Calculate variance of a distribution including a retention - can someone help?

23 Views Asked by At

Imagine we have a random variable X with a certain distribution, being easy to calculate;

Now we construct a retention in it, (or also an insurance sum, so a maximum sum being paid out) - in this case we use a retention and call this variable Y... For example we have then:

$$Y=0\cdot\mathbb{1}_{\{x<50\}}+(X-50)\cdot\mathbb{1}_{\{x\geq50\}} $$

which is now the payout if we have a retention of 50; Now calculating the expectation is easy:

$$\mathbb{E}(Y)=\mathbb{E}(X\cdot\mathbb{1}_{\{x\geq50\}})-50\mathbb{E}(\mathbb{1}_{\{x\geq50\}})= \int_{50}^{\infty}xf_x(x)dx-50 \mathbb{P}(X\geq 50) $$

But now I'm lost, how can I now calculate the variance of Y now - I though of calculating $$\mathbb{E}(Y^2)$$ and then calculate the variance with that but I somehow don't know how to do it - can someone give me an advice?

Thank you so much!!!

EDIT:

Now let's make Y a more compliacated, e.g.

$$Y=0\cdot\mathbb{1}_{\{x<50\}}+(X-50)\cdot\mathbb{1}_{\{x\geq50 \text{ and } x\leq200\}} + (200 + 0.1 \cdot X) \mathbb{1}_{\{x\geq200\}} $$

They payout doesn't really make sense but just that I understand it better... How do I calculate the variance here?

1

There are 1 best solutions below

4
On BEST ANSWER

I would proceed as you suggested and use $Var(Y) = \mathbb{E}[Y^2] - \mathbb{E}[Y]^2$.

For $\mathbb{E}[Y^2]$ we have $Y^2 = (X-50)^2 \mathbb{1}_{X \geq 50}(X)$, so

$$ \mathbb{E}[Y^2] = \int_{x=50}^\infty (x-50)^2 f_X(x) \mathrm{d} x. $$

The variance becomes

$$ Var(Y) = \int_{x=50}^\infty (x-50)^2 f_X(x) \mathrm{d} x - \left( \int_{x=50}^\infty (x-50) f_X(x) \mathrm{d} x \right)^2.$$