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

Confused on the general addition rule for variances.

Why is it that when two random variables, X and Y, have a perfect positive correlation (p=1) their standard deviations add. But when they are uncorrelated (p=0) their variances add?

share|cite|improve this question

The equation $\text{Var}(X+Y) = \text{Var}(X) + \text{Var}(Y) + 2 \text{Cov}(X,Y)$ covers all cases. Note that $ \text{Cov}(X,Y) = \rho\; \sigma_X\; \sigma_Y$.

Where it comes from is the linearity of expected value:

$$\eqalign{E[X+Y] &= E[X] + E[Y]\cr E[(X+Y)^2] &= E[X^2 + 2 X Y + Y^2] = E[X^2] + 2 E[XY] + E[Y^2]\cr}$$

and the definitions of variance and covariance in terms of expected values.

share|cite|improve this answer
thanks for the help. so: $[std(x+y) = sqrt[std(x)^2 + std(y)^2 + 2.std(x).std(y)] = sqrt[(std(x)^2 + std(y))^2] = std(x) + std(y)]$ Correct? – crimsontide Nov 29 '12 at 3:45
No, the middle expression (in the case $\rho = 1$) is $\sqrt{(std(x)+std(y))^2}$. – Robert Israel Nov 29 '12 at 6:33
Wow, this has gone a long time without an upvote. Community threw it at the front page. (+1) – robjohn May 22 '13 at 22:22

Further illumination on Robert Israel's answer:

Starting with the formula $$ \begin{align} \mathrm{Var}(X) &=\mathrm{E}\left((X-\mathrm{E}(X))^2\right)\\ &=\mathrm{E}(X^2)-\mathrm{E}(X)^2\\ \end{align} $$ we get $$ \begin{align} \mathrm{Var}(X+Y) &=\mathrm{E}\left(X^2+2XY+Y^2\right)-\left(\mathrm{E}(X)^2+2\mathrm{E}(X)\mathrm{E}(Y)+\mathrm{E}(y)^2\right)\\ &=\left(\mathrm{E}(X^2)-\mathrm{E}(X)^2\right)+\left(\mathrm{E}(Y^2)-\mathrm{E}(Y)^2\right)+2\Big(\mathrm{E}(XY)-\mathrm{E}(X)\mathrm{E}(Y)\Big)\\[5pt] &=\mathrm{Var}(X)+\mathrm{Var}(Y)+2\mathrm{Cov}(X,Y) \end{align} $$ Furthermore, using Hölder, $$ \begin{align} \left|\mathrm{Cov}(X,Y)\right|^2 &=\left|\mathrm{E}\Big((X-\mathrm{E}(X))(Y-\mathrm{E}(Y))\Big)\right|^2\\ &\le\mathrm{E}\left((X-\mathrm{E}(X))^2\right)\mathrm{E}\left((Y-\mathrm{E}(Y))^2\right)\\ &=\mathrm{Var}(X)\mathrm{Var}(Y) \end{align} $$ For perfect positive correlation, we have $\mathrm{Cov}(X,Y)=\mathrm{Var}(X)^{1/2}\mathrm{Var}(Y)^{1/2}$; therefore, $$ \begin{align} \mathrm{Var}(X+Y) &=\mathrm{Var}(X)+\mathrm{Var}(Y)+2\mathrm{Var}(X)^{1/2}\mathrm{Var}(Y)^{1/2}\\ \mathrm{Std}(X+Y)^2 &=\mathrm{Std}(X)^2+\mathrm{Std}(Y)^2+2\mathrm{Std}(X)\mathrm{Std}(Y)\\ \mathrm{Std}(X+Y) &=\mathrm{Std}(X)+\mathrm{Std}(Y) \end{align} $$

For independence, we have $\mathrm{Cov}(X,Y)=0$; therefore, $$ \mathrm{Var}(X+Y)=\mathrm{Var}(X)+\mathrm{Var}(Y) $$

share|cite|improve this answer

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.