Consider the unit operator
$T: \ell^2(\mathbb Z) \to \ell^2(\mathbb Z), \, T(x_n)_{n \in \mathbb Z} = (x_{n+1})_{n \in \mathbb Z}$. I'm trying to understand the proof that $\sigma(T) = S^1$ given here.
First we have the following results:
$\forall \lambda \in \mathbb C, T - \lambda I$ is injective.
If $E$ is a Banach Space, $S \in \mathcal B(E), z \in \mathbb C, z \neq 0, \, \sigma(zS) = z\sigma(S)$.
If $S,U \in \mathcal B(E), \, \sigma(U^{-1}SU) = \sigma(S)$.
Given $z \in S^1$, we define the bounded operator in $\ell^2(\mathbb Z)$, $$M_z(x_n)_{n \in \mathbb Z} = (...,\overline z ^{2}x_{-2}, \overline z ^{-1}x_{-1},x_0, zx_1, z^2 x_2,...).$$
Then, we have that $M_z^{-1}T M_z= zT$. So, for each $z \in S^1$,
$$ \sigma(T) = \sigma(M_z ^{-1}T M_z) = \sigma(zT) = z\sigma(T). $$
In the proof given in the above link, Jonas Meyer used that exists $z_0 \in \sigma(T)$ with $|z_0| = 1.$ However, I couldn't see why this happens.
QUESTION: If $|\lambda| < 1$, how one can proof that $T - \lambda I$ is invertible?
In particular, the above question implies in $\sigma(T) S^1$, since $\sigma(T) \subset B[0, 1]$. With this, I can conclude the argument, as follows:
Since $\sigma(T) \neq \emptyset$, then $\exists z_0 \in \sigma(T)$. Given $z \in S^1$, we have that $\sigma(T) = z\overline{z_0}\sigma(T)$. Hence,
$$ z_0 \in \sigma(T) \Rightarrow z = z(\overline{z_0}z_0) (z\overline{z_0})z_0 \in z\overline{z_0}\sigma(T) = \sigma(T) \Rightarrow S^1\subset \sigma(T).$$
Help?