feat: 리밸런싱 엔진 V1 + GAS 버그 수정 (2026-06-13)
주요 변경: - tools/build_rebalance_engine_v1.py: REBALANCE_ENGINE_V1 신규 * account_snapshot 직접 합산(_build_snap_position_map) → 소수주 분리 행 병합 * 레짐 소스 macro.REGIME_PRELIM 최우선 (GAS 와 동일) - src/gas_adapter_parts/gdf_06_rebalance.gs: runRebalanceSheet_() 신규 * Logger.log / getSpreadsheet_() 로 run_all 연동 수정 - src/gas_adapter_parts/gdc_01_fetch_fundamentals.gs * _mergePositionRecord_(): 소수주 중복 행 합산 신규 * parseInt → parseFloat (qty, availQty) - src/gas_adapter_parts/gdf_01_price_metrics.gs * 미보유 종목 SELL_READY → WATCH_EXIT_SIGNAL - spec/41_release_dag.yaml: build_rebalance_sheet 노드 추가 (step_count 63) - spec/51_formula_lifecycle_registry.yaml: REBALANCE_ENGINE_V1 등록 Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
This commit is contained in:
@@ -0,0 +1,812 @@
|
||||
#!/usr/bin/env python3
|
||||
"""
|
||||
compute_formula_outputs.py
|
||||
───────────────────────────────────────────────────────────────────────────────
|
||||
Python 공식 계산 엔진
|
||||
|
||||
GAS가 아직 구현하지 않은 핵심 공식을 Python으로 결정론적으로 계산한다.
|
||||
동일 입력 → 동일 출력. 텍스트 판단 없음. 수치만 출력.
|
||||
|
||||
계산 대상:
|
||||
- VELOCITY_V1 : velocity_1d, velocity_5d
|
||||
- PROFIT_LOCK_STAGE : profit_pct → 단계 분류
|
||||
- ANTI_CHASING_VELOCITY_V1 : velocity_1d 기반 뒷박 차단
|
||||
- PULLBACK_ENTRY_TRIGGER_V1 : MA20 기반 눌림목 기준가
|
||||
- SELL_PRICE_SANITY_V1 : 매도가 역전·비현실가 검증
|
||||
- TICK_NORMALIZER_V1 : KRX 호가 단위 정규화
|
||||
- CASH_RECOVERY_OPTIMIZER_V1 : 최소 주식가치 훼손 매도조합
|
||||
- PROFIT_RATCHET_TIERED_V2 : APEX_SUPER ATR×1.2 trailing
|
||||
|
||||
사용법:
|
||||
python tools/compute_formula_outputs.py [GatherTradingData.json]
|
||||
python tools/compute_formula_outputs.py --output computed_harness.json
|
||||
"""
|
||||
|
||||
from __future__ import annotations
|
||||
|
||||
import json
|
||||
import math
|
||||
import sys
|
||||
from pathlib import Path
|
||||
|
||||
ROOT = Path(__file__).resolve().parents[2]
|
||||
if str(ROOT) not in sys.path:
|
||||
sys.path.insert(0, str(ROOT))
|
||||
|
||||
from src.quant_engine.exit_decisions import (
|
||||
compute_cash_shortfall_harness as _compute_cash_shortfall_harness,
|
||||
compute_dynamic_heat_thresholds as _compute_dynamic_heat_thresholds,
|
||||
compute_final_decision as _compute_final_decision,
|
||||
compute_sell_decision as _compute_sell_decision,
|
||||
compute_timing_decision as _compute_timing_decision,
|
||||
compute_stop_action_ladder as _compute_stop_action_ladder,
|
||||
compute_stop_price_core as _compute_stop_price_core,
|
||||
normalize_tick as _normalize_tick,
|
||||
)
|
||||
|
||||
# Windows cp949 터미널 호환
|
||||
if sys.stdout.encoding and sys.stdout.encoding.lower() not in ("utf-8", "utf8"):
|
||||
sys.stdout = open(sys.stdout.fileno(), mode="w", encoding="utf-8", buffering=1)
|
||||
if sys.stderr.encoding and sys.stderr.encoding.lower() not in ("utf-8", "utf8"):
|
||||
sys.stderr = open(sys.stderr.fileno(), mode="w", encoding="utf-8", buffering=1)
|
||||
|
||||
SEP = "=" * 70
|
||||
|
||||
|
||||
# ─────────────────────────────────────────────────────────────────────────────
|
||||
# KRX 호가 단위 테이블 (TICK_NORMALIZER_V1)
|
||||
# ─────────────────────────────────────────────────────────────────────────────
|
||||
KRX_TICK_TABLE = [
|
||||
( 2_000, 1),
|
||||
( 5_000, 5),
|
||||
( 20_000, 10),
|
||||
( 50_000, 50),
|
||||
( 200_000, 100),
|
||||
( 500_000, 500),
|
||||
(math.inf, 1000),
|
||||
]
|
||||
|
||||
def krx_tick_unit(price: float) -> int:
|
||||
for threshold, tick in KRX_TICK_TABLE:
|
||||
if price < threshold:
|
||||
return tick
|
||||
return 1000
|
||||
|
||||
def normalize_tick(price: float) -> int:
|
||||
return _normalize_tick(price)
|
||||
|
||||
|
||||
# ─────────────────────────────────────────────────────────────────────────────
|
||||
# PROFIT_LOCK_STAGE 분류기
|
||||
# ─────────────────────────────────────────────────────────────────────────────
|
||||
def classify_profit_lock_stage(profit_pct: float) -> str:
|
||||
if profit_pct >= 60:
|
||||
return "APEX_SUPER"
|
||||
elif profit_pct >= 40:
|
||||
return "APEX_TRAILING"
|
||||
elif profit_pct >= 30:
|
||||
return "PROFIT_LOCK_30"
|
||||
elif profit_pct >= 20:
|
||||
return "PROFIT_LOCK_20"
|
||||
elif profit_pct >= 10:
|
||||
return "PROFIT_LOCK_10"
|
||||
elif profit_pct >= 0:
|
||||
return "BREAKEVEN_RATCHET"
|
||||
else:
|
||||
return "NORMAL"
|
||||
|
||||
|
||||
# ─────────────────────────────────────────────────────────────────────────────
|
||||
# PROFIT_RATCHET_TIERED_V2
|
||||
# ─────────────────────────────────────────────────────────────────────────────
|
||||
def compute_trailing_stop_v2(
|
||||
profit_pct: float,
|
||||
highest_close: float,
|
||||
atr20: float,
|
||||
ratchet_stop: float | None,
|
||||
average_cost: float,
|
||||
) -> dict:
|
||||
stage = classify_profit_lock_stage(profit_pct)
|
||||
ratchet_stop = ratchet_stop or average_cost
|
||||
|
||||
if stage == "APEX_SUPER":
|
||||
raw = highest_close - 1.2 * atr20
|
||||
trailing_stop = normalize_tick(max(ratchet_stop, raw))
|
||||
tp_action = "강제 10% 익절 권고"
|
||||
elif stage == "APEX_TRAILING":
|
||||
raw = highest_close - 1.5 * atr20
|
||||
trailing_stop = normalize_tick(max(ratchet_stop, raw))
|
||||
tp_action = "부분익절 검토"
|
||||
elif stage in ("PROFIT_LOCK_30", "PROFIT_LOCK_20"):
|
||||
raw = highest_close - 2.0 * atr20
|
||||
trailing_stop = normalize_tick(max(ratchet_stop, raw))
|
||||
tp_action = "래칫 유지"
|
||||
else:
|
||||
trailing_stop = None
|
||||
tp_action = "적용 안함"
|
||||
|
||||
return {
|
||||
"ratchet_stage_v2": stage,
|
||||
"auto_trailing_stop_v2": trailing_stop,
|
||||
"tp_ladder_action": tp_action,
|
||||
"apex_super_active": stage == "APEX_SUPER",
|
||||
}
|
||||
|
||||
|
||||
def compute_stop_price_core(entry_price: float | None, atr20: float | None, current_price: float | None) -> dict:
|
||||
return _compute_stop_price_core(entry_price, atr20, current_price)
|
||||
|
||||
|
||||
def compute_stop_action_ladder(context: dict) -> dict:
|
||||
return _compute_stop_action_ladder(context)
|
||||
|
||||
|
||||
def compute_dynamic_heat_thresholds(regime: str) -> dict:
|
||||
return _compute_dynamic_heat_thresholds(regime)
|
||||
|
||||
|
||||
def compute_cash_shortfall_harness(as_result: dict, total_asset: float, cash_floor_info: dict, mrs_score: float) -> dict:
|
||||
return _compute_cash_shortfall_harness(as_result, total_asset, cash_floor_info, mrs_score)
|
||||
|
||||
|
||||
def compute_timing_decision(ctx: dict) -> dict:
|
||||
return _compute_timing_decision(ctx)
|
||||
|
||||
|
||||
def compute_sell_decision(ctx: dict) -> dict:
|
||||
return _compute_sell_decision(ctx)
|
||||
|
||||
|
||||
def compute_final_decision(ctx: dict) -> dict:
|
||||
return _compute_final_decision(ctx)
|
||||
|
||||
|
||||
# ─────────────────────────────────────────────────────────────────────────────
|
||||
# ANTI_CHASING_VELOCITY_V1
|
||||
# ─────────────────────────────────────────────────────────────────────────────
|
||||
def compute_anti_chasing(velocity_1d: float) -> dict:
|
||||
if velocity_1d >= 3.0:
|
||||
verdict = "BLOCK_CHASE"
|
||||
status = "BLOCKED"
|
||||
elif velocity_1d >= 1.5:
|
||||
verdict = "PULLBACK_WAIT"
|
||||
status = "WAIT"
|
||||
else:
|
||||
verdict = "CLEAR"
|
||||
status = "PASS"
|
||||
return {
|
||||
"anti_chasing_verdict": verdict,
|
||||
"anti_chasing_velocity_status": status,
|
||||
"velocity_1d_input": round(velocity_1d, 4),
|
||||
}
|
||||
|
||||
|
||||
# ─────────────────────────────────────────────────────────────────────────────
|
||||
# PULLBACK_ENTRY_TRIGGER_V1
|
||||
# ─────────────────────────────────────────────────────────────────────────────
|
||||
def compute_pullback_trigger(close: float, ma20: float, atr20: float) -> dict:
|
||||
trigger_price = normalize_tick(ma20 - 0.5 * atr20)
|
||||
upper_band = ma20 * 1.03
|
||||
|
||||
if close <= upper_band:
|
||||
verdict = "PULLBACK_ZONE"
|
||||
state = "PASS"
|
||||
else:
|
||||
verdict = "ABOVE_PULLBACK_ZONE"
|
||||
state = "BLOCKED"
|
||||
|
||||
return {
|
||||
"pullback_entry_verdict": verdict,
|
||||
"pullback_state": state,
|
||||
"pullback_entry_trigger_price": trigger_price,
|
||||
"pullback_upper_band": normalize_tick(upper_band),
|
||||
}
|
||||
|
||||
|
||||
# ─────────────────────────────────────────────────────────────────────────────
|
||||
# SELL_PRICE_SANITY_V1
|
||||
# ─────────────────────────────────────────────────────────────────────────────
|
||||
def check_sell_price_sanity(
|
||||
sell_limit_price: float,
|
||||
stop_loss_price: float | None,
|
||||
prev_close: float,
|
||||
ticker: str = "",
|
||||
) -> dict:
|
||||
issues: list[str] = []
|
||||
status = "PASS"
|
||||
|
||||
if stop_loss_price is not None and sell_limit_price < stop_loss_price:
|
||||
issues.append(
|
||||
f"INVALID_PRICE_INVERSION: sell={sell_limit_price:,} < stop={stop_loss_price:,}"
|
||||
)
|
||||
status = "INVALID_PRICE_INVERSION"
|
||||
|
||||
upper_limit = prev_close * 1.30
|
||||
if sell_limit_price > upper_limit:
|
||||
issues.append(
|
||||
f"INVALID_UNREALISTIC_PRICE: sell={sell_limit_price:,} > prev_close*1.30={upper_limit:,.0f}"
|
||||
)
|
||||
if status == "PASS":
|
||||
status = "INVALID_UNREALISTIC_PRICE"
|
||||
|
||||
tick_unit = krx_tick_unit(sell_limit_price)
|
||||
if sell_limit_price % tick_unit != 0:
|
||||
corrected = normalize_tick(sell_limit_price)
|
||||
issues.append(
|
||||
f"INVALID_TICK: sell={sell_limit_price:,} 호가단위={tick_unit}원 → 정규화={corrected:,}"
|
||||
)
|
||||
if status == "PASS":
|
||||
status = "INVALID_TICK"
|
||||
|
||||
return {
|
||||
"sell_price_sanity_status": status,
|
||||
"sell_price_sanity_issues": issues,
|
||||
"hts_allowed": status == "PASS",
|
||||
"shadow_ledger": status != "PASS",
|
||||
"ticker": ticker,
|
||||
}
|
||||
|
||||
|
||||
# ─────────────────────────────────────────────────────────────────────────────
|
||||
# CASH_RECOVERY_OPTIMIZER_V1
|
||||
# ─────────────────────────────────────────────────────────────────────────────
|
||||
def compute_cash_recovery_optimizer(
|
||||
sell_candidates: list[dict], # sorted by h2_priority_rank asc
|
||||
cash_shortfall_min_krw: float,
|
||||
) -> dict:
|
||||
plan: list[dict] = []
|
||||
cumulative_krw = 0.0
|
||||
|
||||
for cand in sell_candidates:
|
||||
if cumulative_krw >= cash_shortfall_min_krw:
|
||||
break
|
||||
ticker = cand.get("Ticker", "")
|
||||
name = cand.get("Name", "")
|
||||
qty = cand.get("Sell_Qty") or 0
|
||||
limit_price = cand.get("Sell_Limit_Price") or cand.get("current_price", 0)
|
||||
preserve_ratio = cand.get("Cash_Preserve_Ratio", 100)
|
||||
style = cand.get("Cash_Preserve_Style", "FULL")
|
||||
|
||||
if qty and limit_price:
|
||||
expected_krw = qty * limit_price * (preserve_ratio / 100)
|
||||
else:
|
||||
expected_krw = 0
|
||||
|
||||
plan.append({
|
||||
"ticker": ticker,
|
||||
"name": name,
|
||||
"qty": qty,
|
||||
"limit_price": normalize_tick(limit_price) if limit_price else None,
|
||||
"preserve_style": style,
|
||||
"preserve_ratio": preserve_ratio,
|
||||
"expected_krw": round(expected_krw),
|
||||
})
|
||||
cumulative_krw += expected_krw
|
||||
|
||||
shortfall_met = cumulative_krw >= cash_shortfall_min_krw
|
||||
return {
|
||||
"cash_recovery_plan_json": {
|
||||
"sell_sequence": plan,
|
||||
"expected_total_krw": round(cumulative_krw),
|
||||
"cash_shortfall_min_krw": cash_shortfall_min_krw,
|
||||
"shortfall_met": shortfall_met,
|
||||
"items_needed": len(plan),
|
||||
}
|
||||
}
|
||||
|
||||
|
||||
# ─────────────────────────────────────────────────────────────────────────────
|
||||
# 메인 계산 루틴
|
||||
# ─────────────────────────────────────────────────────────────────────────────
|
||||
def main() -> int:
|
||||
output_path: Path | None = None
|
||||
args = sys.argv[1:]
|
||||
if "--output" in args:
|
||||
idx = args.index("--output")
|
||||
output_path = Path(args[idx + 1])
|
||||
args = [a for i, a in enumerate(args) if i != idx and i != idx + 1]
|
||||
|
||||
json_path = Path(args[0]) if args else ROOT / "GatherTradingData.json"
|
||||
if not json_path.exists():
|
||||
print(f"[ERROR] {json_path} not found")
|
||||
return 1
|
||||
|
||||
raw = json.loads(json_path.read_text(encoding="utf-8"))
|
||||
hc = None
|
||||
try:
|
||||
hc = raw["data"]["_harness_context"]
|
||||
except (KeyError, TypeError):
|
||||
pass
|
||||
if not isinstance(hc, dict):
|
||||
hc = {}
|
||||
|
||||
account_snapshot = raw.get("data", {}).get("account_snapshot", []) or []
|
||||
sell_priority = raw.get("data", {}).get("sell_priority", []) or []
|
||||
core_satellite = raw.get("data", {}).get("core_satellite", []) or []
|
||||
|
||||
# ── TOTAL ASSET CALCULATION ─────────────────────────────────────────────
|
||||
# Sum market_value of all holdings + available_cash (if present in context)
|
||||
holdings_value = sum(float(r.get("market_value", 0) or 0) for r in account_snapshot if r.get("market_value"))
|
||||
|
||||
# Try to find cash in snapshot or context
|
||||
cash_d2 = float(hc.get("settlement_cash_d2_krw") or hc.get("available_cash") or 0)
|
||||
total_asset = holdings_value + cash_d2
|
||||
|
||||
hc["total_asset_krw"] = round(total_asset)
|
||||
hc["total_asset"] = round(total_asset)
|
||||
|
||||
regime = str(hc.get("market_regime", "NEUTRAL"))
|
||||
|
||||
# ticker → account row lookup
|
||||
acct_map: dict[str, dict] = {
|
||||
row["ticker"]: row
|
||||
for row in account_snapshot
|
||||
if row.get("ticker")
|
||||
}
|
||||
# ticker → core_satellite row lookup
|
||||
cs_map: dict[str, dict] = {
|
||||
row["Ticker"]: row
|
||||
for row in core_satellite
|
||||
if row.get("Ticker")
|
||||
}
|
||||
|
||||
computed: dict[str, object] = {}
|
||||
per_ticker: list[dict] = []
|
||||
|
||||
cash_shortfall = hc.get("cash_shortfall_min_krw", 0) or 0
|
||||
|
||||
for sp_row in sell_priority:
|
||||
ticker = sp_row.get("Ticker", "")
|
||||
if not ticker:
|
||||
continue
|
||||
|
||||
acct = acct_map.get(ticker, {})
|
||||
cs = cs_map.get(ticker, {})
|
||||
|
||||
close = cs.get("Close") or acct.get("current_price") or 0
|
||||
prev_close = cs.get("PrevClose") or close
|
||||
ma20 = cs.get("MA20") or close
|
||||
atr20 = cs.get("ATR20") or 0
|
||||
avg_cost = acct.get("average_cost") or 0
|
||||
qty_held = acct.get("holding_quantity") or 0
|
||||
highest = acct.get("highest_price_since_entry") or close
|
||||
stop_price = acct.get("stop_price")
|
||||
sell_limit = sp_row.get("Sell_Limit_Price") or 0
|
||||
ret5d = cs.get("Ret5D") or 0
|
||||
|
||||
# ── VELOCITY ────────────────────────────────────────────────────────
|
||||
velocity_1d = ((close - prev_close) / prev_close * 100) if prev_close else 0
|
||||
velocity_5d = float(ret5d) if ret5d else 0
|
||||
|
||||
# ── PROFIT PCT & STAGE ──────────────────────────────────────────────
|
||||
profit_pct = ((close - avg_cost) / avg_cost * 100) if avg_cost else (
|
||||
acct.get("return_pct") or 0
|
||||
)
|
||||
profit_lock = classify_profit_lock_stage(profit_pct)
|
||||
|
||||
# ── RATCHET V2 ──────────────────────────────────────────────────────
|
||||
ratchet = compute_trailing_stop_v2(
|
||||
profit_pct=profit_pct,
|
||||
highest_close=highest,
|
||||
atr20=atr20,
|
||||
ratchet_stop=stop_price,
|
||||
average_cost=avg_cost,
|
||||
)
|
||||
|
||||
# ── ANTI_CHASING ────────────────────────────────────────────────────
|
||||
anti_chase = compute_anti_chasing(velocity_1d)
|
||||
|
||||
# ── PULLBACK TRIGGER ─────────────────────────────────────────────────
|
||||
pullback = compute_pullback_trigger(close, ma20, atr20)
|
||||
|
||||
# ── SELL PRICE SANITY ────────────────────────────────────────────────
|
||||
sanity = {}
|
||||
if sell_limit:
|
||||
sanity = check_sell_price_sanity(
|
||||
sell_limit_price=sell_limit,
|
||||
stop_loss_price=stop_price,
|
||||
prev_close=prev_close or close,
|
||||
ticker=ticker,
|
||||
)
|
||||
|
||||
# ── SELL DECISION ───────────────────────────────────────────────────
|
||||
sell_ctx = {
|
||||
"close": close,
|
||||
"stopPrice": stop_price,
|
||||
"trailingStop": ratchet.get("auto_trailing_stop_v2"),
|
||||
"tp1Price": sp_row.get("TP1_Price"),
|
||||
"tp2Price": sp_row.get("TP2_Price"),
|
||||
"profitPct": profit_pct,
|
||||
"rwPartial": sp_row.get("RW_Partial"),
|
||||
"timingExitScore": sp_row.get("Timing_Score_Exit"),
|
||||
"daysToTimeStop": sp_row.get("Days_To_Time_Stop"),
|
||||
"timingAction": sp_row.get("Timing_Action"),
|
||||
"regime": regime,
|
||||
"atr20": atr20,
|
||||
}
|
||||
r_sell = compute_sell_decision(sell_ctx)
|
||||
|
||||
# ── FINAL DECISION ──────────────────────────────────────────────────
|
||||
decision_ctx = {
|
||||
"sellAction": r_sell.get("action"),
|
||||
"sellValidation": r_sell.get("validation"),
|
||||
"allowedAction": sp_row.get("Allowed_Action"),
|
||||
"timingAction": sp_row.get("Timing_Action"),
|
||||
"timingScoreEntry": sp_row.get("Timing_Score_Entry"),
|
||||
"timingExitScore": sp_row.get("Timing_Score_Exit"),
|
||||
"ss001Total": sp_row.get("SS001_Total"),
|
||||
"flowCredit": sp_row.get("Flow_Credit"),
|
||||
"leaderTotal": sp_row.get("Leader_Scan_Total"),
|
||||
"rwPartial": sp_row.get("RW_Partial"),
|
||||
"profitPct": profit_pct,
|
||||
"daysToTimeStop": sp_row.get("Days_To_Time_Stop"),
|
||||
"weightPct": sp_row.get("Weight_Pct"),
|
||||
"acGate": sp_row.get("AC_Gate"),
|
||||
"liquidityStatus": sp_row.get("Liquidity_Status"),
|
||||
"spreadStatus": sp_row.get("Spread_Status"),
|
||||
"dartRisk": bool(sp_row.get("DART_Risk")),
|
||||
"missingFields": sp_row.get("Missing_Fields"),
|
||||
}
|
||||
r_final = compute_final_decision(decision_ctx)
|
||||
|
||||
row_result = {
|
||||
"ticker": ticker,
|
||||
"name": sp_row.get("Name", ""),
|
||||
"close": close,
|
||||
"prev_close": prev_close,
|
||||
"ma20": ma20,
|
||||
"atr20": atr20,
|
||||
"avg_cost": avg_cost,
|
||||
"profit_pct": round(profit_pct, 2),
|
||||
"velocity_1d": round(velocity_1d, 4),
|
||||
"velocity_5d": round(velocity_5d, 4),
|
||||
"profit_lock_stage": profit_lock,
|
||||
**ratchet,
|
||||
**anti_chase,
|
||||
**pullback,
|
||||
**(sanity if sanity else {}),
|
||||
**r_sell,
|
||||
**r_final,
|
||||
}
|
||||
per_ticker.append(row_result)
|
||||
|
||||
# ── ORDER BLUEPRINT GENERATION ──────────────────────────────────────────
|
||||
blueprint_rows = []
|
||||
for r in per_ticker:
|
||||
if r.get("final_action") == "SELL_READY":
|
||||
blueprint_rows.append({
|
||||
"ticker": r["ticker"],
|
||||
"name": r["name"],
|
||||
"order_type": r.get("order_type", "LIMIT_SELL"),
|
||||
"limit_price": r.get("limit_price"),
|
||||
"quantity": int(acct_map.get(r["ticker"], {}).get("holding_quantity", 0) * (r.get("ratio_pct", 0) / 100)),
|
||||
"validation_status": "PASS",
|
||||
"rationale_code": r.get("reason"),
|
||||
"formula_id": "ORDER_BLUEPRINT_V1",
|
||||
})
|
||||
|
||||
computed["order_blueprint_json"] = blueprint_rows
|
||||
|
||||
# ── CASH_RECOVERY_OPTIMIZER ─────────────────────────────────────────────
|
||||
if cash_shortfall > 0:
|
||||
recovery = compute_cash_recovery_optimizer(sell_priority, cash_shortfall)
|
||||
computed.update(recovery)
|
||||
|
||||
computed["per_ticker"] = per_ticker
|
||||
computed["meta"] = {
|
||||
"source_file": str(json_path.name),
|
||||
"computed_by": "tools/compute_formula_outputs.py",
|
||||
"formulas_run": [
|
||||
"VELOCITY_V1", "PROFIT_LOCK_STAGE_V1", "PROFIT_RATCHET_TIERED_V2",
|
||||
"ANTI_CHASING_VELOCITY_V1", "PULLBACK_ENTRY_TRIGGER_V1",
|
||||
"SELL_PRICE_SANITY_V1", "CASH_RECOVERY_OPTIMIZER_V1",
|
||||
],
|
||||
"deterministic": True,
|
||||
"llm_computed": False,
|
||||
}
|
||||
|
||||
# ── 출력 ────────────────────────────────────────────────────────────────
|
||||
print(SEP)
|
||||
print(" Python 공식 계산 엔진 — compute_formula_outputs")
|
||||
print(f" 소스: {json_path.name} | 현금부족: {cash_shortfall:,.0f}원")
|
||||
print(SEP)
|
||||
|
||||
print(f"\n{'ticker':<10} {'종목명':<14} {'수익률%':>7} {'단계':<20} "
|
||||
f"{'velocity_1d%':>12} {'뒷박판정':<15} {'trailing_stop':>14}")
|
||||
print("-" * 100)
|
||||
for r in per_ticker:
|
||||
ts = r.get("auto_trailing_stop_v2") or r.get("auto_trailing_stop_v2")
|
||||
ts_str = f"{ts:>14,}" if ts else f"{'(없음)':>14}"
|
||||
print(
|
||||
f"{r['ticker']:<10} {r['name']:<14} {r['profit_pct']:>7.1f}% "
|
||||
f"{r['profit_lock_stage']:<20} {r['velocity_1d']:>12.2f}% "
|
||||
f"{r['anti_chasing_verdict']:<15} {ts_str}"
|
||||
)
|
||||
|
||||
print()
|
||||
|
||||
# 매도가 역전 경보
|
||||
invalid_prices = [r for r in per_ticker if r.get("sell_price_sanity_status", "PASS") != "PASS"]
|
||||
if invalid_prices:
|
||||
print(f"[!] SELL_PRICE_SANITY 경보 — {len(invalid_prices)}개 종목 HTS 입력 차단")
|
||||
for r in invalid_prices:
|
||||
for issue in r.get("sell_price_sanity_issues", []):
|
||||
print(f" {r['ticker']} {r['name']}: {issue}")
|
||||
else:
|
||||
print("[✔] SELL_PRICE_SANITY — 전 종목 가격 정상")
|
||||
|
||||
# 현금회복 계획
|
||||
crp = computed.get("cash_recovery_plan_json")
|
||||
if crp:
|
||||
print(f"\n[현금회복 최적 매도조합] 부족분 {cash_shortfall:,.0f}원")
|
||||
print(f" {'#':<3} {'ticker':<10} {'종목명':<14} {'수량':>6} {'지정가':>10} {'예상회수':>12}")
|
||||
print(" " + "-" * 58)
|
||||
for i, s in enumerate(crp.get("sell_sequence", []), 1):
|
||||
lp = s.get("limit_price") or 0
|
||||
ek = s.get("expected_krw") or 0
|
||||
print(f" {i:<3} {s['ticker']:<10} {s['name']:<14} {s.get('qty', 0):>6} "
|
||||
f"{lp:>10,} {ek:>12,}")
|
||||
met_str = "✔ 달성" if crp.get("shortfall_met") else "✗ 부족"
|
||||
print(f" 합계: {crp.get('expected_total_krw', 0):>12,}원 ({met_str})")
|
||||
|
||||
# ── JSON 파일 저장 ─────────────────────────────────────────────────────
|
||||
if output_path is None:
|
||||
output_path = ROOT / "Temp" / "computed_harness.json"
|
||||
output_path.parent.mkdir(parents=True, exist_ok=True)
|
||||
output_path.write_text(
|
||||
json.dumps(computed, ensure_ascii=False, indent=2), encoding="utf-8"
|
||||
)
|
||||
print(f"\n → 결과 저장: {output_path}")
|
||||
print(f" → 결정론적 계산 완료. LLM 추정 없음.\n")
|
||||
|
||||
return 0
|
||||
|
||||
|
||||
# ─────────────────────────────────────────────────────────────────────────────
|
||||
# IMPUTED_DATA_EXPOSURE_GATE_V1
|
||||
# weighted_coverage = Σ(weight × coverage); ifr = 1 - wc; ech = raw × (0.4 + 0.6 × wc)
|
||||
# gate: ifr ≥ 0.50 → BLOCK / ≥ 0.25 → WARN / else PASS
|
||||
# ─────────────────────────────────────────────────────────────────────────────
|
||||
_IDEG_WEIGHTS = {"fundamental_core": 0.30, "realized_outcome": 0.30,
|
||||
"trade_quality": 0.15, "pattern": 0.10, "alpha_eval": 0.15}
|
||||
|
||||
|
||||
def compute_imputed_data_exposure(domain_coverage: dict, raw_cap: float) -> dict:
|
||||
wc = sum(_IDEG_WEIGHTS.get(k, 0) * v for k, v in domain_coverage.items())
|
||||
ifr = round(1.0 - wc, 4)
|
||||
ech = round(raw_cap * (0.4 + 0.6 * wc), 1)
|
||||
gate = "IMPUTED_DATA_BLOCK" if ifr >= 0.50 else ("IMPUTED_DATA_WARN" if ifr >= 0.25 else "PASS")
|
||||
return {"gate_status": gate, "imputed_field_ratio": round(ifr, 4),
|
||||
"weighted_coverage": round(wc, 4), "effective_confidence_honest": ech}
|
||||
|
||||
|
||||
# ─────────────────────────────────────────────────────────────────────────────
|
||||
# TRAILING_STOP_PRICE_V1
|
||||
# trailing_stop = highest_price - atr20 × multiplier
|
||||
# ─────────────────────────────────────────────────────────────────────────────
|
||||
def compute_trailing_stop_price(highest_price_since_entry: float, atr20: float,
|
||||
trailing_atr_multiplier: float) -> dict:
|
||||
return {"trailing_stop_price": round(highest_price_since_entry - atr20 * trailing_atr_multiplier)}
|
||||
|
||||
|
||||
# ─────────────────────────────────────────────────────────────────────────────
|
||||
# EXPECTED_EDGE_V1
|
||||
# edge = ((tp - entry) / (entry - stop)) × confidence
|
||||
# ─────────────────────────────────────────────────────────────────────────────
|
||||
def compute_expected_edge(target_price: float, entry_price: float,
|
||||
stop_price: float, bayesian_confidence: float) -> dict:
|
||||
r_multiple = (target_price - entry_price) / (entry_price - stop_price) if (entry_price - stop_price) != 0 else 0
|
||||
edge = round(r_multiple * bayesian_confidence, 4)
|
||||
return {"expected_edge": edge, "r_multiple": round(r_multiple, 4)}
|
||||
|
||||
|
||||
# ─────────────────────────────────────────────────────────────────────────────
|
||||
# TP_VALIDITY_CHECK_V1
|
||||
# tp_price > current_price → valid; else null (triggered)
|
||||
# ─────────────────────────────────────────────────────────────────────────────
|
||||
def compute_tp_validity(tp_price: float | None, current_price: float) -> dict:
|
||||
if tp_price is None:
|
||||
return {"tp_validated_price": None, "tp_state": "UNKNOWN_NO_CLOSE"}
|
||||
if tp_price > current_price:
|
||||
return {"tp_validated_price": tp_price, "tp_state": "PENDING"}
|
||||
return {"tp_validated_price": None, "tp_state": "TP1_ALREADY_TRIGGERED"}
|
||||
|
||||
|
||||
# ─────────────────────────────────────────────────────────────────────────────
|
||||
# RS_RATIO_V1
|
||||
# rs_ratio = stock_5d_return / kospi_5d_return
|
||||
# ─────────────────────────────────────────────────────────────────────────────
|
||||
def compute_rs_ratio(stock_close_5d_return: float, kospi_close_5d_return: float) -> dict:
|
||||
if kospi_close_5d_return == 0:
|
||||
return {"rs_ratio": None, "note": "kospi_return=0"}
|
||||
return {"rs_ratio": round(stock_close_5d_return / kospi_close_5d_return, 4)}
|
||||
|
||||
|
||||
# ─────────────────────────────────────────────────────────────────────────────
|
||||
# RATCHET_TRAILING_AUTO_V1
|
||||
# PROFIT_LOCK_20: max(ratchet_stop, highest_close - 1.5×ATR20)
|
||||
# PROFIT_LOCK_30/APEX: max(ratchet_stop, highest_close - 2.0×ATR20)
|
||||
# Others: null
|
||||
# ─────────────────────────────────────────────────────────────────────────────
|
||||
def compute_ratchet_trailing_auto(profit_lock_stage: str, ratchet_stop: float,
|
||||
highest_close: float, atr20: float) -> dict:
|
||||
_mult = {"PROFIT_LOCK_20": 1.5, "PROFIT_LOCK_30": 2.0, "APEX_TRAILING": 2.0}
|
||||
m = _mult.get(profit_lock_stage)
|
||||
if m is None:
|
||||
return {"auto_trailing_stop": None, "note": f"{profit_lock_stage}: no trailing"}
|
||||
atr_stop = highest_close - m * atr20
|
||||
result = max(ratchet_stop, atr_stop)
|
||||
return {"auto_trailing_stop": round(result), "atr_stop": round(atr_stop),
|
||||
"ratchet_stop": ratchet_stop, "multiplier": m}
|
||||
|
||||
|
||||
# ─────────────────────────────────────────────────────────────────────────────
|
||||
# STOP_PRICE_CORE_V1
|
||||
# max(entry × 0.92, entry - atr20 × multiplier)
|
||||
# ─────────────────────────────────────────────────────────────────────────────
|
||||
def compute_stop_price_core(entry_price: float, atr20: float, atr_multiplier: float) -> dict:
|
||||
return _compute_stop_price_core(entry_price, atr20, atr_multiplier=atr_multiplier)
|
||||
|
||||
|
||||
# ─────────────────────────────────────────────────────────────────────────────
|
||||
# TARGET_CASH_PCT_V1
|
||||
# max(5 + (mrs/10)×15, cash_floor_regime_min_pct)
|
||||
# ─────────────────────────────────────────────────────────────────────────────
|
||||
def compute_target_cash_pct(market_risk_score: float, cash_floor_regime_min_pct: float) -> dict:
|
||||
formula_result = 5 + (market_risk_score / 10) * 15
|
||||
target = max(formula_result, cash_floor_regime_min_pct)
|
||||
return {"target_cash_pct": round(target, 2), "formula_result": round(formula_result, 2)}
|
||||
|
||||
|
||||
# ─────────────────────────────────────────────────────────────────────────────
|
||||
# FLOW_CREDIT_V1
|
||||
# C1×0.30 + C2×0.30 + C3×0.40
|
||||
# ─────────────────────────────────────────────────────────────────────────────
|
||||
def compute_flow_credit(c1_price_action: float, c2_volume_action: float,
|
||||
c3_flow_action: float) -> dict:
|
||||
credit = round(c1_price_action * 0.30 + c2_volume_action * 0.30 + c3_flow_action * 0.40, 4)
|
||||
return {"flow_credit": credit}
|
||||
|
||||
|
||||
# ─────────────────────────────────────────────────────────────────────────────
|
||||
# MARKET_RISK_SCORE_V1
|
||||
# min(10, vix + kospi + usd_krw + usd_jpy + credit)
|
||||
# ─────────────────────────────────────────────────────────────────────────────
|
||||
def compute_market_risk_score(vix_score: float, kospi_score: float, usd_krw_score: float,
|
||||
usd_jpy_score: float, credit_score: float) -> dict:
|
||||
raw = vix_score + kospi_score + usd_krw_score + usd_jpy_score + credit_score
|
||||
return {"market_risk_score": min(10, raw), "raw_sum": raw}
|
||||
|
||||
|
||||
# ─────────────────────────────────────────────────────────────────────────────
|
||||
# PORTFOLIO_BETA_V1
|
||||
# beta = Σ(beta_i × mv_i) / total_equity
|
||||
# ─────────────────────────────────────────────────────────────────────────────
|
||||
def compute_portfolio_beta(holdings: list, total_equity_value: float) -> dict:
|
||||
if total_equity_value <= 0:
|
||||
return {"portfolio_beta": None}
|
||||
weighted = sum(h.get("beta", 0) * h.get("market_value", 0) for h in holdings
|
||||
if h.get("beta") is not None)
|
||||
return {"portfolio_beta": round(weighted / total_equity_value, 4)}
|
||||
|
||||
|
||||
# ─────────────────────────────────────────────────────────────────────────────
|
||||
# RISK_BUDGET_CASCADE_V1
|
||||
# base × feedback_mult × brake_mult
|
||||
# ─────────────────────────────────────────────────────────────────────────────
|
||||
def compute_risk_budget_cascade(base_risk_budget: float, net_return_feedback_multiplier: float,
|
||||
performance_brake_multiplier: float) -> dict:
|
||||
result = base_risk_budget * net_return_feedback_multiplier * performance_brake_multiplier
|
||||
return {"risk_budget": round(result, 6)}
|
||||
|
||||
|
||||
# ─────────────────────────────────────────────────────────────────────────────
|
||||
# MEAN_REVERSION_GATE_V1
|
||||
# deviation_ratio = close / ma20; gate by ratio
|
||||
# ─────────────────────────────────────────────────────────────────────────────
|
||||
def compute_mean_reversion_gate(close_price: float, ma20: float) -> dict:
|
||||
if ma20 <= 0:
|
||||
return {"deviation_ratio": None, "gate": "DATA_MISSING"}
|
||||
ratio = round(close_price / ma20, 4)
|
||||
gate = "OVEREXTENDED" if ratio >= 1.10 else ("NORMAL" if ratio >= 0.95 else "OVERSOLD")
|
||||
return {"deviation_ratio": ratio, "gate": gate}
|
||||
|
||||
|
||||
# ─────────────────────────────────────────────────────────────────────────────
|
||||
# T1_FORCED_SELL_RISK_V1
|
||||
# min(100, sell_action_active*40 + timing_exit_ge_50*25 + rw_ge_2*25 + dist_ge_70*30)
|
||||
# ─────────────────────────────────────────────────────────────────────────────
|
||||
def compute_t1_forced_sell_risk(sell_action_active: int, timing_exit_ge_50: int,
|
||||
rw_ge_2: int, distribution_ge_70: int) -> dict:
|
||||
score = min(100, sell_action_active * 40 + timing_exit_ge_50 * 25 +
|
||||
rw_ge_2 * 25 + distribution_ge_70 * 30)
|
||||
gate = "HIGH_SELL_RISK" if score >= 70 else ("MODERATE" if score >= 40 else "LOW")
|
||||
return {"t1_forced_sell_risk_score": score, "gate": gate}
|
||||
|
||||
|
||||
# ─────────────────────────────────────────────────────────────────────────────
|
||||
# SELL_CONFLICT_AWARE_RECOMMENDATION_V1
|
||||
# min(100, sell_signal_active*55 + cash_preserve_active*20 + no_add_gate*20)
|
||||
# ─────────────────────────────────────────────────────────────────────────────
|
||||
def compute_sell_conflict_recommendation(sell_signal_active: int, cash_preserve_active: int,
|
||||
no_add_gate: int) -> dict:
|
||||
score = min(100, sell_signal_active * 55 + cash_preserve_active * 20 + no_add_gate * 20)
|
||||
recommendation = "SELL_PRIORITY" if score >= 55 else ("REDUCE" if score >= 20 else "HOLD")
|
||||
return {"conflict_score": score, "recommendation": recommendation}
|
||||
|
||||
|
||||
# ─────────────────────────────────────────────────────────────────────────────
|
||||
# DIVERGENCE_SCORE_V1
|
||||
# divergence = price_above_ma20 × (frg_sell×0.40 + inst_sell×0.35 + vol_surge×0.25)
|
||||
# ─────────────────────────────────────────────────────────────────────────────
|
||||
def compute_divergence_score(price_above_ma20: int, foreign_net_sell: float,
|
||||
institution_net_sell: float, vol_surge: float) -> dict:
|
||||
score = price_above_ma20 * (foreign_net_sell * 0.40 + institution_net_sell * 0.35 +
|
||||
vol_surge * 0.25)
|
||||
score = round(min(100, max(-100, score)), 2)
|
||||
gate = "DISTRIBUTION_SIGNAL" if score >= 50 else ("NEUTRAL" if score >= 0 else "ACCUMULATION")
|
||||
return {"divergence_score": score, "gate": gate}
|
||||
|
||||
|
||||
# ─────────────────────────────────────────────────────────────────────────────
|
||||
# SEMICONDUCTOR_CLUSTER_SYNC_V1
|
||||
# is_mandatory = cluster_pct > cluster_limit_pct × 2
|
||||
# ─────────────────────────────────────────────────────────────────────────────
|
||||
def compute_semiconductor_cluster_sync(cluster_pct: float, cluster_limit_pct: float) -> dict:
|
||||
is_mandatory = cluster_pct > cluster_limit_pct * 2
|
||||
ratio = round(cluster_pct / cluster_limit_pct, 3) if cluster_limit_pct > 0 else None
|
||||
gate = "MANDATORY_REDUCE" if is_mandatory else "PASS"
|
||||
return {"is_mandatory": is_mandatory, "cluster_ratio": ratio, "gate": gate}
|
||||
|
||||
|
||||
# ─────────────────────────────────────────────────────────────────────────────
|
||||
# GOAL_RETIREMENT_V1
|
||||
# achievement_pct = round(total_asset / GOAL_KRW * 1000) / 10
|
||||
# goal_remaining_krw = max(0, GOAL_KRW - total_asset)
|
||||
# ─────────────────────────────────────────────────────────────────────────────
|
||||
def compute_goal_retirement(total_asset_krw: float, goal_krw: float = 500_000_000) -> dict:
|
||||
achievement_pct = round(total_asset_krw / goal_krw * 1000) / 10
|
||||
remaining_krw = max(0.0, goal_krw - total_asset_krw)
|
||||
status = "ACHIEVED" if achievement_pct >= 100 else "IN_PROGRESS"
|
||||
return {
|
||||
"goal_achievement_pct": achievement_pct,
|
||||
"goal_remaining_krw": round(remaining_krw),
|
||||
"goal_status": status,
|
||||
}
|
||||
|
||||
|
||||
# ─────────────────────────────────────────────────────────────────────────────
|
||||
# VELOCITY_1D_V1 (renamed from VELOCITY_V1 for clarity)
|
||||
# velocity = (close - prev_close) / prev_close × 100
|
||||
# ─────────────────────────────────────────────────────────────────────────────
|
||||
def compute_velocity_1d(close: float, prev_close: float) -> dict:
|
||||
if prev_close <= 0:
|
||||
return {"velocity_1d": None}
|
||||
v = round((close - prev_close) / prev_close * 100, 4)
|
||||
return {"velocity_1d": v}
|
||||
|
||||
|
||||
# ─────────────────────────────────────────────────────────────────────────────
|
||||
# POSITION_SIZE_V1
|
||||
# position_size = min(atr_qty, cash_limit_qty, weight_limit_qty, sector_limit_qty)
|
||||
# atr_qty = floor(risk_budget_krw / (atr20 × atr_mult × close))
|
||||
# ─────────────────────────────────────────────────────────────────────────────
|
||||
def compute_position_size(risk_budget_krw: float, atr20: float, atr_mult: float,
|
||||
close: float, cash_limit_qty: int,
|
||||
weight_limit_qty: int, sector_limit_qty: int) -> dict:
|
||||
import math
|
||||
if atr20 <= 0 or close <= 0 or atr_mult <= 0:
|
||||
return {"position_size_qty": 0, "binding_constraint": "DATA_MISSING"}
|
||||
atr_qty = math.floor(risk_budget_krw / (atr20 * atr_mult * close))
|
||||
constraints = {
|
||||
"atr": atr_qty, "cash": cash_limit_qty,
|
||||
"weight": weight_limit_qty, "sector": sector_limit_qty
|
||||
}
|
||||
min_val = min(constraints.values())
|
||||
binding = [k for k, v in constraints.items() if v == min_val][0]
|
||||
return {"position_size_qty": max(0, min_val), "atr_qty": atr_qty,
|
||||
"binding_constraint": binding, "constraints": constraints}
|
||||
|
||||
|
||||
if __name__ == "__main__":
|
||||
raise SystemExit(main())
|
||||
Reference in New Issue
Block a user