If $Y$ is symmetrically distributed around $E(Y)$ then $E(Y^j)=E(Y)^j$ for every odd $j$, why not for $j$ even?

130 Views Asked by At

In this statement "If a random variable Y is symmetrically distributed around its mean $\mu_Y$, that is, $Y=\mu_Y+U$, where U is symmetrically distributed aronud zero- then the odd moments of Y are powers of $\mu_Y$, i.e. $E[(Y)^j]= (\mu_Y)^j $ for any odd integer j, assuming the moment exists".

I think this statement is true for both even and odd moments. What is the significance of the odd moment? And is it true when it comes to the even moment?

1

There are 1 best solutions below

10
On BEST ANSWER

Hint: All you need to know is that $U$ is symmetrically distributed with zero mean. This implies that all odd moments of $U$ will be 0.

So, we know that

$$E[(Y-\mu_Y)^3]=E[Y^3 -3 Y^2\mu_Y+3 Y\mu_Y^2-\mu_Y^3]=0 \implies$$ $$ E[Y^3] = E[3 Y^2\mu_Y-3 Y\mu_Y^2+\mu_Y^3]=3 \mu_Y E[Y^2]-3 \mu_Y^2 E[Y]+\mu_Y^3 \implies $$

$$E[Y^3]= 3 \mu_Y E[Y^2]-2\mu_Y^3 $$

This value depends on the value for $E[Y^2]$, which we can derive as follows:

$$Var[Y] = E[Y^2]-E[Y]^2 \implies E[Y^2] = Var[Y] + \mu_Y^2$$

However, $Var[Y]=Var[U]$ so

$$E[Y^2] = Var[U] + \mu_Y^2$$

So, assuming $Var[U]>0$, what can you conclude?