Title:
(Joint work with Marine Carrasco and Liang Hu)
Optimal Test for Markov Switching
Abstract:
We propose a new class of tests for the stability of parameters. We cover the class of Hamilton models, where regime changes are driven by an
unobservable Markov chain. We derive a class of information matrix-type tests and show that they are equivalent to the likelihood ratio test.
Hence, our tests are asymptotically optimal. Moreover these tests are easy to implement as they do not require the estimation of the model under
the alternative. They are also very general. Indeed, the underlying process driving the regime changes may have a finite or continuous state
space, as long as it is exogenous. The model itself need not be linear. It may be a GARCH model, for instance.
We use this test to investigate the presence of rational collapsing bubbles in stock markets. Using US data, we find evidence in favor of
nonlinearities, which are consistent with periodically collapsing bubbles.