linear and nonlinear causality tests

[upani]

Dear All,

Is there any rocedure to perform linear and nonlinear causality test with frequency bands. In the non linear causality the model can take any of the specification which includes Semi-additive, P-general Taylor based, ANN-based.

This is also done in Wavelet decomposition.

Please suggest me some procedure to perform these test.

With sincere regards,
Upananda

[TomDoan]

Most papers I've seen on non-linear causality range from uninteresting to simply wrong. If you have a specific paper which has demonstrated a technique that is both correct and answers an interesting question about actual data, please cite it.