Structural Break Zivot-Andrews Unit Root Test
Also known as: Zivot-Andrews test, ZA unit root test, endogenous structural break unit root test, ZA breakpoint test
The Zivot-Andrews test is an endogenous structural break unit root test that determines the break point from the data rather than imposing it externally. It tests for a unit root against the alternative of stationarity around a single structural break — in the mean, the trend, or both — choosing the break date that provides the strongest evidence against the null.
Read the full method
Sign in with a free account to read this section.
Method map
The neighbourhood of related methods — select a node to explore.
When to use it
Use the Zivot-Andrews test when you suspect a time series may have experienced a single structural break of unknown date that could confound a standard ADF or PP test. It is appropriate for macroeconomic and financial series that span events such as crises, regime changes, or policy shifts. It is not appropriate when multiple breaks are likely — use Bai-Perron or Lee-Strazicich tests in that case. It should not replace standard unit root tests when there is no theoretical reason to expect a break, and it cannot accommodate breaks in volatility (use ARCH-family models for that).
Strengths & limitations
- Endogenously determines the break date, eliminating pre-selection bias.
- More powerful than ADF and PP tests in the presence of a single structural break.
- Covers three model types (level shift, trend shift, or both), giving flexibility.
- Widely implemented in Stata (zandrews), R (strucchange, tseries), and EViews.
- Critical values are tabulated for common sample sizes and significance levels.
- Designed for only one structural break; multiple breaks reduce test power and may yield a spurious break date.
- The trimming convention (typically 15 %) means breaks very near the start or end of the sample cannot be detected.
- Non-standard critical values make inference sensitive to the model choice (A, B, or C).
- Large lag order (needed for heavily autocorrelated series) can reduce power substantially.
Frequently asked
How is the Zivot-Andrews test different from the standard ADF test?
The ADF test assumes parameter stability throughout the sample. If the series has a structural break, the ADF has low power and tends to fail to reject the unit root null even when the series is stationary. The Zivot-Andrews test searches for the break date endogenously and uses it to augment the regression, restoring power in that setting.
Which model — A, B, or C — should I use?
Model A is appropriate when you expect only a level (intercept) shift; Model B when only the trend slope changes; Model C when both level and slope may shift. In practice Model C is the most conservative and commonly preferred when the type of break is uncertain. Choose based on economic reasoning, not empirical convenience.
What if the test identifies a break date that makes no historical sense?
The identified date is the point with the most extreme t-statistic, not necessarily the true change point. If the date is economically implausible, it is a warning sign — the break may not be real, or the series may have multiple breaks that the test cannot accommodate. Consider Bai-Perron or Lee-Strazicich tests as alternatives.
Can I use the Zivot-Andrews test before cointegration analysis?
Yes. Cointegration tests (Engle-Granger, Johansen) assume the variables are integrated of the same order. If standard unit root tests are unreliable due to suspected breaks, the Zivot-Andrews test provides a more appropriate pre-test for integration order.
What software implements the Zivot-Andrews test?
In Stata use the zandrews command; in R use ur.za() from the urca package or the strucchange package; EViews includes it under the Unit Root test dialog when the break specification option is selected.
Sources
- Zivot, E., & Andrews, D. W. K. (1992). Further evidence on the great crash, the oil-price shock, and the unit-root hypothesis. Journal of Business & Economic Statistics, 10(3), 251–270. DOI: 10.1080/07350015.1992.10509904 ↗
- Perron, P. (1989). The great crash, the oil price shock, and the unit root hypothesis. Econometrica, 57(6), 1361–1401. DOI: 10.2307/1913712 ↗
How to cite this page
ScholarGate. (2026, June 3). Structural Break Zivot-Andrews Unit Root Test. ScholarGate. https://scholargate.app/en/econometrics/structural-break-zivot-andrews-test
Which method?
Set this method beside its closest kin and read them side by side — the library lays the books on the table; the choice is yours.
- Augmented Dickey-Fuller unit root testEconometrics↔ compare
- Engle-Granger Cointegration TestEconometrics↔ compare
- Phillips-Perron unit root testEconometrics↔ compare
- Structural Break ADF Unit Root TestEconometrics↔ compare
- Toda-Yamamoto causality testEconometrics↔ compare
- Zivot-Andrews Structural Break TestEconometrics↔ compare