Thanks to Kit Baum, a new package kapetanios is now available for download from SSC.
To install, type in Stata:
ssc install kapetanios
kapetanios implements Kapetanios' (2005) unit root test, which allows for up to 5 structural breaks with break dates endogenously estimated from the data. Under the alternative hypothesis of trend stationarity, breaks may occur in the level and/or trend of the series. To reflect this, three model specifications are available: intercept breaks only (Model A), trend breaks only (Model B), and both intercept and trend breaks (Model C, the default), each with tabulated critical values.
The test regression augments the standard ADF equation with structural break dummies and lagged differences of the dependent variable to control for serial correlation. To determine the appropriate lag length, kapetanios supports four selection methods: the sequential t-test of Ng and Perron (1995) as used in the original paper (default), a Breusch-Godfrey LM test based general-to-specific (GTS) procedure, AIC, and BIC. For AIC and BIC, a post-selection Breusch-Godfrey test is automatically performed and a warning is issued if residual autocorrelation remains.
For details and worked examples, see the help file:
help kapetanios
The latest version of the package is also available from the author's site:
net install kapetanios, from(https://www.eruygurakademi.com/stata/)
Comments, suggestions, and bug reports are welcome.
Best regards,
H. Ozan Eruygur
AHBV University, Department of Economics (Ankara, Turkey)
Eruygur Akademi ve Danışmanlık
[email protected]
https://www.eruygurakademi.com
To install, type in Stata:
ssc install kapetanios
kapetanios implements Kapetanios' (2005) unit root test, which allows for up to 5 structural breaks with break dates endogenously estimated from the data. Under the alternative hypothesis of trend stationarity, breaks may occur in the level and/or trend of the series. To reflect this, three model specifications are available: intercept breaks only (Model A), trend breaks only (Model B), and both intercept and trend breaks (Model C, the default), each with tabulated critical values.
The test regression augments the standard ADF equation with structural break dummies and lagged differences of the dependent variable to control for serial correlation. To determine the appropriate lag length, kapetanios supports four selection methods: the sequential t-test of Ng and Perron (1995) as used in the original paper (default), a Breusch-Godfrey LM test based general-to-specific (GTS) procedure, AIC, and BIC. For AIC and BIC, a post-selection Breusch-Godfrey test is automatically performed and a warning is issued if residual autocorrelation remains.
For details and worked examples, see the help file:
help kapetanios
The latest version of the package is also available from the author's site:
net install kapetanios, from(https://www.eruygurakademi.com/stata/)
Comments, suggestions, and bug reports are welcome.
Best regards,
H. Ozan Eruygur
AHBV University, Department of Economics (Ankara, Turkey)
Eruygur Akademi ve Danışmanlık
[email protected]
https://www.eruygurakademi.com
