|
11. |
Interactions of real GNP business cycles in a three‐country time‐series model |
|
Journal of Forecasting,
Volume 12,
Issue 3‐4,
1993,
Page 331-344
Masanao Aoki,
Preview
|
PDF (683KB)
|
|
摘要:
AbstractThe paper describes a procedure for examining short‐run dynamic interactions among macroeconomic models by constructing aggregate state‐space submodels for dynamic modes corresponding with short‐run response patterns. Using the quarterly real GNP from the US, Germany, and Japan for 1974. III to 1991. I we examine dynamic interactions in the frequency ranges roughly comparable with a range of business cycle frequencies. We find that there is no (worldwide) shock common to the three countries even though the German and Japanese real GNP are affected by a common
ISSN:0277-6693
DOI:10.1002/for.3980120312
出版商:John Wiley&Sons, Ltd.
年代:1993
数据来源: WILEY
|
12. |
Fixed versus time‐varying transfer functions for modelling business cycles |
|
Journal of Forecasting,
Volume 12,
Issue 3‐4,
1993,
Page 345-364
Per‐Olov Edlund,
Henning T. Søgaards,
Preview
|
PDF (994KB)
|
|
摘要:
AbstractTransfer function models can be used to model the relationship between leading indicators and the business cycle. The traditional transfer function approach assumes that the pure delay (lead) as well as the other parameters of the model are constant. Results from studies on Swedish business cycle data indicate that the relationship between the leading indicators and the business cycle may be time varying. In this paper the traditional approach is compared to a recursive estimation procedure that allows the model to change over time. The results show that the recursive procedure is useful and that the estimates of the model parameters indeed vary over time.
ISSN:0277-6693
DOI:10.1002/for.3980120313
出版商:John Wiley&Sons, Ltd.
年代:1993
数据来源: WILEY
|
13. |
Forecasting with generalized bayesian vector auto regressions |
|
Journal of Forecasting,
Volume 12,
Issue 3‐4,
1993,
Page 365-378
K. Rao Kadiyala,
Sune Karlsson,
Preview
|
PDF (837KB)
|
|
摘要:
AbstractThe effects of using different distributions to parameterize the prior beliefs in a Bayesian analysis of vector autoregressions are studied. The well‐known Minnesota prior of Litterman as well as four less restrictive distributions are considered. Two of these prior distributions are new to vector autoregressive models. When the forecasting performance of the different parameterizations of the prior beliefs are compared it is found that the prior distributions that allow for dependencies between the equations of the VAR give rise to better forecast
ISSN:0277-6693
DOI:10.1002/for.3980120314
出版商:John Wiley&Sons, Ltd.
年代:1993
数据来源: WILEY
|
14. |
Masthead |
|
Journal of Forecasting,
Volume 12,
Issue 3‐4,
1993,
Page -
Preview
|
PDF (90KB)
|
|
ISSN:0277-6693
DOI:10.1002/for.3980120301
出版商:John Wiley&Sons, Ltd.
年代:1993
数据来源: WILEY
|
|