A Proposal for a Flexible Trend Specification in DSGE Models

In this paper I propose a flexible trend specification for estimating DSGE models on log differences. I demonstrate this flexible trend specification on a New Keynesian DSGE model of two economies, which I consequently estimate on data from the Czech economy and the euro area, using Bayesian techniq...

Full description

Bibliographic Details
Main Author: Slanicay Martin
Format: Article
Language:English
Published: Sciendo 2016-06-01
Series:Review of Economic Perspectives
Subjects:
Online Access:https://doi.org/10.1515/revecp-2016-0006
_version_ 1818568341512519680
author Slanicay Martin
author_facet Slanicay Martin
author_sort Slanicay Martin
collection DOAJ
description In this paper I propose a flexible trend specification for estimating DSGE models on log differences. I demonstrate this flexible trend specification on a New Keynesian DSGE model of two economies, which I consequently estimate on data from the Czech economy and the euro area, using Bayesian techniques. The advantage of the trend specification proposed is that the trend component and the cyclical component are modelled jointly in a single model. The proposed trend specification is flexible in the sense that smoothness of the trend can be easily modified by different calibration of some of the trend parameters. The results suggest that this method is capable of finding a very reasonable trend in the data. Moreover, comparison of forecast performance reveals that the proposed specification offers more reliable forecasts than the original variant of the model.
first_indexed 2024-12-14T06:34:29Z
format Article
id doaj.art-0941217f0dd8481499bdd5717e9d3cb4
institution Directory Open Access Journal
issn 1804-1663
language English
last_indexed 2024-12-14T06:34:29Z
publishDate 2016-06-01
publisher Sciendo
record_format Article
series Review of Economic Perspectives
spelling doaj.art-0941217f0dd8481499bdd5717e9d3cb42022-12-21T23:13:24ZengSciendoReview of Economic Perspectives1804-16632016-06-01162738510.1515/revecp-2016-0006revecp-2016-0006A Proposal for a Flexible Trend Specification in DSGE ModelsSlanicay Martin0Masaryk University, Faculty of Economics and Administration, Department of Economics, Lipová 41a, Brno 602 00In this paper I propose a flexible trend specification for estimating DSGE models on log differences. I demonstrate this flexible trend specification on a New Keynesian DSGE model of two economies, which I consequently estimate on data from the Czech economy and the euro area, using Bayesian techniques. The advantage of the trend specification proposed is that the trend component and the cyclical component are modelled jointly in a single model. The proposed trend specification is flexible in the sense that smoothness of the trend can be easily modified by different calibration of some of the trend parameters. The results suggest that this method is capable of finding a very reasonable trend in the data. Moreover, comparison of forecast performance reveals that the proposed specification offers more reliable forecasts than the original variant of the model.https://doi.org/10.1515/revecp-2016-0006dsge modeltrend specificationbayesian estimation
spellingShingle Slanicay Martin
A Proposal for a Flexible Trend Specification in DSGE Models
Review of Economic Perspectives
dsge model
trend specification
bayesian estimation
title A Proposal for a Flexible Trend Specification in DSGE Models
title_full A Proposal for a Flexible Trend Specification in DSGE Models
title_fullStr A Proposal for a Flexible Trend Specification in DSGE Models
title_full_unstemmed A Proposal for a Flexible Trend Specification in DSGE Models
title_short A Proposal for a Flexible Trend Specification in DSGE Models
title_sort proposal for a flexible trend specification in dsge models
topic dsge model
trend specification
bayesian estimation
url https://doi.org/10.1515/revecp-2016-0006
work_keys_str_mv AT slanicaymartin aproposalforaflexibletrendspecificationindsgemodels
AT slanicaymartin proposalforaflexibletrendspecificationindsgemodels