }}"
+                          deal_price: $close
+                          freq: day
+                          open_cost: 0.0005
+                          close_cost: 0.0015
+                          min_cost: 5.0
+              risk_analysis_freq: 1d
@@ -0,0 +1,99 @@
+# M3 isolation run: base compact set + GARCH(1,1) vol-regime trio.
+# Exact copy of exp26 (reference: expId=25 run=21afc6afdb674a399b59dd76c97628ce)
+# except feature_fields. 5-seed ensemble.
+{%- set LAKE = TAC_LAKE_DIR %}
+{%- set UNIVERSE = "SPY,QQQ,DIA,IWM,MDY,VTI,VOO,VEA,VWO,VT,EFA,EEM,TLT,IEF,SHY,AGG,BND,LQD,HYG,JNK,EMB,GLD,SLV,USO,UNG,DBA,DBC,XLK,XLF,XLE,XLV,XLI,XLY,XLP,XLU,XLB,XLRE,ARKK,SMH,SOXX,IBB,XBI,ITA,XAR,ICLN,TAN,FDN,IGV,ESPO,REM" %}
+{%- set FEATURES = "$open,$high,$low,$close,$vwap,$volume,sp_ret,sp_jump_ratio,sp_jump_flag,sp_jump_tail,sp_max_move,sp_rv1,sp_rv5,sp_rv22,sp_vol_ratio_5_22,sp_vol_ratio_1_22,sp_trend_slope_5,sp_trend_slope_20,sp_trend_slope_60,sp_logp,sp_hurst_exponent,sp_sig_level1_lead,sp_sig_level1_lag,sp_sig_level2_lead_lag,sp_sig_level2_lag_lead,sp_garch_cond_var,sp_garch_persistence,sp_garch_std_resid" %}
+
+qlib_init:
+    provider_uri: "{{ LAKE }}"
+    region: us
+    expression_cache: null
+    dataset_cache: null
+    calendar_provider:
+        class: tac_qlib.data.providers.LakeCalendarProvider
+        kwargs: { lake_root: "{{ LAKE }}", market: US }
+    instrument_provider:
+        class: tac_qlib.data.providers.LakeInstrumentProvider
+        kwargs: { lake_root: "{{ LAKE }}", market: US, markets: {} }
+    feature_provider:
+        class: tac_qlib.data.providers.LakeFeatureProvider
+        kwargs: { lake_root: "{{ LAKE }}", market: US }
+    exp_manager:
+        class: MLflowExpManager
+        module_path: qlib.workflow.expm
+        kwargs: { uri: "sqlite:///mlruns.db", default_exp_name: "tac-rd-exp31-m3-garch" }
+
+task:
+    model:
+        class: RankICEnsembleLGBModel
+        module_path: tac_qlib.contrib.model.rank_ensemble
+        kwargs:
+            loss: mse
+            learning_rate: 0.02
+            num_leaves: 31
+            n_estimators: 3000
+            num_boost_round: 3000
+            early_stopping_rounds: 200
+            min_data_in_leaf: 20
+            lambda_l2: 0.5
+            colsample_bytree: 0.8
+            subsample: 0.8
+            subsample_freq: 1
+            reg_alpha: 0.1
+            reg_lambda: 1.0
+            seeds: "42,7,2026,99,123"
+
+    dataset:
+        class: DatasetH
+        module_path: qlib.data.dataset
+        kwargs:
+            handler:
+                class: TACHandler
+                module_path: tac_qlib.contrib.data.handler
+                kwargs:
+                    instruments: "{{ UNIVERSE }}"
+                    start_time: 2015-01-03
+                    end_time: 2026-08-10
+                    fit_start_time: 2016-01-04
+                    fit_end_time: 2025-09-01
+                    freq: day
+                    lake_root: "{{ LAKE }}"
+                    market: US
+                    label: "Ref($close,-6)/Ref($close,-1)-1"
+                    feature_fields: "{{ FEATURES }}"
+                    infer_processors:
+                        - { class: DropAllNaN, kwargs: { fit_start_time: "2016-01-04", fit_end_time: "2025-09-01" } }
+                        - { class: ProcessInf, kwargs: {} }
+                        - { class: CSRankNorm, kwargs: {} }
+                        - { class: ZScoreNorm, kwargs: { fit_start_time: "2016-01-04", fit_end_time: "2025-09-01" } }
+                        - { class: Fillna, kwargs: {} }
+            segments:
+                train: [2016-01-04, 2025-09-01]
+                valid: [2025-09-03, 2026-01-03]
+                test: [2026-01-04, 2026-08-10]
+
+    record:
+        - { class: SignalRecord, module_path: qlib.workflow.record_temp, kwargs: {} }
+        - { class: SigAnaRecord, module_path: qlib.workflow.record_temp, kwargs: { ana_long_short: true, ann_scaler: 252 } }
+        - class: PortAnaRecord
+          module_path: qlib.workflow.record_temp
+          kwargs:
+              config:
+                  strategy:
+                      class: TopkDropoutStrategy
+                      module_path: qlib.contrib.strategy
+                      kwargs: { signal: "<PRED>", topk: 10, n_drop: 1, only_tradable: true, risk_degree: 0.95 }
+                  backtest:
+                      start_time: 2026-01-04
+                      end_time: 2026-08-10
+                      account: 1000000
+                      benchmark: SPY
+                      exchange_kwargs:
+                          codes: "{{ UNIVERSE }}"
+                          deal_price: $close
+                          freq: day
+                          open_cost: 0.0005
+                          close_cost: 0.0015
+                          min_cost: 5.0
+              risk_analysis_freq: 1d
Author	SHA1	Message	Date
zhaoli	46ecbbf0dc	start experiment 80 (exp/80-scheduled-algo-retrain-on-2026-09-22-tac)	2026-09-23 12:12:46 +00:00
zhaoli	a37fc0869f	finish experiment 79 (exp/79-scheduled-algo-retrain-on-2026-09-21-tac)	2026-09-22 12:50:58 +00:00
zhaoli	05ccf47226	exp 79: snapshot custom qlib code	2026-09-22 12:18:06 +00:00
zhaoli	080cd73da3	start experiment 79 (exp/79-scheduled-algo-retrain-on-2026-09-21-tac)	2026-09-22 12:18:00 +00:00
zhaoli	278d9830af	start experiment 78 (exp/78-scheduled-algo-retrain-on-2026-09-18-tac)	2026-09-21 12:02:16 +00:00
zhaoli	f9c8fbe945	start experiment 69 (exp/69-scheduled-algo-retrain-on-2026-09-03-tac)	2026-09-04 12:01:21 +00:00
zhaoli	ce2e0c1a7c	start experiment 67 (exp/67-scheduled-algo-retrain-on-2026-09-01-tac)	2026-09-02 12:03:33 +00:00
zhaoli	f1bd6d99c3	finish experiment 63 (exp/63-scheduled-algo-retrain-on-2026-08-26-tac)	2026-08-27 12:21:58 +00:00
zhaoli	b872f64ce6	start experiment 63 (exp/63-scheduled-algo-retrain-on-2026-08-26-tac)	2026-08-27 12:02:21 +00:00
zhaoli	66cf0a149f	finish experiment 33 (exp/33-q01-m2-reproduction-add-spsharpe22-to-th)	2026-08-19 22:40:28 +00:00
zhaoli	6a1b05db2e	add q01 m2 repro workflow	2026-08-19 22:30:10 +00:00
zhaoli	c7fc73180c	start experiment 33 (exp/33-q01-m2-reproduction-add-spsharpe22-to-th)	2026-08-19 22:05:08 +00:00
zhaoli	483a86e47f	finish experiment 31 (exp/31-isolation-run-m3-does-adding-garch11-vol)	2026-08-18 21:41:40 +00:00
zhaoli	e660b4f2dd	finish experiment 30 (exp/30-isolation-run-m2-does-adding-risk-adjust)	2026-08-18 20:13:22 +00:00
zhaoli	4e1debccba	M2 isolation: base + sp_sharpe_22 (trace 30)	2026-08-18 15:54:24 +00:00
zhaoli	cb18467a6d	start experiment 30 (exp/30-isolation-run-m2-does-adding-risk-adjust)	2026-08-18 15:52:33 +00:00
zhaoli	894ac260a6	finish experiment 26 (exp/26-test-whether-reducing-topkdropout-daily)	2026-08-18 14:05:47 +00:00
zhaoli	c455000a1e	exp26: compact stochastic, n_drop 1 (workflow only)	2026-08-18 12:10:29 +00:00
zhaoli	5c7b2265f4	start experiment 26 (exp/26-test-whether-reducing-topkdropout-daily)	2026-08-18 10:03:21 +00:00
zhaoli	c724682f3a	finish experiment 24 (exp/24-run-the-rankic-ensemble-in-mlflow-experi)	2026-08-18 09:00:11 +00:00
zhaoli	59e733d88d	exp 24: add exact compact stochastic feature workflow	2026-08-18 08:03:37 +00:00
zhaoli	1fcadbfe86	start experiment 24 (exp/24-run-the-rankic-ensemble-in-mlflow-experi)	2026-08-18 08:02:27 +00:00
zhaoli	a718424340	finish experiment 23 (exp/23-test-whether-the-5-day-rankic-ensemble-i)	2026-08-18 07:58:20 +00:00
zhaoli	adf0bfa812	exp 23: add general stochastic feature ablation workflow	2026-08-18 07:21:33 +00:00
zhaoli	b5054ccc25	start experiment 23 (exp/23-test-whether-the-5-day-rankic-ensemble-i)	2026-08-18 07:20:27 +00:00
zhaoli	3d845306fe	finish experiment 22 (exp/22-re-run-experiment-16s-5-day-rankic-ensem)	2026-08-18 07:17:36 +00:00
zhaoli	1075525d6e	exp 22: add validated TA SP ensemble workflow	2026-08-18 06:38:55 +00:00
zhaoli	63c1ea763e	start experiment 22 (exp/22-re-run-experiment-16s-5-day-rankic-ensem)	2026-08-18 06:38:28 +00:00
zhaoli	2c2684b103	start experiment 16 (16-scheduled-algo-retrain-on-20260814-tacrd)	2026-08-15 05:14:48 +00:00
zhaoli	32477c7bb8	exp 12: isolation ensemble workflow yaml (ablate-B features + full history) + finish record	2026-08-14 05:37:13 +00:00
zhaoli	e657c58758	start exp 9 (sp5d-feature-family-ablation): baseline all-24 + generic-only 19 workflow YAMLs	2026-08-13 04:32:30 +00:00
				`@@ -0,0 +1 @@`
				`from .tradeac_exchange import TradeACExchange`