Let $A$ be a $n \times n$ matrix over the field $\mathbb C$ and $X \in \mathbb C^{n}$. Let $k$ be the least positive integer such that the set of vectors $\{X, AX, A^{2} X, \dots , A^{k} X \}$ are lineary dependent.
Now let us consider a relation $\sum_{i=1}^{k} c_{i} A^{i} X = 0$ and a polynomial $g(t) = \sum_{i=1}^{k} c_{i} t^{i}$. Then I have proved that each root of the equation $g(t) = 0$ is an eigenvalue of $A$ corresponding to the eigenvector which is in the span of $\{X, AX, A^{2} X, \dots , A^{k-1} X \}$ though $g(x)$ need not be an annihilating polynomial of $A$. But when $\deg\ g(x) = n$, I observed that $g(x)$ becomes characteristics polynomial of $A$ according to the example which I constructed.
But I don't understand why is this happening. Can we construct an example which disprove it or is it a true fact?
Please help me in understanding this fact.Thank you in advance.