Suppose the sequence $\{x_k\}_{k=1}^\infty\subset\mathbb{R}^n$ satisfies $\|x_{k+2}-x_{k+1}\|\le\theta\|x_{k+1}-x_k\|$ for all $k\ge1$, with $0<\theta<1$. Show that $\{x_k\}$ is a Cauchy sequence and hence converges.
So I did some rearranging to try and get something that worked and got this:
For $m,n>1$ such that $m<n$ $$\|x_n-x_m\|=\|x_n+\left(\sum_{i=m+1}^{n-1}(x_i-x_i)\right)-x_m\|$$ $$\le \sum_{i=m+1}^n\|x_i-x_{i-1}\|\text{ (triangle inequality)}$$ $$\le\left(\sum_{i=0}^{n-m}\theta^i\right)\|x_{m+1}-x_m\|$$
Is it sufficient to choose $N$ for any $\epsilon$ such that $\|x_{N+1}-x_N\|<\epsilon$ and since $\left(\sum_{i=0}^{n-m}\theta^i\right)<1$ I can use the above derivation?
A hint: Prove by induction that $$|x_{k+1}-x_k|\leq \theta^k|x_1-x_0|\qquad(k\geq0)\ ,$$ then use (as you did) the triangle inequality and the infinite series $\sum_{k\geq n}\theta^k$ to estimate $|x_{n+p}-x_n|$. Don't be afraid of ${|x_1-x_0|\over 1-\theta}$.