Applied Statistical Arbitrage: Scaling Strategies and Managing Relationship Decay

A pairs trading strategy that works beautifully in backtest and initial live trading will eventually face the challenge that defines advanced statistical arbitrage: the relationship changes. Regulatory changes affect one leg. A corporate event permanently alters the relative value of the pair. The correlation structure shifts during a market regime change. This course prepares you to recognise these events, respond systematically, and manage a portfolio of pairs over a sustained trading horizon. By the end of this course you will be able to monitor a live statistical arbitrage strategy for signs of relationship decay, apply rolling cointegration analysis to detect when a pair's statistical relationship is weakening, expand from single pairs to a portfolio of correlated strategies while managing aggregate risk, evaluate execution quality in market-neutral strategies where both legs must trade near-simultaneously, and design a review and retirement protocol for pairs that no longer meet quality standards. What you will learn: - Rolling cointegration monitoring: detecting decay before it destroys accumulated profit - Spread divergence management: distinguishing temporary widening (opportunity) from structural breakdown (exit signal) - Hedge ratio drift: why static ratios become stale and how to update them without curve-fitting - Portfolio of pairs: managing correlation across multiple positions and controlling aggregate dollar exposure - Execution challenges: leg risk in simultaneous two-leg entries and strategies to minimise it - Mean reversion in different asset classes: how pairs behave differently in equities, ETFs, futures, and fixed income - Regime conditioning: adjusting strategy activity based on broad market volatility regime - Pair retirement protocol: the criteria and documentation process for removing a pair from active trading The course uses extended case studies — a sector ETF pair that weakens after a market-structure change, a two-futures pair disrupted by a change in contract rollover conventions, a portfolio of equity pairs that becomes unintentionally correlated during a market sell-off. Reflection prompts ask you to diagnose each scenario before reading the worked response. This course is designed for algorithmic traders who have built and run at least one pairs strategy and want to develop a more professional long-term management framework. Written for those who are no longer new to statistical arbitrage but face the challenges of sustaining it over time. This course is informational and educational and does not constitute financial or investment advice.