Relationship between the dollar, the price of copper and the IPSA indifferent time scales: An approach through Wavelet

In this study the causality relation between three relevant variables of the Chilean economy is analyzed. These variables are: the exchange rate between the Chilean peso and the US dollar, the price of copper and the Santiago Stock Exchange Index IPSA. To accomplish this, a decomposition by the wavelet transformed with Daubechies filter is realized and then a non linear Granger causality test is applied. We find that the results obtained from complete series are different from the ones obtained when the series were decomposed in different time scales. In general, there is bi-directional Granger causality only in the long term, but in the short term the results depend on the pair analyzed.