Chapter 12

Regression with Time-Series Data:

Nonstationary Variables

Walter R. Paczkowski
Rutgers University
Chapter 12: Regression with Time-Series Data:
Principles of Econometrics, 4th Edition Page 1
Nonstationary Variables
 Chapter Contents

 12.1 Stationary and Nonstationary Variables

 12.2 Spurious Regressions
 12.3 Unit Root Tests for Nonstationarity
 12.4 Cointegration
 12.5 Regression When There Is No Cointegration

Chapter 12: Regression with Time-Series Data:

Principles of Econometrics, 4th Edition Page 2
Nonstationary Variables
 The aim is to describe how to estimate regression
models involving nonstationary variables
– The first step is to examine the time-series
concepts of stationarity (and nonstationarity)
and how we distinguish between them.
– Cointegration is another important related
concept that has a bearing on our choice of a
regression model

Chapter 12: Regression with Time-Series Data:

Principles of Econometrics, 4th Edition Page 3
Nonstationary Variables
 12.1
Stationary and Nonstationary
Variables

Chapter 12: Regression with Time-Series Data:

Principles of Econometrics, 4th Edition Page 4
Nonstationary Variables
 12.1
Stationary and
Nonstationary
Variables

The change in a variable is an important concept

– The change in a variable yt, also known as its
first difference, is given by Δyt = yt – yt-1.
• Δyt is the change in the value of the variable
y from period t - 1 to period t

Chapter 12: Regression with Time-Series Data:

Principles of Econometrics, 4th Edition Page 5
Nonstationary Variables
 12.1
Stationary and
Nonstationary FIGURE 12.1 U.S. economic time series
Variables

Chapter 12: Regression with Time-Series Data:

Principles of Econometrics, 4th Edition Page 6
Nonstationary Variables
 12.1
Stationary and
Nonstationary FIGURE 12.1 (Continued) U.S. economic time series
Variables

Chapter 12: Regression with Time-Series Data:

Principles of Econometrics, 4th Edition Page 7
Nonstationary Variables
 12.1
Stationary and
Nonstationary
Variables

Formally, a time series yt is stationary if its mean

and variance are constant over time, and if the
covariance between two values from the series
depends only on the length of time separating the
two values, and not on the actual times at which
the variables are observed

Chapter 12: Regression with Time-Series Data:

Principles of Econometrics, 4th Edition Page 8
Nonstationary Variables
 12.1
Stationary and
Nonstationary
Variables

That is, the time series yt is stationary if for all

values, and every time period, it is true that:
Eq. 12.1a E  yt   μ (constant mean)
Eq. 12.1b var  yt   σ 2 (constant variance)
Eq. 12.1c cov  yt , yt  s   cov  yt , yt  s   γ s (covariance depends on s, not t )

Chapter 12: Regression with Time-Series Data:

Principles of Econometrics, 4th Edition Page 9
Nonstationary Variables
 12.1
Stationary and
Nonstationary Table 12.1 Sample Means of Time Series Shown in Figure 12.1
Variables

Chapter 12: Regression with Time-Series Data:

Principles of Econometrics, 4th Edition Page 10
Nonstationary Variables
 12.1
Stationary and
Nonstationary
Variables

Nonstationary series with nonconstant means are

often described as not having the property of
mean reversion
– Stationary series have the property of mean
reversion

Chapter 12: Regression with Time-Series Data:

Principles of Econometrics, 4th Edition Page 11
Nonstationary Variables
 12.1
Stationary and
Nonstationary
Variables

12.1.1
The First-Order
Autoregressive
Model

The econometric model generating a time-series

variable yt is called a stochastic or random
process
– A sample of observed yt values is called a
particular realization of the stochastic process
• It is one of many possible paths that the
stochastic process could have taken

Chapter 12: Regression with Time-Series Data:

Principles of Econometrics, 4th Edition Page 12
Nonstationary Variables
 12.1
Stationary and
Nonstationary
Variables

12.1.1
The First-Order
Autoregressive
The autoregressive model of order one, the AR(1)
Model
model, is a useful univariate time series model for
explaining the difference between stationary and
nonstationary series:
Eq. 12.2a yt  yt 1  vt ,  1
– The errors vt are independent, with zero mean
2
σ
and constant variance v , and may be normally
distributed
– The errors are sometimes known as ‘‘shocks’’
or ‘‘innovations’’

Chapter 12: Regression with Time-Series Data:

Principles of Econometrics, 4th Edition Page 13
Nonstationary Variables
 12.1
Stationary and
Nonstationary FIGURE 12.2 Time-series models
Variables

12.1.1
The First-Order
Autoregressive
Model

Chapter 12: Regression with Time-Series Data:

Principles of Econometrics, 4th Edition Page 14
Nonstationary Variables
 12.1
Stationary and
Nonstationary FIGURE 12.2 (Continued) Time-series models
Variables

12.1.1
The First-Order
Autoregressive
Model

Chapter 12: Regression with Time-Series Data:

Principles of Econometrics, 4th Edition Page 15
Nonstationary Variables
 12.1
Stationary and
Nonstationary
Variables

12.1.1
The First-Order
Autoregressive
The value ‘‘zero’’ is the constant mean of the
Model
series, and it can be determined by doing some
algebra known as recursive substitution
– Consider the value of y at time t = 1, then its
value at time t = 2 and so on
– These values are:
y1  y0  v1

y2  y1  v2  (y0  v1 )  v2   2 y0  v1  v2

yt  vt  vt 1   2 vt  2  .....  t y0
Chapter 12: Regression with Time-Series Data:
Principles of Econometrics, 4th Edition Page 16
Nonstationary Variables
 12.1
Stationary and
Nonstationary
Variables

12.1.1
The First-Order
Autoregressive
Model
The mean of yt is:
E  yt   E  vt  vt 1  2 vt  2     0

Real-world data rarely have a zero mean

– We can introduce a nonzero mean μ as:
( yt  )  ( yt 1  )  vt
– Or
Eq. 12.2b yt     yt 1  vt ,  1

Chapter 12: Regression with Time-Series Data:

Principles of Econometrics, 4th Edition Page 17
Nonstationary Variables
 12.1
Stationary and
Nonstationary
Variables

12.1.1
The First-Order
Autoregressive
Model

With α = 1 and ρ = 0.7:

E ( yt )     / (1  )  1 / (1  0.7)  3.33

Chapter 12: Regression with Time-Series Data:

Principles of Econometrics, 4th Edition Page 18
Nonstationary Variables
 12.1
Stationary and
Nonstationary
Variables

12.1.1
The First-Order
Autoregressive
Model

An extension to Eq. 12.2a is to consider an AR(1)

model fluctuating around a linear trend: (μ + δt)
– Let the ‘‘de-trended’’ series (yt -μ - δt) behave
like an autoregressive model:
( yt    t )  ( yt 1    (t  1))  vt ,  1
Or:
Eq. 12.2c yt    yt 1  t  vt

Chapter 12: Regression with Time-Series Data:

Principles of Econometrics, 4th Edition Page 19
Nonstationary Variables
 12.1
Stationary and
Nonstationary
Variables

12.1.2
Random Walk
Models
Consider the special case of ρ = 1:
Eq. 12.3a yt  yt 1  vt

– This model is known as the random walk model

• These time series are called random walks
because they appear to wander slowly upward
or downward with no real pattern
• the values of sample means calculated from
subsamples of observations will be dependent
on the sample period
– This is a characteristic of nonstationary
series
Chapter 12: Regression with Time-Series Data:
Principles of Econometrics, 4th Edition Page 20
Nonstationary Variables
 12.1
Stationary and
Nonstationary
Variables

12.1.2
Random Walk
Models

We can understand the ‘‘wandering’’ by recursive

substitution:
y1  y0  v1
2
y2  y1  v2  ( y0  v1 )  v2  y0   vs
s 1


t
yt  yt 1  vt  y0   vs
s 1

Chapter 12: Regression with Time-Series Data:

Principles of Econometrics, 4th Edition Page 21
Nonstationary Variables
 12.1
Stationary and
Nonstationary
Variables

12.1.2
Random Walk
Models

The term  s 1 vs is often called the stochastic

t

trend
– This term arises because a stochastic
component vt is added for each time t, and
because it causes the time series to trend in
unpredictable directions

Chapter 12: Regression with Time-Series Data:

Principles of Econometrics, 4th Edition Page 22
Nonstationary Variables
 12.1
Stationary and
Nonstationary
Variables

12.1.2
Random Walk
Models

Recognizing that the vt are independent, taking the

expectation and the variance of yt yields, for a
fixed initial value y0:
E ( yt )  y0  E (v1  v2  ...  vt )  y0

var( yt )  var(v1  v2  ...  vt )  tσv2

– The random walk has a mean equal to its initial
value and a variance that increases over time,
eventually becoming infinite

Chapter 12: Regression with Time-Series Data:

Principles of Econometrics, 4th Edition Page 23
Nonstationary Variables
 12.1
Stationary and
Nonstationary
Variables

12.1.2
Random Walk
Models

Another nonstationary model is obtained by

adding a constant term:
Eq. 12.3b yt    yt 1  vt
– This model is known as the random walk with
drift

Chapter 12: Regression with Time-Series Data:

Principles of Econometrics, 4th Edition Page 24
Nonstationary Variables
 12.1
Stationary and
Nonstationary
Variables

12.1.2
Random Walk
Models

A better understanding is obtained by applying

recursive substitution:
y1    y0  v1
2
y2    y1  v2    (  y0  v1 )  v2  2  y0   vs
s 1


t
yt    yt 1  vt  t   y0   vs
s 1

Chapter 12: Regression with Time-Series Data:

Principles of Econometrics, 4th Edition Page 25
Nonstationary Variables
 12.1
Stationary and
Nonstationary
Variables

12.1.2
Random Walk
Models

The term tα a deterministic trend component

– It is called a deterministic trend because a fixed
value α is added for each time t
– The variable y wanders up and down as well as
increases by a fixed amount at each time t

Chapter 12: Regression with Time-Series Data:

Principles of Econometrics, 4th Edition Page 26
Nonstationary Variables
 12.1
Stationary and
Nonstationary
Variables

12.1.2
Random Walk
Models

The mean and variance of yt are:

E ( yt )  t   y0  E (v1  v2  v3  ...  vt )  t   y0
var( yt )  var(v1  v2  v3  ...  vt )  t v2

Chapter 12: Regression with Time-Series Data:

Principles of Econometrics, 4th Edition Page 27
Nonstationary Variables
 12.1
Stationary and
Nonstationary
Variables

12.1.2
Random Walk
Models

We can extend the random walk model even

further by adding a time trend:
Eq. 12.3c yt    t  yt 1  vt

Chapter 12: Regression with Time-Series Data:

Principles of Econometrics, 4th Edition Page 28
Nonstationary Variables
 12.1
Stationary and
Nonstationary
Variables

12.1.2
Random Walk
Models
The addition of a time-trend variable t strengthens
the trend behavior:
y1      y0  v1
2
y2    2  y1  v2    2  (    y0  v1 )  v2  2  3  y0   vs
s 1


 t (t  1)  t
yt    t  yt 1  vt  t       y 0   vs
 2  s 1

where we used:

1  2  3    t  t  t  1 2

Chapter 12: Regression with Time-Series Data:

Principles of Econometrics, 4th Edition Page 29
Nonstationary Variables
 12.2
Spurious Regressions

Chapter 12: Regression with Time-Series Data:

Principles of Econometrics, 4th Edition Page 30
Nonstationary Variables
 12.2
Spurious
Regressions

The main reason why it is important to know

whether a time series is stationary or nonstationary
before one embarks on a regression analysis is that
there is a danger of obtaining apparently
significant regression results from unrelated data
when nonstationary series are used in regression
analysis
– Such regressions are said to be spurious

Chapter 12: Regression with Time-Series Data:

Principles of Econometrics, 4th Edition Page 31
Nonstationary Variables
 12.2
Spurious
Regressions

Consider two independent random walks:

rw1 : yt  yt 1  v1t
rw2 : xt  xt 1  v2t
– These series were generated independently and,
in truth, have no relation to one another
– Yet when plotted we see a positive relationship
between them

Chapter 12: Regression with Time-Series Data:

Principles of Econometrics, 4th Edition Page 32
Nonstationary Variables
 12.2
Spurious
Regressions
FIGURE 12.3 Time series and scatter plot of two random walk variables

Chapter 12: Regression with Time-Series Data:

Principles of Econometrics, 4th Edition Page 33
Nonstationary Variables
 12.2
Spurious FIGURE 12.3 (Continued) Time series and scatter plot of two random
Regressions walk variables

Chapter 12: Regression with Time-Series Data:

Principles of Econometrics, 4th Edition Page 34
Nonstationary Variables
 12.2
Spurious
Regressions

A simple regression of series one (rw1) on series

two (rw2) yields:
  17.818  0.842 rw ,
rw R 2  0.70
1t 2t

(t ) (40.837)

– These results are completely meaningless, or

spurious
• The apparent significance of the relationship
is false

Chapter 12: Regression with Time-Series Data:

Principles of Econometrics, 4th Edition Page 35
Nonstationary Variables
 12.2
Spurious
Regressions

When nonstationary time series are used in a

regression model, the results may spuriously
indicate a significant relationship when there is
none
– In these cases the least squares estimator and
least squares predictor do not have their usual
properties, and t-statistics are not reliable
– Since many macroeconomic time series are
nonstationary, it is particularly important to
take care when estimating regressions with
macroeconomic variables
Chapter 12: Regression with Time-Series Data:
Principles of Econometrics, 4th Edition Page 36
Nonstationary Variables
 12.3
Unit Root Tests for Stationarity

Chapter 12: Regression with Time-Series Data:

Principles of Econometrics, 4th Edition Page 37
Nonstationary Variables
 12.3
Unit Root Tests for
Stationarity

There are many tests for determining whether a

series is stationary or nonstationary
– The most popular is the Dickey–Fuller test

Chapter 12: Regression with Time-Series Data:

Principles of Econometrics, 4th Edition Page 38
Nonstationary Variables
 12.3
Unit Root Tests for
Stationarity

12.3.1
Dickey-Fuller Test 1
(No constant and No
Trend)
The AR(1) process yt = ρyt-1 + vt is stationary when
|ρ| < 1
– But, when ρ = 1, it becomes the nonstationary
random walk process
– We want to test whether ρ is equal to one or
significantly less than one
• Tests for this purpose are known as unit root
tests for stationarity

Chapter 12: Regression with Time-Series Data:

Principles of Econometrics, 4th Edition Page 39
Nonstationary Variables
 12.3
Unit Root Tests for
Stationarity

12.3.1
Dickey-Fuller Test 1
(No constant and No
Trend)

Consider again the AR(1) model:

Eq. 12.4 yt  yt 1  vt
– We can test for nonstationarity by testing the
null hypothesis that ρ = 1 against the alternative
that |ρ| < 1
• Or simply ρ < 1

Chapter 12: Regression with Time-Series Data:

Principles of Econometrics, 4th Edition Page 40
Nonstationary Variables
 12.3
Unit Root Tests for
Stationarity

12.3.1
Dickey-Fuller Test 1
(No constant and No
Trend)
A more convenient form is:
yt  yt 1  yt 1  yt 1  vt
Eq. 12.5a yt     1 yt 1  vt
  yt 1  vt
– The hypotheses are:
H0 :   1  H0 :   0

H1 :   1  H1 :   0

Chapter 12: Regression with Time-Series Data:

Principles of Econometrics, 4th Edition Page 41
Nonstationary Variables
 12.3
Unit Root Tests for
Stationarity

12.3.2
Dickey-Fuller Test 2
(With Constant but
No Trend)

The second Dickey–Fuller test includes a constant

term in the test equation:
Eq. 12.5b yt     yt 1  vt
– The null and alternative hypotheses are the
same as before

Chapter 12: Regression with Time-Series Data:

Principles of Econometrics, 4th Edition Page 42
Nonstationary Variables
 12.3
Unit Root Tests for
Stationarity

12.3.3
Dickey-Fuller Test 3
(With Constant and
With Trend)

The third Dickey–Fuller test includes a constant

and a trend in the test equation:
Eq. 12.5c yt    yt 1  t  vt
– The null and alternative hypotheses are
H0: γ = 0 and H1:γ < 0 (same as before)

Chapter 12: Regression with Time-Series Data:

Principles of Econometrics, 4th Edition Page 43
Nonstationary Variables
 12.3
Unit Root Tests for
Stationarity

12.3.4
The Dickey-Fuller
Critical Values

To test the hypothesis in all three cases, we simply

estimate the test equation by least squares and
examine the t-statistic for the hypothesis that
γ=0
– Unfortunately this t-statistic no longer has the
t-distribution
– Instead, we use the statistic often called a τ
(tau) statistic

Chapter 12: Regression with Time-Series Data:

Principles of Econometrics, 4th Edition Page 44
Nonstationary Variables
 12.3
Unit Root Tests for
Stationarity
Table 12.2 Critical Values for the Dickey–Fuller Test

12.3.4
The Dickey-Fuller
Critical Values

Chapter 12: Regression with Time-Series Data:

Principles of Econometrics, 4th Edition Page 45
Nonstationary Variables
 12.3
Unit Root Tests for
Stationarity

12.3.4
The Dickey-Fuller
Critical Values

To carry out a one-tail test of significance, if τc is

the critical value obtained from Table 12.2, we
reject the null hypothesis of nonstationarity if
τ ≤ τc
– If τ > τc then we do not reject the null
hypothesis that the series is nonstationary

Chapter 12: Regression with Time-Series Data:

Principles of Econometrics, 4th Edition Page 46
Nonstationary Variables
 12.3
Unit Root Tests for
Stationarity

12.3.4
The Dickey-Fuller
Critical Values

An important extension of the Dickey–Fuller test

allows for the possibility that the error term is
autocorrelated
– Consider the model:
m
Eq. 12.6 yt     yt 1   as yt  s  vt
s 1
where
yt 1   yt 1  yt  2  , yt  2   yt  2  yt 3  , 

Chapter 12: Regression with Time-Series Data:

Principles of Econometrics, 4th Edition Page 47
Nonstationary Variables
 12.3
Unit Root Tests for
Stationarity

12.3.4
The Dickey-Fuller
Critical Values

The unit root tests based on Eq. 12.6 and its

variants (intercept excluded or trend included) are
referred to as augmented Dickey–Fuller tests
– When γ = 0, in addition to saying that the series
is nonstationary, we also say the series has a
unit root
– In practice, we always use the augmented
Dickey–Fuller test

Chapter 12: Regression with Time-Series Data:

Principles of Econometrics, 4th Edition Page 48
Nonstationary Variables
 12.3
Unit Root Tests for
Stationarity Table 12.3 AR processes and the Dickey-Fuller Tests

12.3.5
The Dickey-Fuller
Testing Procedures

Chapter 12: Regression with Time-Series Data:

Principles of Econometrics, 4th Edition Page 49
Nonstationary Variables
 12.3
Unit Root Tests for
Stationarity

12.3.5
The Dickey-Fuller The Dickey-Fuller testing procedure:
Testing Procedures
– First plot the time series of the variable and select a
suitable Dickey-Fuller test based on a visual
inspection of the plot
• If the series appears to be wandering or fluctuating
around a sample average of zero, use test equation
(12.5a)
• If the series appears to be wandering or fluctuating
around a sample average which is nonzero, use test
equation (12.5b)
• If the series appears to be wandering or fluctuating
around a linear trend, use test equation (12.5c)

Chapter 12: Regression with Time-Series Data:

Principles of Econometrics, 4th Edition Page 50
Nonstationary Variables
 12.3
Unit Root Tests for
Stationarity

12.3.5
The Dickey-Fuller
Testing Procedures

The Dickey-Fuller testing procedure (Continued):

– Second, proceed with one of the unit root tests
described in Sections 12.3.1 to 12.3.3

Chapter 12: Regression with Time-Series Data:

Principles of Econometrics, 4th Edition Page 51
Nonstationary Variables
 12.3
Unit Root Tests for
Stationarity

12.3.6
The Dickey-Fuller
Tests: An Example

As an example, consider the two interest rate

series:
– The federal funds rate (Ft)
– The three-year bond rate (Bt)
Following procedures described in Sections 12.3
and 12.4, we find that the inclusion of one lagged
difference term is sufficient to eliminate
autocorrelation in the residuals in both cases

Chapter 12: Regression with Time-Series Data:

Principles of Econometrics, 4th Edition Page 52
Nonstationary Variables
 12.3
Unit Root Tests for
Stationarity

12.3.6
The Dickey-Fuller
Tests: An Example
The results from estimating the resulting equations
are:  F  0.173  0.045 F  0.561F
 t t 1 t 1

(tau ) (  2.505)

 B  0.237  0.056 B  0.237B

 t t 1 t 1

(tau ) (  2.703)

– The 5% critical value for tau (τc) is -2.86

– Since -2.505 > -2.86, we do not reject the null
hypothesis of non-stationarity.
Chapter 12: Regression with Time-Series Data:
Principles of Econometrics, 4th Edition Page 53
Nonstationary Variables
 12.3
Unit Root Tests for
Stationarity

12.3.7
Order of Integration Recall that if yt follows a random walk, then γ = 0
and the first difference of yt becomes:
yt  yt  yt 1  vt
– Series like yt, which can be made stationary by
taking the first difference, are said to be
integrated of order one, and denoted as I(1)
• Stationary series are said to be integrated of
order zero, I(0)
– In general, the order of integration of a series is
the minimum number of times it must be
differenced to make it stationary
Chapter 12: Regression with Time-Series Data:
Principles of Econometrics, 4th Edition Page 54
Nonstationary Variables
 12.3
Unit Root Tests for
Stationarity

12.3.7
Order of Integration

The results of the Dickey–Fuller test for a random

walk applied to the first differences are:
  F    0.447  F 
 t t 1

(tau ) (  5.487)
  B    0.701 B 
 t t 1

(tau ) (  7.662)

Chapter 12: Regression with Time-Series Data:

Principles of Econometrics, 4th Edition Page 55
Nonstationary Variables
 12.3
Unit Root Tests for
Stationarity

12.3.7
Order of Integration

Based on the large negative value of the tau

statistic (-5.487 < -1.94), we reject the null
hypothesis that ΔFt is nonstationary and accept the
alternative that it is stationary
– We similarly conclude that ΔBt is stationary
(-7.662 < -1.94)

Chapter 12: Regression with Time-Series Data:

Principles of Econometrics, 4th Edition Page 56
Nonstationary Variables
 12.4
Cointegration

Chapter 12: Regression with Time-Series Data:

Principles of Econometrics, 4th Edition Page 57
Nonstationary Variables
 12.4
Cointegration

As a general rule, nonstationary time-series

variables should not be used in regression models
to avoid the problem of spurious regression
– There is an exception to this rule

Chapter 12: Regression with Time-Series Data:

Principles of Econometrics, 4th Edition Page 58
Nonstationary Variables
 12.4
Cointegration

There is an important case when

et = yt - β1 - β2xt is a stationary I(0) process
– In this case yt and xt are said to be cointegrated
• Cointegration implies that yt and xt share
similar stochastic trends, and, since the
difference et is stationary, they never diverge
too far from each other

Chapter 12: Regression with Time-Series Data:

Principles of Econometrics, 4th Edition Page 59
Nonstationary Variables
 12.4
Cointegration

The test for stationarity of the residuals is based

on the test equation:
Eq. 12.7 eˆt  γeˆt 1  vt
– The regression has no constant term because
the mean of the regression residuals is zero.
– We are basing this test upon estimated values of
the residuals

Chapter 12: Regression with Time-Series Data:

Principles of Econometrics, 4th Edition Page 60
Nonstationary Variables
 12.4
Cointegration Table 12.4 Critical Values for the Cointegration Test

Chapter 12: Regression with Time-Series Data:

Principles of Econometrics, 4th Edition Page 61
Nonstationary Variables
 12.4
Cointegration

There are three sets of critical values

– Which set we use depends on whether the
residuals are derived from:
Eq. 12.8a Equation 1: eˆt  yt  bxt
Eq. 12.8b Equation 2 : eˆt  yt  b2 xt  b1
Eq. 12.8c Equation 3 : eˆt  yt  b2 xt  b1  ˆ t

Chapter 12: Regression with Time-Series Data:

Principles of Econometrics, 4th Edition Page 62
Nonstationary Variables
 12.4
Cointegration

12.4.1
An Example of a
Cointegration Test

Consider the estimated model:

Bˆt  1.140  0.914 Ft , R 2  0.881
Eq. 12.9
(t ) (6.548) (29.421)
– The unit root test for stationarity in the
estimated residuals is:
eˆt  0.225eˆt 1  0.254eˆt 1
(tau ) ( 4.196)

Chapter 12: Regression with Time-Series Data:

Principles of Econometrics, 4th Edition Page 63
Nonstationary Variables
 12.4
Cointegration

12.4.1
An Example of a
Cointegration Test

The null and alternative hypotheses in the test for

cointegration are:
H 0 : the series are not cointegrated  residuals are nonstationary

H1 : the series are cointegrated  residuals are stationary

– Similar to the one-tail unit root tests, we reject

the null hypothesis of no cointegration if τ ≤ τc,
and we do not reject the null hypothesis that the
series are not cointegrated if τ > τc

Chapter 12: Regression with Time-Series Data:

Principles of Econometrics, 4th Edition Page 64
Nonstationary Variables
 12.4
Cointegration

12.4.2
The Error Correction
Model

Consider a general model that contains lags of y

and x
– Namely, the autoregressive distributed lag
(ARDL) model, except the variables are
nonstationary:
yt  δ  θ1 yt 1  δ 0 xt  δ1 xt 1  vt

Chapter 12: Regression with Time-Series Data:

Principles of Econometrics, 4th Edition Page 65
Nonstationary Variables
 12.4
Cointegration

12.4.2
The Error Correction
Model

If y and x are cointegrated, it means that there is a

long-run relationship between them
– To derive this exact relationship, we set
yt = yt-1 = y, xt = xt-1 = x and vt = 0
– Imposing this concept in the ARDL, we obtain:
y  1  θ1   δ   δ0  δ1  x
• This can be rewritten in the form:
y  β1  β 2 x

Chapter 12: Regression with Time-Series Data:

Principles of Econometrics, 4th Edition Page 66
Nonstationary Variables
 12.4
Cointegration

12.4.2
The Error Correction
Model
Add the term -yt-1 to both sides of the equation:

yt  yt 1  δ   θ1  1 yt 1  δ0 xt  δ1 xt 1  vt

– Add the term – δ0xt-1+ δ0xt-1:

yt  δ   θ1  1 yt 1  δ0  xt  xt 1    δ0  δ1  xt 1  vt

– Manipulating this we get:

 δ
yt   θ1  1   yt 1 
 δ0  δ1  
xt 1   δ0 xt  vt
  θ  1  θ  1 
 1 1 

Chapter 12: Regression with Time-Series Data:

Principles of Econometrics, 4th Edition Page 67
Nonstationary Variables
 12.4
Cointegration

12.4.2
The Error Correction
Model
Or:
Eq. 12.10 yt  α  yt 1  β1  β 2 xt 1   δ0 xt  vt

– This is called an error correction equation

– This is a very popular model because:
• It allows for an underlying or fundamental link
between variables (the long-run relationship)
• It allows for short-run adjustments (i.e.
changes) between variables, including
adjustments to achieve the cointegrating
relationship
Chapter 12: Regression with Time-Series Data:
Principles of Econometrics, 4th Edition Page 68
Nonstationary Variables
 12.4
Cointegration

12.4.2
The Error Correction
Model

For the bond and federal funds rates example, we

have:
Bˆt  0.142  Bt 1  1.429  0.777 Ft 1   0.842Ft  0.327Ft 1
 t  2.857   9.387   3.855 
– The estimated residuals are

eˆt 1   Bt 1  1.429  0.777 Ft 1 

Chapter 12: Regression with Time-Series Data:

Principles of Econometrics, 4th Edition Page 69
Nonstationary Variables
 12.4
Cointegration

The result from applying the ADF test for

stationarity is:

eˆt  0.169eˆt 1  0.180eˆt 1

 t   3.929 
– Comparing the calculated value (-3.929) with
the critical value, we reject the null hypothesis
and conclude that (B, F) are cointegrated

Chapter 12: Regression with Time-Series Data:

Principles of Econometrics, 4th Edition Page 70
Nonstationary Variables
 12.5
Regression with No-Cointegration

Chapter 12: Regression with Time-Series Data:

Principles of Econometrics, 4th Edition Page 71
Nonstationary Variables
 12.5
Regression When
There is No
Cointegration

How we convert nonstationary series to stationary

series, and the kind of model we estimate, depend
on whether the variables are difference stationary
or trend stationary
– In the former case, we convert the
nonstationary series to its stationary counterpart
by taking first differences
– In the latter case, we convert the nonstationary
series to its stationary counterpart by de-
trending

Chapter 12: Regression with Time-Series Data:

Principles of Econometrics, 4th Edition Page 72
Nonstationary Variables
 12.5
Regression When
There is No
Cointegration

12.5.1
First Difference
Stationary
Consider the random walk model:
yt  yt 1  vt

– This can be rendered stationary by taking the

first difference:
yt  yt  yt 1  vt

• The variable yt is said to be a first difference

stationary series

Chapter 12: Regression with Time-Series Data:

Principles of Econometrics, 4th Edition Page 73
Nonstationary Variables
 12.5
Regression When
There is No
Cointegration

12.5.1
First Difference
Stationary
A suitable regression involving only stationary
variables is:
Eq. 12.11a yt  yt 1  0 xt  1xt 1  et
– Now consider a series yt that behaves like a
random walk with drift:
yt    yt 1  vt
with first difference:
yt    vt
• The variable yt is also said to be a first
difference stationary series, even though it
is stationary around a constant term
Chapter 12: Regression with Time-Series Data:
Principles of Econometrics, 4th Edition Page 74
Nonstationary Variables
 12.5
Regression When
There is No
Cointegration

12.5.1
First Difference
Stationary

Suppose that y and x are I(1) and not cointegrated

– An example of a suitable regression equation is:
Eq. 12.11b yt    yt 1  0 xt  1xt 1  et

Chapter 12: Regression with Time-Series Data:

Principles of Econometrics, 4th Edition Page 75
Nonstationary Variables
 12.5
Regression When
There is No
Cointegration

12.5.2
Trend Stationary

Consider a model with a constant term, a trend

term, and a stationary error term:
yt    t  vt
– The variable yt is said to be trend stationary
because it can be made stationary by removing
the effect of the deterministic (constant and
trend) components:
yt    t  vt

Chapter 12: Regression with Time-Series Data:

Principles of Econometrics, 4th Edition Page 76
Nonstationary Variables
 12.5
Regression When
There is No
Cointegration

12.5.2
Trend Stationary

If y and x are two trend-stationary variables, a

possible autoregressive distributed lag model is:
Eq. 12.12 yt   yt1  0 xt  1 xt1  et

Chapter 12: Regression with Time-Series Data:

Principles of Econometrics, 4th Edition Page 77
Nonstationary Variables
 12.5
Regression When
There is No
Cointegration

12.5.2
Trend Stationary

As an alternative to using the de-trended data for

estimation, a constant term and a trend term can be
included directly in the equation:
yt     t   yt 1  0 xt  1 xt 1  et
where:
  1 (1  1 )   2 (0  1 )  11  1 2

  1 (1  1 )   2 (0  1 )

Chapter 12: Regression with Time-Series Data:

Principles of Econometrics, 4th Edition Page 78
Nonstationary Variables
 12.5
Regression When
There is No
Cointegration

12.5.3
Summary If variables are stationary, or I(1) and cointegrated,
we can estimate a regression relationship between
the levels of those variables without fear of
encountering a spurious regression
If the variables are I(1) and not cointegrated, we
need to estimate a relationship in first differences,
with or without the constant term
If they are trend stationary, we can either de-trend
the series first and then perform regression
analysis with the stationary (de-trended) variables
or, alternatively, estimate a regression relationship
that includes a trend variable
Chapter 12: Regression with Time-Series Data:
Principles of Econometrics, 4th Edition Page 79
Nonstationary Variables
 12.5
Regression When
There is No FIGURE 12.4 Regression with time-series data: nonstationary variables
Cointegration

12.5.3
Summary

Chapter 12: Regression with Time-Series Data:

Principles of Econometrics, 4th Edition Page 80
Nonstationary Variables
 Key Words

Chapter 12: Regression with Time-Series Data:

Principles of Econometrics, 4th Edition Page 81
Nonstationary Variables
 Keywords

autoregressive order of stochastic

process integration process
cointegration random walk stochastic trend
Dickey–Fuller process tau statistic
tests random walk trend stationary
difference with drift unit root tests
stationary spurious
mean reversion regressions
nonstationary stationary

Chapter 12: Regression with Time-Series Data:

Principles of Econometrics, 4th Edition Page 82
Nonstationary Variables