Testing for equivalence of pre-trends in Difference-in-Differences estimation