Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. It's 100% free, no registration required.

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

Does knowing that for a iid $X_j$'s, $EXj = 0$ and $EX_j^2 < \infty$ indicate $E|X_j|$ is finite too? How to show that.

Can we say: $EX_j^2 = E|X_j|^2 < \infty \rightarrow \text{then } E|X_j| < \infty$ I appreciate your help.

share|cite|improve this question
You can use Jensen's inequality for $|X|$, for example: $E|X|^2 - (E|X|)^2 \geq 0$ to create an upper bound for $E|X|$ – yoki Nov 29 '12 at 5:30
Cool...thanks... – Sam Nov 29 '12 at 5:37
No need for jensen's inequality. Use $Var(|X|) \geq 0$. Its the same thing but you get it from here also. – Gautam Shenoy Nov 29 '12 at 6:02

This is a matter of definition: E(X) is undefined when X is not integrable. Hence, to assume that E(X) = 0 is, in particular, to assume that X is integrable, which is needed for E(X) to exist, that is, to assume that (X is measurable and that) E(|X|) is finite.

share|cite|improve this answer

It's not necessary for $E[X_j^2]$ to be finite either. Just knowing $E[X_j]$ is finite is sufficient. For instance, consider $$ X = n \quad \mbox{w.prob} \quad c\frac{1}{n^3} \quad \forall n \geq 1$$ where $c=\sum \frac{1}{n^3}$ which converges by the way.

Here $E[X^2]$ is infinite but the mean is finite if you compute it.

To show that $E[X] < \infty$ implies $E[|X|] < \infty$, observe $X=X^+ - X^-$, where $X^+ = \max(X,0)$ and $X^+ = \max(-X,0)$. Then $$E[X] = E[X^+ - X^-] = EX^+ - EX^-$$ But the way Lebesgue integrals are defined, E[X] is finite if and only if EX^+ and EX^- are finite. This is due to construction. So it is disallowed for either term to be infinite. Thus $$E[|X|] = E[X^+] + E[X^-] < \infty$$

So all in all, you're condition for $E[X^2] < \infty$ is superfluous.

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.