Quant Memo
Statistics/●●●●●

Two random walks with R² of 0.9 and no relationship

Asked at Two Sigma, Point72

You simulate two completely independent random walks (each day's value is the previous value plus fresh, unrelated noise) and regress one on the other. The R2R^2 often comes out at 0.70.7, 0.90.9, sometimes higher, with a "highly significant" t-statistic.

How can two unrelated series produce a huge R2R^2, and what does this teach about trusting R2R^2?

Show a hint

R2R^2 measures how much of yy's variation the line explains in-sample. A random walk wanders far from its mean. What does "variation around the mean" even mean for a series that trends?

Your answer

This one is open-ended. Work it through, then check your reasoning against the full solution.

More Statistics questions