Regime Detection Integration Task List

Critical Issues to Fix

🔴 P0 - Critical (Must Fix Immediately)

1. Start RegimeDetectionService in API

• File: src/alpha_pulse/api/main.py
• Status: ❌ NOT STARTED
• Impact: Service exists but never runs
• Effort: 1 hour
• Code: See src/alpha_pulse/integration/api_regime_startup.py

2. Create RegimeIntegrationHub

• File: NEW - src/alpha_pulse/integration/regime_integration.py
• Status: ✅ CREATED (ready to integrate)
• Impact: Central distribution of regime information
• Effort: Already done

🟡 P1 - High Priority (Fix This Week)

3. Update All Trading Agents

• Files: All files in src/alpha_pulse/agents/
• Status: ❌ Only Technical agent has basic regime detection
• Impact: Agents operate without market context
• Effort: 4 hours
• Tasks:
  • Modify agent factory to pass regime hub
  • Update Technical agent to use full HMM
  • Add regime awareness to Fundamental agent
  • Add regime awareness to Sentiment agent
  • Add regime awareness to Value agent
  • Add regime awareness to Activist agent

4. Integrate Risk Manager

• File: src/alpha_pulse/risk_management/risk_manager.py
• Status: ❌ NOT INTEGRATED
• Impact: Risk limits don’t adapt to market conditions
• Effort: 2 hours
• Code: See src/alpha_pulse/integration/portfolio_risk_regime_integration.py

5. Integrate Portfolio Optimizer

• Files: src/alpha_pulse/portfolio/*.py
• Status: ❌ NOT INTEGRATED
• Impact: Portfolio allocation ignores market regime
• Effort: 3 hours
• Tasks:
  • Update MeanVarianceOptimizer
  • Update HierarchicalRiskParity
  • Update BlackLittermanOptimizer

🟢 P2 - Medium Priority (Fix This Month)

6. Add Regime API Endpoints

• File: NEW - src/alpha_pulse/api/routes/regime.py
• Status: ❌ NOT CREATED
• Impact: No visibility into regime detection
• Effort: 2 hours
• Endpoints:
  • GET /regime/current
  • GET /regime/history
  • POST /regime/analyze
  • GET /regime/parameters

7. Update Backtesting Framework

• File: src/alpha_pulse/backtesting/backtest_engine.py
• Status: ❌ NOT INTEGRATED
• Impact: Can’t analyze historical regime performance
• Effort: 4 hours
• Tasks:
  • Track regime during backtest
  • Generate regime-specific metrics
  • Add regime transition analysis

8. Create Regime Dashboard

• Files: dashboard/src/components/RegimePanel.tsx
• Status: ❌ NOT CREATED
• Impact: No real-time regime visibility
• Effort: 6 hours

🔵 P3 - Nice to Have

9. Advanced Regime Features

• Multi-timeframe regime detection
• Cross-asset regime correlation
• Regime prediction (not just detection)
• Custom regime definitions

10. Performance Optimizations

• Regime calculation caching
• Distributed regime computation
• GPU acceleration for HMM

Implementation Checklist

Week 1 Sprint

• Start RegimeDetectionService in API startup
• Test service is running and detecting regimes
• Update agent factory to pass regime hub
• Modify at least 2 agents to use regime

Week 2 Sprint

• Complete all agent integrations
• Integrate risk manager
• Start portfolio optimizer integration
• Add basic regime endpoints

Week 3 Sprint

• Complete portfolio optimizer integration
• Add regime tracking to backtesting
• Create regime monitoring dashboard
• Run integration tests

Week 4 Sprint

• Performance testing with regime
• Documentation updates
• Production deployment preparation
• Monitor regime detection accuracy

Code Snippets for Quick Implementation

1. Minimal API Startup Change

# In api/main.py startup_event()

from alpha_pulse.services.regime_detection_service import RegimeDetectionService

# After other service initialization
regime_service = RegimeDetectionService(
    config=config.regime_detection,
    metrics_collector=metrics_collector
)
await regime_service.start()
app.state.regime_service = regime_service

2. Quick Agent Update

# In any agent's analyze method

# Get current regime
regime = await app.state.regime_service.get_current_regime()
if regime == MarketRegime.CRISIS:
    # Reduce signal strength or skip
    return []

3. Risk Manager Quick Fix

# In risk_manager.py calculate_position_size()

regime = await self.regime_service.get_current_regime()
regime_multipliers = {
    MarketRegime.BULL: 1.2,
    MarketRegime.BEAR: 0.6,
    MarketRegime.CRISIS: 0.3
}
size *= regime_multipliers.get(regime, 1.0)

Validation Tests

Test 1: Service Running

curl http://localhost:8000/regime/current
# Should return current regime, not 404

Test 2: Agents Using Regime

# Check agent signals include regime metadata
signal = agent.analyze(data)
assert 'regime' in signal.metadata

Test 3: Risk Adjustment

# Verify position sizes change with regime
bull_size = risk_manager.calculate_position_size(signal, MarketRegime.BULL)
crisis_size = risk_manager.calculate_position_size(signal, MarketRegime.CRISIS)
assert crisis_size < bull_size

Success Metrics

• ✅ RegimeDetectionService runs continuously
• ✅ All 6 agents use regime information
• ✅ Risk limits adjust based on regime
• ✅ Portfolio allocation considers regime
• ✅ Regime visible in monitoring dashboard
• ✅ Backtest results include regime analysis

Estimated Timeline

• Total Effort: ~25 hours
• Critical Path: 3-5 hours (just starting the service)
• Full Integration: 2-3 weeks
• With Testing: 4 weeks

Risk Mitigation

1. Gradual Rollout: Start with just service running, add components gradually
2. Feature Flags: Use config to enable/disable regime integration
3. Monitoring: Track regime detection accuracy before full integration
4. Fallback: Keep non-regime code paths as fallback

Questions to Answer

1. Should regime detection block trading if confidence is low?
2. How often should regime be recalculated? (Currently 5 min)
3. Should we alert on every regime change?
4. What’s the minimum historical data for regime detection?

Next Action

START HERE: Open src/alpha_pulse/api/main.py and add the regime service initialization in the startup event. This single change will activate the entire regime detection system that’s currently sitting idle.