Multivariate ARDL Unit Root Test.
mvardlurt: Multivariate ARDL Unit Root Test
R implementation of the multivariate autoregressive distributed lag (ARDL) unit root test proposed by Sam, McNown, Goh, and Goh (2024).
Overview
The mvardlurt package implements a unit root test that augments the standard ADF regression with lagged levels of a covariate (independent variable) to improve power, especially when cointegration exists. Bootstrap critical values ensure correct size regardless of nuisance parameters.
Installation
Install from CRAN:
install.packages("mvardlurt")
Or install the development version from GitHub:
# install.packages("devtools")
Usage
library(mvardlurt)
# Generate example data with cointegration
set.seed(123)
n <- 200
x <- cumsum(rnorm(n)) # I(1) process
y <- 0.5 * x + rnorm(n, sd = 0.5) # Cointegrated with x
# Run the test
result <- mvardlurt(y, x, case = 3, reps = 1000)
print(result)
summary(result)
# Diagnostic plots
plot(result)
The Four-Case Framework
Based on the test results, the package determines one of four cases:
| Case | t-test | F-test | Interpretation |
|---|---|---|---|
| I | Reject | Reject | Cointegration |
| II | Reject | Accept | Degenerate case 1 (y may be I(0)) |
| III | Accept | Reject | Degenerate case 2 (spurious) |
| IV | Accept | Accept | No cointegration |
Deterministic Cases
- Case 1: No deterministic terms
- Case 3: Intercept only (default)
- Case 5: Intercept and linear trend
References
Sam, C. Y., McNown, R., Goh, S. K., & Goh, K. L. (2024). A multivariate autoregressive distributed lag unit root test. Studies in Economics and Econometrics, 1-17. doi:10.1080/03796205.2024.2439101
License
GPL (>= 3)