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

Assume $X = \mbox{random variable} X>0$, $\mathbb{E}X<\infty \implies \exists \phi:\mathbb R_+\to\mathbb R_+$, superlinar, convex with $\mathbb{E}\phi(X)<\infty$

superlinear means $\lim\frac{\phi(x)}{x}=\infty, x\to\infty$

I have a proof for this statement but I would like to know if there exists also a geometric/heuristic argument for this statement.

share|cite|improve this question

We can assume without loss of generality that $\mathbb{E}[X]=1$. Define now an increasing and unbounded sequence $\{x_n\}_{n\in\mathbb{N}}$ through: $$ x_n = \inf\left\{x\in\mathbb{R}^+:\mathbb{P}[X>x]\leq\frac{1}{4^n}\right\}$$ (so $x_0=0$) and $$ i(x) = \max\left\{n\in\mathbb{N}:x_n\leq x\right\}.$$ You can take $$ \phi(x) = x\cdot\frac{2^{i(x)+1}(x-x_{i(x)})+2^{i(x)}(x_{i(x)+1}-x)}{x_{i(x)+1}-x_{i(x)}}$$ as a non-negative continous, increasing convex function and easily prove that with such choices $$ \mathbb{P}[\phi(X)>t]\leq 4\cdot\mathbb{P}[X>t],$$ so $\mathbb{E}[\phi(X)]$ is necessarily finite.

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.