Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

Wiki states that standard deviation of $X-Y$ is:

$$\sigma_{x-y} = \sqrt { \sigma_x^2 + \sigma_y^2 - 2\rho\sigma_x\sigma_y }$$

I have a number (say 3) correlated random variables to be subtracted from another correlated random variable.

All random variables have identical correlation $\rho$.

Can I subtract each one in turn like this:

$$\sigma_{x-1} = \sqrt { \sigma_x^2 + \sigma_1^2 - 2\rho\sigma_x\sigma_1 }$$

$$\sigma_{x-2} = \sqrt { \sigma_{x-1}^2 + \sigma_2^2 - 2\rho\sigma_{x-1}\sigma_2 }$$

$$\sigma_{x-3} = \sqrt { \sigma_{x-2}^2 + \sigma_3^2 - 2\rho\sigma_{x-2}\sigma_3 }$$

The application is determining carrier-to-interference ratio of multiple co-channel interferers in an environment where it can be assumed that each interferer has identical correlated variation.

share|cite|improve this question
up vote 2 down vote accepted

$\newcommand{\Var}{\operatorname{Var}} \newcommand{\Cov}{\operatorname{Cov}}$ I think it is easier to work with variance and covariance (which are easily computed from correlation and standard deviation, and vice versa). The key identity is $$\Var(X \pm Y) = \Var(X) + \Var(Y) \pm 2 \Cov(X,Y).$$ (This is very simple to prove; it is nothing but linearity of expectation and some algebra.) Similarly, you can show $$\Cov(X \pm Y, Z) = \Cov(X,Z) \pm \Cov(Y,Z).$$ Now by using these identities repeatedly, we could show $$\begin{align*} \Var(X_1 - X_2 - X_3) &= \Var(X_1 - (X_2 + X_3)) \\ &= \Var(X_1) + \Var(X_2 + X_3) - 2 \Cov(X_1, X_2 + X_3) \\ &= \Var(X_1) + \Var(X_2) + \Var(X_3) + 2 \Cov(X_2, X_3) \\ &\quad\quad\quad\quad- 2 \Cov(X_1, X_2) - 2 \Cov(X_1, X_3). \end{align*}$$ I'll leave you to perform the corresponding computation for 4 random variables.

share|cite|improve this answer
@Didier: Thanks for the typo fix. – Nate Eldredge Aug 2 '11 at 0:58
Thanks. I get Var(X1)+Var(X2)+Var(X3)+Var(X4) - interestingly, co-variance terms cancel. – philcolbourn Aug 2 '11 at 9:50

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.