A note on a problem of Abramovich, Aliprantis and Burkinshaw (Q655975)

An operator \(Q\) between Banach spaces \(X\) and \(Y\) is called a quasi-affinity if \(Q\) is one-to-one and has dense range. An operator \(T\in L(X)\) is said to be a quasi-affine transform of an operator \(S\in L(Y)\) if there exists a quasi-affinity \(Q\in L(X,Y)\) such that \(QT = SQ\). If both \(T\) and \(S\) are quasi-affine transforms of each other, the operators \(T\in L(X)\) and \(S\in L(Y)\) are called quasi-similar. Two operators \(S\in L(E)\), \(T\in L(F)\) are said to be positive quasi-affinities if there exist positive quasi-affinities \(P\in L(E,F)\) and \(Q\in L[F,E)\) such that \(TP = PS\) and \(QT = SQ\). One of the main results of the paper is the following. Theorem. Let \(B,T\) be positive operators on a Banach lattice \(E\) such that \(B\) is compact friendly and \(T\) is locally quasi-nilponent at a positive element of \(E\). If \(B\) is positive quasi-similar to \(T\), then \(T\) has a nontrivial closed invariant ideal. Moreover, the authors prove the following interesting result. Theorem. Let \(B\) be a positive operator on a Banach lattice \(E\). If \(B\) is positively quasi-similar to a positive operator \(R\) on \(E\) which is dominated by a positive compact operator \(K\) on \(E\), then \(B\) has a nontrivial closed invariant subspace.