Quadratic negligibility and the asymptotic normality of operator normed sums (Q1209877)

Let \(X_ 1\), \(X_ 2,\dots\) be independent and identically distributed \(d\)-dimensional random vectors (column); \(V_ n=\sum^ n_{k=1}X_ kX^ T_ k\) (``\(T\)'' denotes a vector transpose). It is shown that \[ \max_{1\leq k\leq n}X^ T_ kV_ n^{-1}X_ k@>\mathbb{P}>>0,\;n\to\infty, \] if and only if \(X_ 1\) belongs to the operator normed domain of attraction of the multivariate normal distribution.