These and other papers out of the same group in Stockholm invoke the name "Parrondo" but do not construct actual Parrondo games. I have attached the original note from Nature. You will note that Parrondo game "B" is composed of two sub-games, one of which has to have a positive expectation. So the Parrondo two game paradox actually consists of three games. It is not applicable to trading because if you have three uncorrelated games, one of which is a winner, you are better off just playing the winner combined with cash rather than with one or both of the losing games.