We analyze a toy class of two-player repeated games with two-sided incomplete information. In our model, two players are facing independent decision problems and each of them holds information that is potentially valuable to the other player. We study to what extent, and how, information can be...