Testing for contagion: a time-scale decomposition

The aim of the paper is to test for financial contagion by estimating a simultaneous equation modelsubject to structural breaks. For this purpose, we use the Maximum Overlapping Discrete WaveletTransform, MODWT, to decompose four asset returns into different scale components (eachassociated with a given frequency range). The decomposition will enable us to obtain the momentconditions necessary to (over)identify a structural form model with a single dummy and the onewith multiple dummies capturing shifts in the co-movement of asset returns occurring duringperiods of financial turmoil. A Montecarlo simulation exercise shows that test based on a singledummy structural form model has good size and power properties in detecting financial contagion.The empirical results for four East Asian stock markets show that, once we account forinterdependence through an (unobservable) common factor, there is no evidence of contagion butonly (limited) empirical support of hypersensitivity and extra-vulnerability during the 1997-1998financial turbulence.