Solution: Detecting Leakage in a Time-Series Split

Exercise: Detecting Leakage in a Time-Series Split

Solution

Random folds can train on future observations while validating on earlier dates. Autocorrelation makes nearby future data informative, so this leaks regime information.

Takeaways

• State the assumption that makes the method valid before trusting the result.
• Connect the computation back to the finance workflow, not only to the code.
• Check whether the result depends on average-case behaviour, validation design, or numerical safeguards.