Skip to content
Get started

R² from pure noise

You regress a pure-noise target on 1010 pure-noise regressors (all mutually independent standard normal) plus an intercept, using n=61n=61 observations.

What is the expected in-sample R2R^2? Give a decimal to 4 places.

Show hints (2)+
  1. After removing the intercept, how many dimensions does the noise vector live in?
  2. By symmetry each dimension carries an equal share of the expected sum of squares.

Answer

Reveal answer →

0.1667 (± 0.002)

Create a free account to see the full worked solution and track your progress.

Interview format: Prop Trading, Systematic & Data-Driven Trading

Related questions