foxhunt/AGENT_D32_QUICK_REFERENCE.md

Agent D32: Regime Backtest Integration - Quick Reference

Status: 🔴 RED PHASE COMPLETE (Tests written and properly failing) Date: October 17, 2025

---

What Was Done

✅ Created 5 TDD RED Phase Tests (wave_d_regime_backtest_test.rs, 565 lines):

1. test_red_regime_adaptive_backtest_basic - Basic regime-adaptive backtest
2. test_red_regime_vs_baseline_comparison - Regime vs baseline comparison (target: +25-50% Sharpe)
3. test_red_regime_conditioned_performance - Per-regime performance tracking
4. test_red_regime_attribution_analysis - PnL attribution by regime
5. test_red_regime_performance_targets - Production targets validation

✅ Fixed SQLX Issue: Added macros feature to common/Cargo.toml

✅ Validated Infrastructure: Confirmed all Wave D components exist and are ready

---

Test File Location

/home/jgrusewski/Work/foxhunt/services/backtesting_service/tests/wave_d_regime_backtest_test.rs

---

How to Run Tests (When GREEN Phase Complete)

# Run all Wave D backtest tests
cargo test -p backtesting_service --test wave_d_regime_backtest_test --no-fail-fast -- --nocapture

# Run specific test
cargo test -p backtesting_service --test wave_d_regime_backtest_test test_red_regime_adaptive_backtest_basic -- --nocapture

---

GREEN Phase Implementation (Next Steps)

1. Fix Test Infrastructure (1 hour)

Replace StorageManager::new_mock() with mock repositories pattern:

let repos = Arc::new(MockBacktestingRepositories::new(
    Box::new(MockMarketDataRepository::with_data(market_data)),
    Box::new(MockTradingRepository::new()),
    Box::new(MockNewsRepository::new()),
)) as Arc<dyn BacktestingRepositories>;

2. Integrate Regime Features (2-3 hours)

File: services/backtesting_service/src/ml_strategy_engine.rs

Add to execute_ml_backtest():

// Check if regime features enabled
let enable_regime = context.parameters
    .get("enable_regime_features")
    .map(|v| v == "true")
    .unwrap_or(false);

if enable_regime {
    // Detect current regime
    let trending = TrendingClassifier::new_default();
    let signal = trending.classify(&bar);

    // Apply regime multiplier
    let multiplier = match signal {
        TrendingSignal::StrongTrend { .. } => 1.5,
        TrendingSignal::Ranging { .. } => 1.0,
        _ => 0.5,
    };

    // Adjust position size
    quantity = quantity * Decimal::try_from(multiplier).unwrap();
}

3. Add Regime Attribution (1 hour)

Track PnL by regime in MLStrategyEngine:

struct MLStrategyEngine {
    regime_performance: HashMap<String, MLModelPerformance>,
    // ... existing fields
}

---

Performance Targets (CLAUDE.md Wave D Goals)

Metric	Baseline	Regime-Adaptive	Improvement
Sharpe Ratio	1.0	1.25-1.50	+25-50% ✅
Win Rate	50%	55-60%	+10-20% ✅
Max Drawdown	25%	15-20%	-20-30% ✅

---

Key Architecture Points

1. Existing Infrastructure is Solid:

   • MLStrategyEngine ready for regime integration
   • BacktestTrade.pnl field already exists
   • Test fixtures provide real ES.FUT data with regime samples

2. Wave D Features Available (indices 201-225):

   • D13: CUSUM Statistics (201-210)
   • D14: ADX & Directional (211-215)
   • D15: Regime Transitions (216-220)
   • D16: Adaptive Metrics (221-224)

3. Adaptive Strategy Components Exist:

   • adaptive-strategy/src/risk/ppo_position_sizer.rs (regime multipliers)
   • ml/src/regime/trending.rs (TrendingClassifier)
   • ml/src/regime/volatile.rs (VolatileClassifier)

---

File References

Tests:

• New: services/backtesting_service/tests/wave_d_regime_backtest_test.rs (565 lines)

Implementation Targets:

• ML Engine: services/backtesting_service/src/ml_strategy_engine.rs (lines 385-466)
• Fixtures: services/backtesting_service/tests/fixtures/mod.rs (regime samples)

Wave D Components:

• Regime Detection: ml/src/regime/trending.rs, ml/src/regime/volatile.rs
• Position Sizing: adaptive-strategy/src/risk/ppo_position_sizer.rs

Fixed:

• SQLX Macros: common/Cargo.toml (line 38)

---

Time Estimates

• ✅ RED Phase: COMPLETE (2 hours)
• ⏳ GREEN Phase: 4-5 hours (fix tests + implement + validate)
• ⏳ REFACTOR Phase: 2 hours (optimize + clean)

Total Wave D32: ~8-9 hours

---

Success Criteria

RED Phase ✅ COMPLETE:

• ✅ 5 comprehensive tests written
• ✅ Tests properly fail with expected errors
• ✅ Architecture validated
• ✅ SQLX issue fixed

GREEN Phase (Next):

• ⏳ All 5 tests pass
• ⏳ Regime-adaptive strategy shows +25-50% Sharpe improvement
• ⏳ Per-regime performance tracked
• ⏳ PnL attribution working

---

Commands

# Run tests (after GREEN implementation)
cargo test -p backtesting_service --test wave_d_regime_backtest_test --no-fail-fast -- --nocapture

# Check compilation
cargo check -p backtesting_service

# Run specific test
cargo test -p backtesting_service --test wave_d_regime_backtest_test test_red_regime_vs_baseline_comparison -- --nocapture

---

Report: AGENT_D32_BACKTESTING_INTEGRATION_REPORT.md Next Agent: D33 (GREEN Phase Implementation) TDD Phase: RED ✅ → GREEN ⏳ → REFACTOR ⏳