Money Creation in a Random Matching Model

I study money creation in versions of the Trejos-Wright (1995) and Shi (1995) models with indivisible money and individual holdings bounded at two units. I work with the same class of policies as in Deviatov and Wallace (2001), who study money creation in that model. However, I consider an alternative notion of implementability - the ex ante pairwise core. I compute a set of numerical examples to determine whether money creation is beneficial. I find beneficial effects of money creation if individuals are sufficiently risk averse (obtain sufficiently high utility gains from trade) and impatient.