From Cointegration to Out-of-Sample Failure: A Pairs-Trading Case Study on PEP/KO

Battle Count: 0

Highly relevant as it provides a rigorous case study on the pitfalls of backtesting, specifically highlighting out-of-sample failure, the impact of transaction costs, and the importance of robustness checks like Deflated Sharpe Ratios in pairs trading.

Moderate complexity. Requires knowledge of time series econometrics (ADF, Engle-Granger), state-space models (Kalman Filter), and backtesting frameworks. The logic is rule-based, but careful handling of look-ahead bias and parameter estimation is crucial.

The paper provides detailed parameter settings (thresholds, costs, windows), statistical tests, and references an extended report with implementation details on GitHub. However, specific data sources beyond 'adjusted daily closing prices' are not explicitly linked in the extract, though standard financial data is implied.