**S+/R – Yet Another Blog in Statistical Computing**, and kindly contributed to R-bloggers)

In a project of developing PPNR balance projection models, I tried to use the Phillips-Ouliaris (PO) test to investigate the cointegration between the historical balance and a set of macro-economic variables and noticed that implementation routines of PO test in various R packages, e.g. urca and tseries, would give different results. After reading through the original paper “Asymptotic Properties of Residual Based Tests for Co-Integration” by P. Phillips again, I started realizing that the po.test() function in the tseries package and the ca.po() function in the urca package are implementing different types of Phillips-Ouliaris cointegration tests. In other words, the so-called “Phillips-Ouliaris Cointegration test” is not A statistical test but a set of statistical tests with different assumptions, formulations, critical values, and implications.

Let’s start with simulating cointegrated series, as below.

set.seed(2019) x <- cumsum(rnorm(200, sd = 0.5)) y <- cumsum(rnorm(200, sd = 0.5)) + 1 z <- x + y + rnorm(200, sd = 0.5)

First of all, the po.test() function from the tseries package is applied to simulated series with following observations:

1. As the position of each series is changed in the po.test() function, we will get different testing results.

2. Results are determined by which series on the most left-hand side.

The reason is that the po.test() function is testing the cointegration with Phillip’s Z_alpha test, which is the second residual-based test described in P171 of the paper. For this test, critical values in tables Ia – Ic in P189 are used to reject the Null of No Cointegration. Because the po.test() will use the series at the first position to derive the residual used in the test, results would be determined by the series on the most left-hand side.

tseries::po.test(cbind(x, y, z), demean = TRUE, lshort = TRUE) # Phillips-Ouliaris demeaned = -186.03, Truncation lag parameter = 1, p-value = 0.01 tseries::po.test(cbind(z, x, y), demean = TRUE, lshort = TRUE) # Phillips-Ouliaris demeaned = -204.7, Truncation lag parameter = 1, p-value = 0.01 tseries::po.test(cbind(z, y, x), demean = TRUE, lshort = TRUE) # Phillips-Ouliaris demeaned = -204.7, Truncation lag parameter = 1, p-value = 0.01

The Phillips-Ouliaris test implemented in the ca.po() function from the urca package is different. In the ca.po() function, there are two cointegration tests implemented, namely “Pu” and “Pz” tests. Although both the ca.po() function and the po.test() function are supposed to do the Phillips-Ouliaris test，outcomes from both functions are completely different.

Below shows results of the Pu test, which is a Variance Ratio test and the fourth residual-based test described in P171 of the paper. For this test, critical values in tables IIIa – IIIc in P191 are used to reject the Null of No Cointegration. Similar to Phillip’s Z_alpha test, the Pu test also is not invariant to the position of each series and therefore would give different outcomes based upon the series on the most left-hand side.

urca::ca.po(cbind(x, y, z), demean = "constant", lag = "short", type = "Pu") # The value of the test statistic is: 72.8124 urca::ca.po(cbind(z, x, y), demean = "constant", lag = "short", type = "Pu") # The value of the test statistic is: 194.5645 urca::ca.po(cbind(z, y, x), demean = "constant", lag = "short", type = "Pu") # The value of the test statistic is: 194.5645

At last, let’s look at the Pz test implemented in the ca.po() function. For this test, critical values in tables IVa – IVc in P192 are used to reject the Null of No Cointegration. As a multivariate trace statistic, the Pz test has its appeal that the outcome won’t change by the position of each series, as shown below.

urca::ca.po(cbind(x, y, z), demean = "constant", lag = "short", type = "Pz") # The value of the test statistic is: 219.2746 urca::ca.po(cbind(z, x, y), demean = "constant", lag = "short", type = "Pz") # The value of the test statistic is: 219.2746

**leave a comment**for the author, please follow the link and comment on their blog:

**S+/R – Yet Another Blog in Statistical Computing**.

R-bloggers.com offers

**daily e-mail updates**about R news and tutorials on topics such as: Data science, Big Data, R jobs, visualization (ggplot2, Boxplots, maps, animation), programming (RStudio, Sweave, LaTeX, SQL, Eclipse, git, hadoop, Web Scraping) statistics (regression, PCA, time series, trading) and more...