Generalising the RBC methodology for testing macro models (with Konstantinois Theodoridis)
|Speaker:||Patrick Minford, Cardiff University|
|Date:||Friday 18 February 2005|
|Location:||Lecture Room D, Streatham Court|
We take a large macro model of the UK (the Liverpool model) and estimate it by FIML to obtain the best estimates conditional on the data. The question then is whether the model can replicate the 'facts of the business cycle'. We represent these facts by a VECM-GARCH process fitted to the main 5 macro variables. We then check via the bootstrap whether the Liverpool model is rejected by this representation.