# Cointegration panel data

Hello,
Can someone explan to me how the pco package work and what do the results mean?

pedroni99(gdi, gds)
\$CALL
pedroni99(Y = gdi, X = gds)

\$METHOD
[1] "Pedroni(1999) panel tests for cointegration"

\$STATISTIC
empirical standardized
nipanel 34.688809 -0.024554841
rhopanel -28.703846 0.403612168
tpanelnonpar -8.971574 -0.529853094
tpanelpar -7.637772 0.538385663
rhogroup -43.106818 0.978667063
tgroupnonpar -10.575789 -0.764380042
tgrouppar -9.958862 0.001412969

`pedroni99` is the `bivariate` function in the `pco` package, that takes two matrices as its arguments. `gdi` is a data set of Gross domestic investment as a share of GDP (in percent) for 25 high income OECD countries, 2013-2017. `gds` is Gross domestic savings as a share of GDP (in percent) for 25 high income OECD countries, 1973-2013.

The return values

• \$CALL echos the invocation of the function
• \$METHOD identifies the source of the algorithm
• \$STATISTICS returns nine test statistics

The test statistics are described in Pedroni, Peter, 1999. "Critical Values for Cointegration Tests in Heterogeneous Panels with Multiple Regressors," Oxford Bulletin of Economics and Statistics, Department of Economics, University of Oxford, vol. 61(0), pages 653-70, Special I., the abstract of which states

Asymptotic distributions and critical values are computed for several residualābased tests of the null of no cointegration in panels for the case of multiple regressors, including regressions with individualāspecific fixed effects and time trends. The associated cointegrating vectors and the dynamics of the underlying error processes are permitted considerable heterogeneity across individual members of the panel.

and the full text of which should be consulted for additional information.

This topic was automatically closed 21 days after the last reply. New replies are no longer allowed.

If you have a query related to it or one of the replies, start a new topic and refer back with a link.