An eight-region, stochastic, general equilibrium model that incorporates global money and financial markets is used, first, to identify the distributions of regional supply side shocks in 1954-2016 from real GDP data. Second, in Monte Carlo simulations the model is subjected to random draws from these distributions in all regions simultaneously, under a range of combinations of monetary targets. These experiments begin with a comparison of the volatilities of output and welfare (defined as the purchasing power of disposable income at regional prices) under current de facto and de jure monetary targets. The results show the de facto targets offer significant improvement. A further analysis of targeting games suggests a Nash equilibrium on output risk is dominated by nominal GDP targeting, while the corresponding Nash equilibrium on welfare risk has predominant inflation targeting (IT). Notwithstanding IMF policy to convert developing country central banks to inflation targeting, a fallacy of composition is observed in the spread of IT internationally. This spread tends to raise the volatility of welfare in the established open economies.