2020International Journal of Finance & EconomicsRequires access

Cointegration tests at the quantiles

Marilena Furno

Open publisher page 5 citations

Abstract

Abstract A cointegration test for quantile regressions is proposed and implemented using Italian data. The test relies on auxiliary quantile regression to verify the stationarity of the residuals of the cointegrating equation. According to the problem under analysis, the cointegrating equation may or may not model a structural break, to verify cointegration with or without break. The existing test by Xiao, Journal of Econometrics (2009), 150, 248–260 is a fluctuation type test, which is closely related to Qu, Journal of Econometrics (2008), 146, 170–184 on structural break in quantile regressions. The link between the two tests makes unclear if the fluctuation test verifies cointegration, stability, or possibly cointegration and stability mixed together. Two real data case studies and a Monte Carlo experiment complete the analysis.

About this research paper

What this paper is about

Abstract A cointegration test for quantile regressions is proposed and implemented using Italian data. The test relies on auxiliary quantile regression to verify the stationarity of the residuals of the cointegrating equation. According to the problem under analysis, the cointegrating equation may or may not model a structural break, to verify cointegration with or without break. The existing test by Xiao, Journal of Econometrics (2009), 150, 248–260 is a fluctuation type test, which is closely related to Qu, Journal of Econometrics (2008), 146, 170–184 on structural break in quantile regressions. The link between the two tests makes unclear if the fluctuation test verifies cointegration, stability, or possibly cointegration and stability mixed together. Two real data case studies and a Monte Carlo experiment complete the analysis.

Why it matters

OpenAlex reports 5 citations for this work. Citation counts describe recorded attention and do not establish research quality.

Key contribution

A contribution statement is not available in the OpenAlex record.

Method / approach

Method details are not available in the OpenAlex metadata.

Main findings

Findings are not separately available in the OpenAlex metadata.

Limitations

Limitations are not available in the OpenAlex metadata.

Applications

Application details are not available in the OpenAlex metadata.

Available abstract

Abstract A cointegration test for quantile regressions is proposed and implemented using Italian data. The test relies on auxiliary quantile regression to verify the stationarity of the residuals of the cointegrating equation. According to the problem under analysis, the cointegrating equation may or may not model a structural break, to verify cointegration with or without break. The existing test by Xiao, Journal of Econometrics (2009), 150, 248–260 is a fluctuation type test, which is closely related to Qu, Journal of Econometrics (2008), 146, 170–184 on structural break in quantile regressions. The link between the two tests makes unclear if the fluctuation test verifies cointegration, stability, or possibly cointegration and stability mixed together. Two real data case studies and a Monte Carlo experiment complete the analysis.

Key concepts: Cointegration, Quantile, Econometrics, Quantile regression, Structural break, Monte Carlo method, Stability (learning theory), Economics

Related papers

Back to paper searchBrowse research topicsOriginal source
Cointegration tests at the quantiles — Research Paper | ScholarLens