A matrix version of the Wielandt inequality and its applications to statistics (Q1124947)

Let \(A\) be a positive definite matrix, and let \(X\) and \(Y\) be matrices such that \(X^*Y=0\). The authors prove the inequality \[ X^* AY(Y^* AY)^-Y^*AX \leq\left({a-b \over a+b} \right)^2X^*AX, \] where \(a\) and \(b\) are the largest and smallest eigenvalues of \(A\), respectively, and \(M^-\) stands for a generalized inverse of \(M\).