Abstract:
This paper reconsiders the case of the rivalry between the HP and the Hamilton detrending methods, assessing it from a new angle, that of estimation of the trend itself. An extensive Monte Carlo study based on Euro area GDP is performed which, besides confirming and even reinforcing the superiority of the HP filter in terms of goodness of fit, demonstrates that, when the assessment criteria are designed to measure the success in terms of replicating long-run properties, the positions are switched, Hamilton's method dominating the HP filter. Since, however, both methods exhibit implausible features, a combination of them is employed to derive some empirical evidence concerning the Euro Area economy.
