Use this URL to cite or link to this record in EThOS: http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.606197
Title: Essays on applied time series econometrics
Author: Kurniawan, Ferry
ISNI:       0000 0004 5360 9354
Awarding Body: University of Warwick
Current Institution: University of Warwick
Date of Award: 2014
Availability of Full Text:
Access through EThOS:
Full text unavailable from EThOS. Restricted access.
Access through Institution:
Abstract:
The thesis consists of three self-contained essays. These are Business Cycles in ASEAN-5 Countries (Chapter 1), Nowcasting Indonesian Economy (Chapter 2), and Data Revisions in Indonesia (Chapter 3). In the first chapter, I investigate the extent to which business cycles coincide in five ASEAN (Association of Southeast Asian Nations) countries. I employ unobserved component models to decompose output into its trend and cycle. I find that the business cycles co-movements are high for some pairs of countries. However, the magnitude of the business cycles and the source of output fluctuations are different. The findings suggest that these countries may not yet be ready to step further beyond the formation of ASEAN Economic Community (AEC) by 2015 to a monetary union. In the second chapter, I exploit monthly indicators to forecast the current quarter Indonesian output growth (nowcast). In particular, I employ three nowcasting approaches; mixed-frequency factor model, bridging equation and MIDAS (MIxed DAta Sampling) regression, and evaluate their performance. The nowcasts are computed and evaluated on the basis of real-time and latest available data. In general, the encompassing tests recommend that neither mixed-factor model nowcasts nor MIDAS nowcasts encompass the other. Hence, I adopt a nowcast combination approach and find that the combination increases the accuracy relative to individual nowcasts. The last chapter focuses on real-time data issues. I construct a real-time data set for Indonesia, investigate the nature of data revisions, and assess the impact of the revisions on real-time output gap estimates and through this, the effect on monetary policy. The results suggest that the preliminary data is not an efficient forecast of the revised (final) value. By comparing the output gap estimates, extracted using the Hodrick-Prescott filter to both preliminary and final data, I find that output gap revisions is mainly due to data revisions, rather than due to one-sided nature of the filter. Further, I find that the potential impact of data revisions on monetary policy, measured by the difference between policy rate recommendations based on first released and revised data, can be substantial. Finally, I show that by taking into account data revisions, the reliability of real-time output gap estimates can be improved, and hence policy regret may be reduced.
Supervisor: Not available Sponsor: Bank Indonesia
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID: uk.bl.ethos.606197  DOI: Not available
Keywords: HB Economic Theory
Share: