Cointegration: Regression of A Unit Root Time Series On Another Unit Root Time Series

Cointegration: regression of a unit root time series on another unit root time series
As the regression of a nonstationary time series on another nonstationary time series may produce a spurious regression. Soppose Yt and Xt are non stationary time series say I(1) Yt = 1 + 2 Xt + ut Or, ut = Yt - 1 - 2 Xt (1)
Here, after subjected to unit root analysis, ut is found to be a stationary series i.e. I(0). This is an interesting situation, for although Yt and Xt are non-stationary, I(1), ie they have stochastic trends, their linear combination Is stationary I(0). So to speak, the linear combination cancels out the stochastic trends in the two series. For eg. If we take consumption and income as two I(1) variables, savings defined as (income consumption) could be I(0). As a result, a regression of consumption on income would be meaningful (i.e., not spurious). In this case we say that the two variables are cointegrated. Economically speaking, two variables will be cointegrated if they have a long-term, or equilibrium, relationship between them. Provided the residuals from regressions are I(0) or stationary, the traditional regression methodology (including the t and F tests) is applicable to data involving (nonstationary) time series. # A regression such as (21.11.1) is known as a cointegrating regression and the slope parameter 2 is known as the cointegrating parameter.
Testing for Cointegration

A no. of methods for testing cointegration have been proposed in the literature. The two comparatively simple methods are: (1) The DF or ADF unit root test on the residuals estimated from the cointegrating regression, and (2) The cointegrating regression DurbinWatson (CRDW) test.
EngleGranger (EG) or Augmented EngleGranger (AEG) Test.

We know how to apply the DF or ADF unit root tests. All we have to do is estimate a regression, obtain the residuals, and use the DF or ADF tests. There is one precaution to exercise, however. Since the estimated ut are based on the estimated cointegrating parameter 2, the DF and ADF critical significance values are not quite appropriate. Engle and Granger have calculated these values, the DF and ADF tests in the present context are known as EngleGranger (EG) and Augmented EngleGranger (AEG) tests.
Illustration : We regressed consumption time series on disposable income time series and obtained the following regression: Ct = 171.4412 + 0.9672 It t = (7.4808) (119.8712) R2 = 0.9940 d = 0.5316
(21.11.3)
Since Ct and It are individually nonstationary, there is the possibility that this regression is spurious. But when we performed a unit root test on the residuals obtained from (21.11.3), we obtained the following results:
The EngleGranger 1 % critical value is 2.5899. Since the computed t value is much more negative than this, the conclusion is that the residuals from the regression of Ct on It are I(0); ie they are stationary. Hence, (21.11.3) is a cointegrating regression and this regression is not spurious, even though individually the two variables are nonstationary. Eq (21.11.3) shows the static or long run consumption function and interpret its parameters as long run parameters. Thus 0.9672 represents the long-run, or equilibrium, marginal propensity to consumer (MPC).
Cointegrating Regression DurbinWatson (CRDW) Test.

In this test, we use the DurbinWatson d obtained from the cointegrating regression, such as d = 0.5316 given in (21.11.3). But now the null hypothesis is that d = 0 rather than the standard d = 2. This is because d 2(1 ), so if there is to be a unit root, the estimated will be about 1, which implies that d will be about zero. On the basis of 10,000 simulations formed from 100 observations each, the 1, 5, and 10 % critical values to test the hypothesis that the true d = 0 are 0.511, 0.386, and 0.322, respectively. Thus, if computed d value is smaller than, say, 0.511, we reject the null hypothesis of cointegration at the 1 % level. In the example, the value of 0.5316 is above this critical value, suggesting that Ct & It are cointegrated.
Cointegration and Error Correction Mechanism (ECM) :

As the Ct & It cointegrated; ie there is a long term, or equilibrium relationship between the two. Of course, in short run there may be disequilibrium. Therefore, one can treat the error term in (21.11.2) as equilibrium error. And we can use this error term to tie the short-run behavior of Ct to its long-run value. The error correction mechanism (ECM) first used by Sargan and later popularized by Engle and Granger corrects for disequilibrium. An imp theorem, known as Granger representation theorem, states that if two variables Y and X are cointegrated, then the relationship between the two can be expressed as ECM. In Ct & It example, Consider the following model:
Ct = o + 1 It + 2 ut-1 + t
(5)
Where denotes the first difference operator, t is a random error term, and ut1 = (Ct-1 1 - 2 It-1), ie a one-period lagged value of error from the cointegrating regression (1) ie Ct-1 = 1 + 2 It-1 + ut-1 ECM equation (5) states that Ct depends on It and also on equilibrium error term. If t is nonzero, then the model is out of equilibrium. Suppose It is zero and ut1 is positive. This means Ct-1 is too high to be in equilibrium, ie Ct-1 is above its equilibrium value of (0 + 1It1). Since 2 is expected to be negative, the term 2 ut1 is negative and, therefore, Ct will be negative to restore the equilibrium. That is, if Ct is above the equilibrium value, it will start falling in the next period to correct the equilibrium error; hence the name ECM. By the same token, if ut1 is negative (i.e., Ct is below its equilibrium value), 2ut1 will be positive, which will cause Ct to be positive, leading Ct to rise in period t. Thus, the absolute value of 2 decides how quickly the equilibrium is restored. In our example, Ct = 11.6918 + 0.2906 It 0.0867 ut1 t = (5.3249) (4.1717) (1.6003) R2 = 0.1717 d = 1.9233 (21.11.6)
Statistically, the equilibrium error term is zero, suggesting that C adjusts to changes in I in the same time period. As (21.11.6) shows, short-run changes in I have a positive impact on short-run changes in personal consumption. One can interpret 0.2906 as the short-run marginal propensity to consume (MPC); the long-run MPC is given by the estimated (static) equilibrium relation (21.11.3) as 0.9672.

Cointegration: Regression of A Unit Root Time Series On Another Unit Root Time Series

Uploaded by

Copyright:

Available Formats

Cointegration: Regression of A Unit Root Time Series On Another Unit Root Time Series

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Cointegration: Regression of A Unit Root Time Series On Another Unit Root Time Series

Uploaded by

Copyright:

Available Formats

Cointegration: regression of a unit root time series on another unit root time series

Testing for Cointegration

EngleGranger (EG) or Augmented EngleGranger (AEG) Test.

Cointegrating Regression DurbinWatson (CRDW) Test.

Cointegration and Error Correction Mechanism (ECM) :

You might also like