I've learned that multiplying two generating functions $f(x)$ and $g(x)$ will give the result

\begin{equation*} \sum_{k=0}^\infty\left(\sum_{j=0}^k a_j\,b_{k-j}\right)x^k. \end{equation*}

I've used the result, but it was presented in my class without proof and I'm having some trouble tracking one down. Weak google-foo today, I suppose. Can anyone give me a pointer to a proof? If this is a question better answered in book form, that is fine as well.


2 Answers 2


Casebash is correct that this is a definition and not a theorem. But the motivation from 3.48 (Defintion of product of series) of little Rudin may convince you that this is a good definition:
$\sum_{n=0}^{\inf} a_n z^n \cdot \sum_{n=0}^{\inf} b_n z^n = (a_0+a_1z+a_2z^2+ \cdots)(b_0+b_1z+b_2z^2+ \cdots)$
$=a_0b_0+(a_0b_1 + a_1b_0)z + (a_0b_2+a_1b_1+a_2b_0)z^2 + \cdots$
$=c_0+c_1z+c_2z^2+ \cdots $
where $c_n=\sum_{k=0}^n a_k b_{n-k}$

  • $\begingroup$ Excellent additional info, thanks. $\endgroup$
    – dsolimano
    Jul 28, 2010 at 3:21

It is actually the other way round. A generating function is generally defined to have an addition operation where the components are added and a multiplication operation like that you mentioned. Once we have made these definitions, we observe that polynomials obey the same laws and so that it is convenient to represent generating functions as infinite polynomials rather than just an infinite tuple.


You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .