Show that if $\sum\limits_{k=1}^nx_k=\sum\limits_{k=1}^n{x_k}^2=\dots=\sum\limits_{k=1}^n{x_k}^n$ where $x_k\in\mathbb{R}$ then $\prod\limits_{k=1}^n x_k\le1$.
For example, when $n=3$, the system of equations is:
$$x_1+x_2+x_3$$ $$={x_1}^2+{x_2}^2+{x_3}^2$$ $$={x_1}^3+{x_2}^3+{x_3}^3$$
where $x_1, x_2, x_3 \in\mathbb{R}$. We are to show that $x_1 x_2 x_3 \le1.$
I made up this question. I have found a solution, which I will post, but my solution is rather complicated, and I am hoping for a more elegant solution.
Here is another solution, which only uses the following assumption:
Let $p$ denote the common value of the sums in $\text{(*)}$. Clearly, $p \geq 0$. Moreover, by Cauchy–Schwarz inequality,
\begin{align*} p = \sum_{k=1}^{n} x_k^2 = \frac{1}{n} \left( \sum_{k=1}^{n} x_k^2 \right)\left( \sum_{k=1}^{n} 1 \right) \geq \frac{1}{n} \left( \sum_{k=1}^{n} x_k \right)^2 = \frac{p^2}{n}. \end{align*}
Hence we get $0 \leq p \leq n$. Then by AM–GM inequality,
$$ |x_1 \cdots x_n|^{\frac{2}{n}} \leq \frac{x_1^2 + \cdots + x_n^2}{n} \leq 1, $$
which is enough to conclude that $|x_1 \cdots x_n| \leq 1$.