Abstracts - faqs.org

Abstracts

Economics

Search abstracts:
Abstracts » Economics

Sparse direct methods for model simulation

Article Abstract:

Newton-type sparse direct methods offer valuable solutions for the simulation and evaluation of large macroeconomic models with forward-looking variables. These methods actually involve subdividing a stacked model into recursive submodels and finding the solution through the help of Newton algorithm. However, the model's original block pattern must not be altered during the process. Such methods are highly efficient under a linear system wherein an LU factorization is used. Other solution methods, particularly the Gauss-Seidel algorithm may also be used in determining the L and U matrices.

Author: Gilli, Manfred, Pauletto, Giorgio
Publisher: Elsevier B.V.
Publication Name: Journal of Economic Dynamics & Control
Subject: Economics
ISSN: 0165-1889
Year: 1997
Economics, Models, Analysis, Evaluation, Macroeconomics, Recursive functions, Linear systems

User Contributions:

Comment about this article or add new information about this topic:

CAPTCHA


Krylov methods for solving models with forward-looking variables

Article Abstract:

An analysis of Krylov method use in simulations of large macroeconomic models with forward-looking variables reveals that the methods feature low computational complexity and low storage requirements. The study determines an alternative to the exact Newton method with sparse Gaussian elimination and finds that inexact Newton algorithms using Krylov methods viable alternatives.

Author: Gilli, Manfred, Pauletto, Giorgio
Publisher: Elsevier B.V.
Publication Name: Journal of Economic Dynamics & Control
Subject: Economics
ISSN: 0165-1889
Year: 1998
Econometrics & Model Building, Management Science, Business models

User Contributions:

Comment about this article or add new information about this topic:

CAPTCHA


An algorithm competition: first-order iterations versus Newton-based techniques

Article Abstract:

A comparison of Newton-based techniques and first order iterations used in solving nonlinear forward-looking models reveals that Newton-based techniques are considerably faster and less prone to simulation failure. The study compares the two techniques by solving them on MULTIMOD and finds that the Newton-Raphson iterative method is better than the Fair-Taylor algorithm.

Author: Laxton, Douglas, Juillard, Michel, McAdam, Peter, Pioro, Hope
Publisher: Elsevier B.V.
Publication Name: Journal of Economic Dynamics & Control
Subject: Economics
ISSN: 0165-1889
Year: 1998
Research and Development in the Physical, Engineering, and Life Sciences, Statistics, Statistics (Mathematics)

User Contributions:

Comment about this article or add new information about this topic:

CAPTCHA


Subjects list: Methods, Econometrics, Mathematical models, Simulation methods, Simulation
Similar abstracts:
  • Abstracts: Uncertainty aversion and preference for randomisation. A theorem on portfolio separation with general preferences
  • Abstracts: A model of multiproduct price competition. On the dynamic efficiency of Bertrand and Cournot equilibria. Comparing Cournot and Bertrand in a homogenous product market
  • Abstracts: Microfoundations for a stable demand for money function. Microfoundations and the demand for money. Which road leads to stable money demand?
  • Abstracts: Coalition-proof implementation. Approximately competitive equilibria in large finite economies
  • Abstracts: Labor market segmentation and the implications for public sector retrenchment programs. Learning from transition economies: assessing labor market policies across Central and Eastern Europe
This website is not affiliated with document authors or copyright owners. This page is provided for informational purposes only. Unintentional errors are possible.
Some parts © 2025 Advameg, Inc.