Improved methods for calculating and estimating maximal stop-loss premiums