On the Measurement of the International Propagation of Shocks

In this paper I offer an alternative identification assumption that allows one to test for changing patterns regarding the international propagation of shocks when endogenous variables, omitted variables, and heteroskedasticity are present in the data. Using this methodology, I demonstrate that the propagation mechanisms of 36 stock markets remained relatively stable throughout the last three major international crises which have been associated with 'contagion' (i.e., Mexico 1994, Hong Kong 1997, and Russia 1998). These findings cast considerable doubt upon theories that suggest that the propagation of shocks is crisis contingent, and driven by endogenous liquidity issues, multiple equilibria, and political contagion. Rather, these findings would seem to support theories that identify such matters as trade, learning, and aggregate shocks as the primary transmission mechanisms in this process.