Can Indeterminacy Resolve the Consumption Correlation Puzzle?

Do international business cycles only represent optimal responses by rational agents to erratic changes in technology, or are they also influenced by factors that are unrelated to fundamentals? This paper addresses this issue by exploring the role of such non-fundamental factors in improving open-economy real business cycle (RBC) models. One major shortcoming of existing RBC models is the "consumption correlation puzzle" - models tend to generate cross-country consumption correlations that are too high and output correlations that are too low. I show that with empirically supported level of increasing returns, an otherwise standard model possesses multiple equilibria and indeterminate convergent paths to the steady states, which allows for "self-fulfilling beliefs" to influence the economy. The model displays time series properties that in many ways match the data better than the conventional model. It is especially successful in generating realistic consumption and output correlations.