Quantitative Finance Stack Exchange is a question and answer site for finance professionals and academics. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

Below is a spread built with two ETFs that pass the cointegration test i.e. Adjusted Dickey Fuller, adfTest(type="nc") in R's fUnitRoots with a p-value < 0.01.

The red line is the trendline.

What test can I use to proove that: (1) both securities are cointegrated and (2) they are mean reversing and the mean is constantly 0 (i.e. stationary, not trended)?


EWA and EZA spread

share|improve this question
Depends on what you mean by "prove". – LazyCat Apr 12 '12 at 20:50
You can include a constant and a deterministic trend in your initial regression. You then test for the presence of the trend (ie the significance of the parameter before the trend term). – Zarbouzou Apr 13 '12 at 10:27
up vote 0 down vote accepted

Here is an empirical strategy to test for cointegration.

FIRST, check whether both $X_t$ and $Y_t$ contain an unit root.

  • If they are both stationary then model $Y_t$ or $X_t$ in levels (and nothing is wrong).
  • If one of the two is $I(1)$ (non-stationary for one level), then take differences to ensure stationarity.
  • If they are both non-stationary, and hence $I(1)$, then test for co-integration:

    1. if the residuals are $I(0)$, then we speak of the presence of cointegration. Estimate then an ECM model ($Y_t = \beta_0 + \beta_1 X_t + \eta_t$ obtaining $\hat{\beta_0}$ and $\hat{\beta_1}$ and using it in: $\Delta Y_t = \Delta X_t'\phi - \psi(Y_{t-1}-\hat{\beta_0} - \hat{\beta_1}X_t) + \varepsilon_t$. When $\varepsilon_t \sim N(0,1)$ then both $\psi$ and $\phi$ are asymptotically valid.
    2. if the residuals are $I(1)$ then we speak of spurious regression. In that case you should model both variables by taking the first differences.
share|improve this answer
This post is identical to this and this. Don't repost. – chrisaycock Apr 13 '12 at 2:33
I disagree with the latest part of you explanation that "2. (...) In that case you should model both variables by taking the first differences." – edouard Apr 14 '12 at 10:31
Yes I understand why you say that, but given the two variables you have I think that is the best you can do. What do you propose? – JohnAndrews Apr 14 '12 at 11:24

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.