To solve the problem of determining for which invertible matrices \( M \), the polynomial \( p_M(x) - p_{M^{-1}}(x) \) is constant, we need to examine the properties of the characteristic polynomial of a matrix and its inverse. The characteristic polynomial of a matrix \( A \), \( p_A(x) \), is given by:
\(\det(xI - A)\)
where \( I \) is the identity matrix of the same dimension as \( A \).
For an invertible matrix \( M \), suppose the eigenvalues are \(\lambda_1, \lambda_2, \ldots, \lambda_n\). Then, the eigenvalues of \( M^{-1} \) are \(\frac{1}{\lambda_1}, \frac{1}{\lambda_2}, \ldots, \frac{1}{\lambda_n}\).
The characteristic polynomial of \( M \) is:
\(p_M(x) = (x - \lambda_1)(x - \lambda_2) \cdots (x - \lambda_n)\)
And the characteristic polynomial of \( M^{-1} \) is:
\(p_{M^{-1}}(x) = (x - \frac{1}{\lambda_1})(x - \frac{1}{\lambda_2}) \cdots (x - \frac{1}{\lambda_n})\)
We need \( p_M(x) - p_{M^{-1}}(x) \) to be a constant. For this to happen, under normal circumstances, the eigenvalues of \( M \) and \( M^{-1} \) must combine such that their differences are constant for all values of \( x \).
Upon evaluation, the matrices that satisfy the condition are:
In these cases, the properties of the eigenvalues and their reciprocals allow the polynomial difference to remain constant (typically zero for similar matrices' inverse reciprocal pairs).