A New Ridge Regression Causality Test in the Presence of Multicollinearity

This paper analyzes and compares the properties of the most commonly applied versions of the Granger causality (GC) test to a new ridge regression GC test (RRGC), in the presence of multicollinearity. The investigation has been carried out using Monte Carlo simulations. A large number of models have been investigated where the number of observations, strength of collinearity, and data generating processes have been varied. For each model we have performed 10000 replications and studied seven different versions of the test. The main conclusion from our study is that the traditional OLS version of the GC test over-rejects the true null hypothesis when there are relatively high (but empirically common levels of) multicollinearity, while it is established that the new RRGC test will remedy or substantially decrease this problem.