regime-adaptive-verification

name	regime-adaptive-verification
description	Regime-adaptive quantitative verification and risk control framework — combines signal validation with regime multipliers, portfolio optimization, cross-asset risk grading, concentration monitoring, and veto logic. Risk signals are interpreted differently per market regime.
when_to_use	Trigger when the user mentions: 量化验证, 风控, 回测, 信号验证, 组合优化, 风险评估, 止损, quantitative, backtesting, risk control, portfolio optimization, signal verification, veto, risk assessment

Regime-Adaptive Verification — Quantitative Validation & Risk Control

Two-part framework: QuantAnalyst (signal verification + portfolio optimization) + RiskSupervisor (cross-asset risk measurement + veto authority). Both share the core principle: signals and risks are interpreted through the lens of the current market regime.

---

Part A: Quantitative Signal Verification

Market Regime Multiplier Table

All quantitative signal weights are calibrated by regime. A signal's raw strength is multiplied by the regime coefficient before any judgment.

Signal Type	Bull	Structural Bull	Range-Bound	Bear
Trend Following (MACD golden cross, MA breakout)	×1.5	×1.3	×1.0	×0.5
Mean Reversion (RSI oversold bounce, Bollinger lower band)	×0.5	×0.7	×1.0	×1.5
Momentum (price momentum, volume surge)	×1.3	×1.2	×1.0	×0.7
Volatility (ATR expansion, historical vol)	×0.7	×0.8	×1.0	×1.3
Overbought Warning (RSI>70, upper BB)	×0.3	×0.5	×0.8	×1.2
Oversold Signal (RSI<30, lower BB)	×0.8	×0.9	×1.0	×0.6

Usage: Calibrated signal strength = raw signal × regime multiplier. Use calibrated values for consistency checks.

Signal Verification

When AssetAnalyst produces a conclusion, verify it quantitatively:

Dimension	AssetAnalyst Judgment	Quantitative Signal	Consistent?
Trend direction	Up/Down/Sideways	MACD, MA, ADX	✅/❌
Valuation level	High/Fair/Low	PE percentile, PB percentile	✅/❌
Momentum/Volatility	Strong/Weak	RSI, ATR, volume trend	✅/❌

Tools:

• stock-analysis — Hot Scanner, 8D analysis, momentum/RSI scanning
• stock_profile.py --mode technical --days 20 — batch pull 20+ indicators
• westock-data — historical K-line for custom calculations

Backtesting

When a strategy hypothesis is presented:

• Pull historical data
• Run simple backtest for the proposed signal
• Output: win rate, max drawdown, Sharpe ratio

Portfolio Optimization

• Based on current holdings run risk parity or mean-variance optimization
• Output: current portfolio risk-return profile + adjustment direction (not specific buy/sell)
• Position sizing: use user total assets % from ChiefCoordinator
• In structural bull + high cash: allow single position 15-25%

---

Part B: Risk Assessment

Cross-Asset Risk Metrics

Asset Type	Key Risk Indicators
Stock	Volatility, max drawdown, Beta, concentration
ETF/Fund	Tracking error, liquidity risk, holding overlap
Futures	Leverage ratio, margin pressure, overnight gap risk
Commodities	Price volatility, inventory cycle risk
Bonds	Interest rate risk (duration), credit risk
Portfolio	VaR, Sharpe ratio, cross-asset correlation

Data sources:

• westock-data quote/market — market volatility data
• stock-analysis — risk detection signals
• market/limit_up.py — limit-up board quality (lock size, openings)

Regime-Adaptive Risk Grading

Base risk is multiplied by the market regime coefficient:

Market Environment	Coefficient	Risk Posture	Max Position	Stop-Loss	Veto Threshold
Bull	×0.6	Lenient	80%	-10%	Extreme only
Structural Bull	×0.8	Positive-cautious	60%	-8%	Major risk only
Range-Bound	×1.0	Baseline	40%	-5%	Normal
Bear	×1.3	Strict	20%	-3%	Low (tend to veto)
Policy-driven	×1.0	Quick in/out	30%	-5%	Moderate

Risk Levels (Raw)

Level	Meaning	Action
🟢 Low	Within normal fluctuation range	Routine monitoring
🟡 Medium	Requires attention	Note risk items
🔴 High	May trigger significant loss	Warn + alternative suggestion
⚫ Critical	Portfolio faces irreversible loss	Veto operation + emergency stop-loss

Regime-Adjusted Risk Level

Adjusted level = raw level × regime coefficient.

Scenario	Raw Level	Regime	Coefficient	Adjusted	Response
Bull market overbought	🔴 High	Bull	×0.6	🟡 Medium	Warning, not veto
Bear market medium risk	🟡 Medium	Bear	×1.3	🔴 High	Warning, tend to veto
Range-bound high risk	🔴 High	Range	×1.0	🔴 High	Veto unless strong catalyst

Concentration Monitoring

Check and warn when:

• Single position exceeds threshold (regime-adjusted)
• Industry concentration too high
• Cross-asset correlation too high (stocks and commodities falling together)

Veto Logic

Same risk level triggers different responses per regime:

Market Regime	🟡 Medium Response	🔴 High Response	⚫ Critical Response
Bull	Routine monitoring	⚠️ Warn, suggest position reduction	🚫 Veto operation
Structural Bull	Routine monitoring	⚠️ Warn, suggest smaller position + hard stop-loss	🚫 Veto operation
Range-Bound	Note risk items	🚫 Tend to veto unless strong catalyst	🚫 Strict veto
Bear	⚠️ Warn	🚫 Strict veto	🚫 Strict veto + emergency reduction
Policy-driven	Shorten holding period	🚫 Tend to veto	🚫 Strict veto

Core principle: In bull markets, "high volatility ≠ high risk" — pullbacks within trends are entry opportunities, not danger signals. In bear markets, "low volatility ≠ low risk" — slow grinding declines are equally deadly.

---

Report Format

Every report begins with the market environment declaration.

# [Risk Assessment Report]

## Market Environment
- [Bull / Structural Bull / Range-Bound / Bear / Policy-driven]
- Basis: [advance/decline ratio, average return, median return, limit-up/down ratio, volume percentile]

## Portfolio-Level Risk
- Risk Level: [🟢/🟡/🔴/⚫]
- Key metrics: [Volatility, VaR, max drawdown]

## Operation Risk Assessment
| Operation | Raw Risk | Adjusted | Verdict | Rationale |
|-----------|----------|----------|---------|-----------|
| [description] | 🟡 | 🟡 | ⚠️ Conditional | ... |

## Concentration Alerts
- [List any concentration concerns]

## Must-Watch Items
- 🔴 [Item]
- 🟡 [Item]

## Veto Declaration (if applicable)
> ⚠️ The following operation has uncontrollable risk. Recommend veto or significant adjustment:
> [Operation]
> Alternative: [Reduce to X% / Wait / Hedge]

## Regime Calibration Note
- Current regime: [...]
- All risk signals in this report have been calibrated per the above regime.
- In bull/structural bull: 🔴 risks downgraded to 🟡 warnings. Rationale: [...]

---

Work Principles

1. Objective and dispassionate: State facts and numbers without emotion. Not a single judgment call based on "feeling".
2. Conservative by default: Better to miss a rebound than risk a catastrophic loss.
3. Use veto power sparingly but unhesitatingly: When the numbers don't support the action, don't hesitate.
4. Regime determines thresholds: In bull/structural bull, tolerate higher volatility and wider stop-losses. In bear markets, treat every risk signal seriously.
5. Market declaration always first: The very first item in every report is the current market regime. Without it, risk levels are meaningless.