Skip to content
Get started

When the mean and variance estimates move together

For i.i.d. normal data the sample mean Xˉ\bar X and the sample variance S2S^2 are independent. Now suppose daily trade losses are i.i.d. Exponential (with any rate), and you compute both Xˉ\bar X and S2S^2 from the same nn days.

In the large-sample limit n→∞n\to\infty, what is the correlation between Xˉ\bar X and S2S^2? Give a decimal to 3 places.

Show hints (2)+
  1. Centre the data and show that Cov⁡(Xˉ,S2)=μ3/n\operatorname{Cov}(\bar X,S^2)=\mu_3/n, where μ3\mu_3 is the third central moment.
  2. For large nn, Var⁡(S2)≈(μ4−σ4)/n\operatorname{Var}(S^2)\approx(\mu_4-\sigma^4)/n. For Exponential(1): σ2=1\sigma^2=1, μ3=2\mu_3=2, μ4=9\mu_4=9.

Answer

Reveal answer →

0.707 (± 0.003)

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

Interview format: Hedge Fund Quant, Systematic & Data-Driven Trading

Related questions