Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. It's 100% free, no registration required.

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

If $X$ is a Banach space, and $T:X \to X$ is a bounded linear operator with norm < $1$, then $I-T$ has a bounded inverse defined by $(I-T)^{-1} = \sum_{n=0}^\infty T^n$.

Thinking in terms of a converse, if $T$ is any bounded linear operator defined on $X$, then does the existence of a bounded inverse $S=(I-T)^{-1}$ imply that $S$ can be represented as $S=\sum_{n=0}^\infty T^n$?

share|cite|improve this question
Just added some tags. – AD. Jan 22 '12 at 15:19
up vote 13 down vote accepted

No, not even in the finite-dimensional case. If $T$ is a linear map from $\mathbb{R}^n$ to itself with all eigenvalues $>1$ in absolute value, then $I-T$ is invertible, but $\sum T^n$ certainly does not converge.

share|cite|improve this answer
not even in dimension 1: 1-2 is invertible but $\sum_n 2^n$ does not converge – Plop Nov 11 '10 at 21:56

Well there are slightly weaker condition that improves the result you cite.

Suppose $T$ is a bounded operator with spectral radius $r(T)< 1$ Then by the spectral radius formula we have $$\lim_{n\to\infty}\|T^n\|^{1/n} =\inf_n\|T^n\|^{1/n}=r(T)< 1$$ which ensure the convergence of $$\sum_{n=0}^\infty\|T^n\|$$ which in turn, by the triangle inequality, bounds $$\left\|\sum_{n=0}^\infty T^n\right\|$$ Now, it is a standard exercise to show that $$\sum_{n=0}^\infty T^n=(I-T)^{-1}$$ If there is any doubt at all do not hesitate to ask.. **Edit:** By the Banach algebra inequality we have $\|T^n\|\le\|T\|^n$, which means that $$r(T)=\inf_k\|T^k\|^{1/k}\le\|T^n\|^{1/n}\le\|T\|$$ Hence $\|T\|<1$ implies not only $r(T)<1$, but also $r(T)\le\|T\|<1$. Also, this is not the case in the example of Robin above, because we also have $r(T)=\sup{|\lambda|:\lambda\in\sigma(T)}$ where $\sigma(T)$ is the spectrum of $T$ (the set of all $\lambda\in\mathbb{C}$ such that $\lambda I-T$ is not invertible) and the eigenvaules of $T$ is certainly in the spectrum. (Note that $r(T)$ is the radius of the smallest closed disc that contain $\sigma(T)$ - hence the name spectral radius).

share|cite|improve this answer
Where is the text??? – AD. Nov 11 '10 at 7:58
An extreme example is a nilpotent operator (i.e. $T^n=0$ for some $n$). – AD. Nov 11 '10 at 8:06
@AD: Fixed. I don't know why but it seems you have to type &lt; for <. – kennytm Nov 11 '10 at 9:07
@KennyTM♦ Great - Thanks! – AD. Nov 11 '10 at 9:37
BTW - I do not claim that this is my spectrum. This set called the spectrum is used in linear algebra, Fourier analysis, operator theory and functional analysis etc. I suggest that you look it up in Wikipedia. – AD. Nov 12 '10 at 14:41

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.