I am trying to find out the distribution of $n$ i.i.d. random variables $X_1, ..., X_n$ so that their product $X_1 \cdot ... \cdot X_n$ follows a Gamma distribution.

Or in other words, assume that $X_1, ..., X_n$ are i.i.d. distributed random variables and that the product of the $n$ random variables $X_1\cdot ... \cdot X_n$ follows a Gamma distribution. What is the distribution of $X_i$?

I already know that the product of Gamma distributed random variables does not seem to follow a Gamma distribution, since the pdf of two Gamma distributed random variables has a complicated pdf that includes a BesselK function. PDF of the product of two independent Gamma random variables

I have the feeling that this is a difficult question, therefore, any help is highly appreciated.

Thank you.


If a product of variables has pdf $f(x)$, the product's logarithm, which is a sum, has pdf $e^x f(e^x)$, in this case over the whole of $\mathbb{R}$. Can you get the characteristic function for that, in terms of the Gamma function? You can get the CF of $\ln X_1$ from that.

  • $\begingroup$ Thank you very much for your help @J.G. I definitively need to think more about it to better understand your idea and to find the solution. $\endgroup$ – Mathias Jul 25 '18 at 20:10

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.