docs(book): Complete Chapter 11 Pairs Trading pedagogical expansion

- Expanded from 5,720 to 13,831 words (8,111 words added)
- Added 1,955 lines of production-ready OVSM code

NEW SECTIONS:
- 11.5.3: Ornstein-Uhlenbeck parameter estimation (~800 words, 150 lines)
  * AR(1) approximation for discrete-time OU process
  * Half-life calculation and validation
  * Complete worked example with GS/MS pair

- 11.5.4: Dynamic hedge ratios with Kalman filter (~1,400 words, 240 lines)
  * State-space formulation for time-varying beta
  * Predict-update cycle implementation
  * Static vs. Kalman comparison framework
  * August 2007 lesson: static betas obsolete in 48 hours

- 11.5.5: Complete backtesting framework (~1,500 words, 300 lines)
  * Walk-forward pair selection (formation/trading periods)
  * Event-driven execution (avoid look-ahead bias)
  * 5-component transaction cost model (38 bps reality)
  * References Chapter 9 methodology

- 11.5.6: Production risk management system (~2,200 words, 350 lines)
  * Multi-layered controls: position limits, stop-losses, circuit breakers
  * Correlation monitoring (detect regime changes)
  * VaR calculation, drawdown tracking
  * Kill switch implementation
  * August 2007 survivor analysis: Renaissance (-5%), AQR (-13%) vs. failures (-65%)
  * Cost-benefit: $200/mo → survival ROI = ∞

- 11.6: Summary and key takeaways (~1,500 words)
  * Evidence-based success factors
  * Common failure modes with disaster analysis
  * Realistic 2024 expectations (Sharpe 0.8-1.2 vs. 2.0 historically)
  * How to avoid disaster list ($155B in preventable losses)

- 11.7: Exercises (~400 words)
  * Mathematical derivations (half-life, optimal thresholds)
  * Coding projects (Johansen test, HMM regime detection, risk dashboard)
  * Empirical research (cross-asset analysis, strategy decay)

- 11.8: Expanded references (~600 words, 20+ citations)
  * Gatev et al. (2006), Engle-Granger (1987), Khandani-Lo (2007)
  * Cointegration theory, OU processes, risk management
  * ML extensions, industry perspectives

- 11.9: Conclusion (~400 words)
  * Theory vs. implementation gap
  * August 2007 as survival lesson
  * Defensive engineering imperative

DIAGRAMS ADDED:
- Walk-forward backtesting flowchart
- Risk management architecture (multi-layer controls)
- Kalman filter state machine
- Risk-return quadrant chart (position sizing)

KEY THEMES:
- Disaster-driven learning (August 2007 quant quake throughout)
- Production-ready implementations (not toy examples)
- Risk management emphasis (cost $0-200/mo, benefit = survival)
- Cross-chapter integration (Chapters 8, 9, 10 referenced)
- Evidence-based guidance (20+ academic citations)

PEDAGOGICAL APPROACH:
- 60% explanation, 40% code (established ratio)
- WHAT/WHY/HOW function documentation
- Real disaster analysis (Aug 2007: Renaissance survived, others lost 65%)
- Practical cost-benefit (implement all controls: 20-30 hours, $200/mo)
- Realistic expectations (Sharpe 0.8-1.2 in 2024 vs. 2.0 in 1990s)

STATUS: Production-ready, comprehensive pairs trading guide

Author: 0xrinegade