Spatial correlation plays an important role in spatial economics. However, there is still room for improvement of this method. We address the question of whether the correlation may be trivial consequence of the trends that are entirely unrelated to spatial adjacency. Application of this analysis to selected economic data demonstrates that some trends are not sufficient to account for spatial correlation properties. Then we modify the spatial adjacency matrix using the detrended cross-correlation analysis method for stock data. The modified spatial matrix is introduced to spatial Dubin model. Application of this method to stock data demonstrates that the spatial DCCA coefficient matrix is an optional method.