Crandi

Pdf Error Correction Factor Models For High Dimensional Cointegrated T

Leo Migdal
-
pdf error correction factor models for high dimensional cointegrated t

Academia.edu no longer supports Internet Explorer. To browse Academia.edu and the wider internet faster and more securely, please take a few seconds to upgrade your browser. Cointegration inference is often built on the correct specification for the short-run dynamic vector autoregression. However, this specification is unknown a priori. A too small lag length leads to erroneous inference due to misspecification, while using too many lags leads to dramatic increase in the number of parameters, especially when the dimension of time series is... In this paper, we develop a new methodology which adds an error correction term for long-run equilibrium to a latent factor model for modeling short-run dynamic relationship.

Two eigenanalysis based methods for estimating, respectively, cointegration and latent factor process consist of the cornerstones of the inference. The proposed error correction factor model does not require to specify the short-run dynamics explicitly, and is particularly effective for high-dimensional cases when the standard error-correction suffers from overparametrization. It also increases the predictability over a pure factor model. Asymptotic properties of the proposed methods are established when the dimension of the time series is either fixed or diverging slowly as the length of time series goes to infinity. Illustration with both simulated and real data sets is also reported. This paper considers cointegration tests for dynamic systems where the number of variables is large relative to the sample size.

Typical examples include tests for unit roots in panels where the units are linked by complicated dynamic relationships. It is well known that conventional cointegration tests based on a parametric (vector autoregressive) representation of the system break down if the number of variables approaches the number of time periods. To sidestep this difficulty we propose nonparametric cointegration tests based on eigenvalue problems that are asymptotically free of nuisance parameters. It turns out that the nonparametric tests outperform their parametric (likelihood-ratio based) counterparts by a clear margin. RePEc: Research Papers in Economics, 2000

People Also Search

Academia.edu No Longer Supports Internet Explorer. To Browse Academia.edu And

Academia.edu no longer supports Internet Explorer. To browse Academia.edu and the wider internet faster and more securely, please take a few seconds to upgrade your browser. Cointegration inference is often built on the correct specification for the short-run dynamic vector autoregression. However, this specification is unknown a priori. A too small lag length leads to erroneous inference due to m...

Two Eigenanalysis Based Methods For Estimating, Respectively, Cointegration And Latent

Two eigenanalysis based methods for estimating, respectively, cointegration and latent factor process consist of the cornerstones of the inference. The proposed error correction factor model does not require to specify the short-run dynamics explicitly, and is particularly effective for high-dimensional cases when the standard error-correction suffers from overparametrization. It also increases th...

Typical Examples Include Tests For Unit Roots In Panels Where

Typical examples include tests for unit roots in panels where the units are linked by complicated dynamic relationships. It is well known that conventional cointegration tests based on a parametric (vector autoregressive) representation of the system break down if the number of variables approaches the number of time periods. To sidestep this difficulty we propose nonparametric cointegration tests...