Quantifying different degrees of coupling in detrended cross-correlation analysis

Detrended cross-correlation analysis provides a scaling exponent that should characterize the power-law cross-correlation of two simultaneously recorded series. This exponent by itself is not able to guarantee the presence of cross-correlation, being strongly influenced by the auto-correlation properties of the single series. Through the use of ?DCCA coefficients and simulation with ARFIMA models we built families of curves that can be used as templates to correctly detect the power-law behaviour and quantify the degree of coupling between series with any degree of auto-correlation.