返回Google圖書搜尋

Measuring the Output Gap in Real-Time

其他書名	Evidence for Switzerland
出版	Universität St. Gallen, 2016
URL	http://books.google.com.hk/books?id=7M6eAQAACAAJ&hl=&source=gbs_api

註釋In this thesis I use a linear opinion pool to construct real-time ensemble nowcast densities for the Swiss output gap over an out-of-sample period from 2003q1 to 2015q4. The model space consists of a large number of bivariate vector autoregressive specifications for inflation and the output gap, with each specification using a different univariate output gap estimation method, lag order and structural break. The ensemble nowcast densities for the output gap are constructed by combining the predictive densities of the individual VAR specifications weighted by their ability to provide accurate density forecasts for inflation. I assess the robustness of the real-time output gap nowcasts by the linear opinion pool's predictive power for inflation over the evaluation period and by the size of the ex post revisions to the output gap nowcasts due to data revisions and newly added data. I find that the linear opinion pool approach does not generate more accurate density or point inflation forecasts than a number of correctly specified univariate benchmark models. Further, the real-time estimate of the output gap provided by the opinion pool is not more robust to ex post revisions than the real-time estimates of the individual univariate output gap estimation methods. The results put into question the robustness of the real-time output gap nowcasts provided by the linear opinion pool over the evaluation period. I also find that revisions to Swiss GDP price deflator data are large and complicate inflation forecasting.