Multifrequency network for SADC exchange rate markets using EEMD-based DCCA

We used the detrended cross-correlation analysis (DCCA) method based on ensemble empirical mode decomposition (EEMD) to study the dynamic interdependence structure of daily domestic currency to US dollar exchange rates of 15 Southern African Development Community (SADC) exchange rate markets. We first decompose all series into intrinsic mode functions using EEMD and reconstruct the series into three frequency modes: high-, medium- and low frequency, and residue. The DCCA method was used to analyze the cross-correlation between the various frequencies, residues and original series. These were meant to address the nonlinearity and nonstationarity in observed exchange rate data. Finally, we formed a correlation network from the cross-correlation coefficients in all cases which revealed rich than would have been obtained from the original series. We observed similarities between the nature of cross-correlation between high-frequency series mimic the original series and the significant cross-correlation among the long-term trend of most SADC countries exchange rate markets. The innovation of this paper is to combine EEMD with DCCA to study the multifrequency cross-correlations of exchange rate markets, which can provide policymakers a deeper understanding of the dynamics of exchange rate markets toward the formation of currency unions.