##
**What is Stationarity in Time Series Analysis?**

##
In econometrics, time series data are frequently used and they
often pose distinct problems for econometricians. As it will be discussed with
examples, most empirical work based on time series data assumes that the
underlying series is stationary. Stationarity of a series (that is, a variable)
implies that its mean, variance and covariance are constant over time. That is,
these do not vary systematically over time. In order words, they are *time* *invariant*.
However, if that is not the case, then the series is nonstationary. We will
discuss some possible scenarios where two series, *Y* and *X*,
are nonstationary and the error term, *u*, is also nonstationary. In
that case, the error term will exhibit autocorrelation. Another likely scenario
is where *Y* and *X* are nonstationary, but *u* is
stationary. The implications of this will also be explored. In time series
analysis, the words *nonstationary*, *unit root* or *random
walk* *model* are used synonymously. In essence, of a series
is considered to be nonstationary, it implies that such exhibit a unit root and
exemplifies a random walk series.

Regressing two series that
are nonstationary, likewise, yields a spurious (or nonsense) regression. That
is, a regression whose outcome cannot be used for inferences or forecasting. In
short, such results should not be taken seriously and must be discarded. A
stationary series will tend to return to its mean (called

*mean reversion*) and ﬂuctuations around this mean (measured by its variance) will have a broadly constant breadth. But if a time series is not stationary in the sense just explained, it is called a nonstationary time series such will have a time-varying mean or a time-varying variance or both. In summary, a stationary time series is important because if such is nonstationary, its behaviour can be studied only for the time period under consideration. That is, each set of time series data will therefore be for a particular episode. As a result, it is not possible to generalise its relevance to other time periods. Therefore, for the purpose of forecasting, such (nonstationary) time series may be of little practical value**How to detect unit root in a series?**

In a bivariate (2 variables)
model or that involving multiple variables (called a multiple regression
model), it is assumed that all the variables are stationary at level (that is,
the order of integration of each of the variable is zero,

*I*(0). It is important to state at this point, that the order of integration of a series in a regression model is determined by the outcome of a unit root test (or stationarity test). If the series is stationary at level after performing unit root test, then it is*I*(0), otherwise it is*I*(*d*) where*d*represents the number of times the series is differenced before it becomes stationary. But what if the assumption of*stationarity at level*of the series in a bivariate or multiple regression model is relaxed and we consequently allow for a unit root in each of the variables in the model, how can this be corrected? In general, this would require a different treatment from a conventional regression with stationary variables at*I*(0).
In particular, we focus on a
class of linear combination of unit root processes known as cointegrated
process. The generic representation for the order of integration of series
is

*I*(*d*) where*d*is the number of differencing to render the series stationary. Hence, a stationary series at level,*d*= 0 is a series with an*I*(0) process. Although, for any non-stationary series,*‘d’*can assume any value greater than zero, however, in applied research, only the unit root process of*I*(1) process is allowed, otherwise such series with higher order of integration (*d*> 1) should be excluded in the model as no meaningful policy implications or relevance can be drawn from such series.
Here is an example of a
bivariate linear regression model:

*Y*= 𝛂₀ +

_{t }*bX*+

_{t}*u*[1]

_{t}
Assume

*Y*and_{t}*X*are two random walk models that are_{t}*I*(1) processes and are independently distributed as:*Y*= ρ

_{t }*Y*+

_{t-1}*v*, -1 ≤ ρ ≤ 1 [2]

_{t}*X*= ղ

_{t }*X*+

_{t-1}*e*, -1 ≤ ղ ≤ 1 [3]

_{t}
and

*v*and_{t}*e*have zero mean, a constant variance and are orthogonal (these are_{t}*white noise*error terms).
We also assumed that

*v*and_{t}*e*are serially uncorrelated as well as mutually uncorrelated. As stated in [2] and [3], both these time series are nonstationary; that is, they are_{t}*I*(1) or exhibit stochastic trends. Suppose we regress*Y*on_{t}*X*. Since_{t}*Y*on_{t}*X*are uncorrelated_{t}*I*(1) processes, the*R*^{2}from the regression of*Y*on*X*should tend to zero; that is, there should not be any relationship between the two variables. Equations [2] and [3] resemble the Markov ﬁrst-order autoregressive model. If ρ and ղ = 1, the equations become a random walk model without drift. If ρ and ղ are in fact 1, then a unit root problem surfaces, that is, a situation of nonstationarity; because we already know that in this case the variance of*Y*is not stationary. The name unit root is due to the fact that ρ = 1. Again, the terms nonstationary, random walk, and unit root can be treated as synonymous. If, however, |ρ| ≤ 1, and |ղ| ≤ 1, that is if their absolute values are less than one, then it can be shown that both series_{t}*Y*and_{t}*X*are stationary. In practice, then, it is important to ﬁnd out if a time series possesses a unit root._{t}
Given equations [2] and [3],
there should be no systematic relationship between

*Y*and_{t}*X*as they both drift away from equilibrium (i.e. they do not converge), and therefore, we should expect that an ordinary least squares (OLS) estimate of_{t}*b*should be close to zero, or insignificantly different from zero, at least as the sample size increases. But this is not usually the case. The fitted coefficients in this case may be statistically significant even when there is no true relationship between the dependent variable and the regressors. This is regarded as a spurious regression or correlation where, in the case of our example,*b*takes any value randomly, and its*t*-statistic indicates significance of the estimate.**But how can unit root be detected?**There are some clues that tell you if a series is nonstationary and if the regression of bivariate or multivariate relationships are spurious. Some of these are:

1. Do a graphical plot of the series to visualise
the nature. Is it trending upwards or downwards? Does it exhibit a
mean-reversion or not? Or are there fluctuations around its mean?

2. Or carry out a regression analysis on two series
and observe the

*R*^{2}. If it is above 0.9, it may suggest that the variables are nonstationary.
3. The rule-of-thumb: if the

*R*^{2}obtained from the regression is higher than the Durbin Watson (DW) statistic. The low DW statistic evidences positive first order auto-correlation of the error terms.
Using Gujarati and Porter Table 21.1 quarterly data of 1970q1 to 1991q4, examples of
nonstationary series and spurious regression can be seen from the

*pce*,*pdi*and*gdp*relationship. Since the series are measured in billions of US dollars, the natural logarithms of the variables will be used in analysing their essential features.**Nonstationary series:**the graphical plot of the three variables shows an upward trend and none of the variables revert to their means. That is, all three variables do not exhibit mean reversions. That clearly tells us that the series are nonstationary.

Stata: Plot of nonstationary series Source: CrunchEconometrix |

**Note:**Use this syntax to generate the graph:

*line lnpce lnpdi lngdp qtrly, legend(size(medsmall))*

**What is a spurious regression?**Sometimes we expect to find no relationship between two variables, yet a regression of one on the other variable often shows a signiﬁcant relationship. This situation exempliﬁes the problem of spurious, or nonsense, regression. The regression of

*lnpce*on

*lnpdi*shows how spurious regressions can arise if time series are not stationary. As expected, because both variables are nonstationary, the result evidences that a spurious regression has been undertaken.

But how do we know this? Take
a look at the

*R*^{2}the value of**0.9944**is higher than the Durbin Watson statistic of**0.57**. So, whenever the*R*^{2}> DW, a spurious regression has occurred because the variables are nonstationary.*Stata syntax: regress lnpce lnpdi*

Stata: Example of a spurious regression Source: CrunchEconometrix |

As you can see, the coefﬁcient
of

*lnpdi*is highly statistically signiﬁcant, and the*R*^{2}value is statistically signiﬁcantly different from zero. From these results, you may be tempted to conclude that there is a signiﬁcant statistical relationship between both variables, whereas*a priori*there may or may*not*be none. This is simply the phenomenon of**spurious or nonsense regression**, ﬁrst discovered by Yule (1926). He showed that (spurious) correlation could persist in nonstationary time series even if the sample is very large. That there is something wrong in the preceding regression is suggested by the extremely low Durbin–Watson value, which suggests very strong ﬁrst-order autocorrelation. According to Granger and Newbold,*R*^{2}> DW is a good rule of thumb to suspect that the estimated regression is spurious, as in the given example.**Why is it important to test for stationarity?**

We usually consider a
nonstationary series for the following reasons:

1. To evaluate the behaviour of series over time. Is the series
trending upward or downward? This can be verified from performing a
stationarity test. In other words, the test can be used to evaluate the
stability or predictability of time series. If a series is nonstationary, that
means the series is unstable or unpredictable and therefore may not be valid
for inferences, prediction or forecasting.

2. To know how a series responds to shocks requires carrying out a
stationarity test. If such series is nonstationary, the impact of shocks to the
series are more likely to be permanent. Consequently, if a series is
stationary, impact of shocks will be temporary or brief.

**How to correct for nonstationarity?**

What can be done with
nonstationarity in a time series knowing that performing OLS on such a model
yields spurious regression?

**The Unit Root Test**

We begin with equations [2]
and [3] which are unit root (stochastic) processes with white noise error
terms. If the parameters of the models are equal to 1, that is, in the case of
the unit root, both equations become random walk models without drift, which we
know is a nonstationary stochastic process. So, what can be done to correct
this? For instance, for equation [2], simply regress

*Y*on its (one-period) lagged value_{t}*Y*and ﬁnd out if the estimated ρ is statistically equal to 1? If it is, then_{t−1}*Y*is nonstationary. Repeat same for the_{t}*X*series. This is the general idea behind the unit root test of stationarity._{t}
For theoretical reasons,
equation [2] is manipulated as follows: Subtract

*Y*from both sides of [2] to obtain:_{t−1}*Y*-

_{t }*Y*= ρ

_{t-1 }*Y*-

_{t-1}*Y*+

_{t-1 }*v*[4]

_{t}
= (ρ - 1)

*Y*+_{t-1 }*v*_{t}
and this can be stated
alternatively as:

⃤

*Y*= δ_{t}*Y*+_{t-1 }*v*[5]_{t}
where δ = (ρ − 1)
and ⃤, as usual, is the ﬁrst-difference operator. In practice, therefore,
instead of estimating [2], we estimate [5] and test the null hypothesis
that δ = 0. If δ = 0, then ρ = 1, that is we have a unit
root, meaning the time series under consideration is nonstationary.

Before we proceed to
estimate [5], it may be noted that if δ = 0, [5] will become:

⃤

*Y*=_{t}*Y*-_{t-1}*Y*=_{t-1 }*v*[6]_{t}
(Remember to do the same
for

*X*series)_{t}
Since

*v*_{t}is a white noise error term, it is stationary, which means that**the ﬁrst difference of a random walk time series is stationary**.Stata: Example of a stationary series Source: CrunchEconometrix |

Visual observation of the
differenced series shows that the three variables are stationary around the mean.
They all exhibit constant mean-reversions. That is, they fluctuate around 0. If
we are to draw a trend line, such a line will be horizontal at 0.01.

Okay, having said all that.
Let us return to estimating equation [5]. This is quite simple, all that is
required is to take the ﬁrst differences of

*Y*and regress on_{t}*Y*and see if the estimated slope coefﬁcient in this regression is statistically different from is zero or not. If it is zero, we conclude that_{t−1}*Y*is nonstationary. But if it is negative, we conclude that_{t}*Y*is stationary._{t}**Note:**Since δ = (ρ − 1), for stationarity ρ must be less than one. For this to happen δ must be negative!

The only question is which
test do we use to ﬁnd out if the estimated coefﬁcient of

*Y*in [5] is zero or not? You might be tempted to say, why not use the usual_{t−1}*t*test? Unfortunately, under the null hypothesis that δ = 0 (i.e., ρ = 1), the*t*value of the estimated coefﬁcient of*Y*does not follow the_{t−1}*t*distribution even in large samples; that is, it does not have an asymptotic normal distribution.
What is the alternative?
Dickey and Fuller (DF) have shown that under the null hypothesis that δ = 0,
the estimated

*t*value of the coefﬁcient of*Y*in [5] follows the_{t−1}**statistic. These authors have computed the critical values of the***τ (tau)**tau statistic*on the basis of Monte Carlo simulations.**Note:**Interestingly, if the hypothesis that δ = 0 is rejected (i.e., the time series is stationary), we can use the usual (Student’s)

*t*test.

The unit root test can be
computed under three (3) different null hypotheses. That is, under different
model specifications such as if the series is a:

1. random walk (that is, model has no
constant, no trend)

2. random walk with drift (that is, model
has a constant)

3. random walk with drift and a trend (that
is, model has a constant and trend)

In all cases, the null
hypothesis is that δ = 0; that is, there is a unit root and the alternative
hypothesis is that δ is less than zero; that is, the time series is stationary.
If the null hypothesis is rejected, it means that

*Y*is a stationary time series with zero mean in the case of [5], that_{t}*Y*is stationary with a nonzero mean in the case of a random walk with drift model, and that_{t}*Y*is stationary around a deterministic trend in the case of random walk with drift around a trend._{t}
It is extremely important to
note that the critical values of the

*tau test*to test the hypothesis that δ = 0, are different for each of the preceding three speciﬁcations of the DF test, which are now computed by all econometric packages. In each case, if the computed absolute value of the*tau statistic*(|τ|) exceeds the DF or MacKinnon critical*tau values*, the null hypothesis of a unit root is rejected, in order words the time series is stationary. On the other hand, if the computed |τ| does not exceed the critical*tau*value, we fail to reject the null hypothesis, in which case the time series is nonstationary.**Note:**Students often get confused in interpreting the outcome of a unit root test. For instance, if the calculated

*tau*statistic is -2.0872 and the MacKinnon

*tau*statistic is -3.672, you cannot reject the null hypothesis. Hence, the conclusion is that the series is nonstationary. But if the calculated

*tau*statistic is -5.278 and the MacKinnon

*tau*statistic is -3.482, you reject the null hypothesis in favour of the alternative. Hence, the conclusion is that the series is stationary.

*Always use the appropriate critical τ values
for the indicated model specification.

**How to Perform Unit Root Test in Stata (see here for EViews)**

Several tests have
been developed in the literature to test for unit root. Prominent among these
tests are Augmented Dickey-Fuller, Phillips-Perron, Dickey-Fuller Generalised
Least Squares (DF-GLS) and so on. But this tutorials limits testing to the use
of ADF and PP tests. Once the reader has good basic knowledge of these two
techniques, they can progress to conducting other stationarity test on their
time series variables.

**How to Perform the Augmented Dickey-Fuller (ADF) Test**

An important assumption of
the DF test is that the error terms are independently and identically
distributed. The ADF test adjusts the DF test to take care of possible serial
correlation in the error terms by adding the lagged difference terms of the
outcome (dependent) variable. For

*Y*series, in conducting the DF test, it is assumed that the error term_{t}*v*is uncorrelated. But in case where it is correlated, Dickey and Fuller have developed a test, known as the augmented Dickey–Fuller (ADF) test. This test is conducted by “augmenting” the preceding three model specifications stated above by adding the lagged values of the dependent variable._{t}
As mentioned earlier,
approaches will be limited to using the ADF and PP tests. Either of these tests
can be used and when both are used, the reader can compare the outcomes to see
if there are similarities or differences in the results.

1.
Using Gujarati and Porter Table 21.1 quarterly data on

*pce, pdi*and*gdp*
2.
Load data into Stata

3.
We are considering the pair of

*lnpce*and*lnpdi*in natural logarithms (because variables are measured in US$ billions)
4.
Inform Stata that you are about to perform a time series
analysis by typing this code into the

**Command**box:*tsset qtrly*
and
you will obtain this:

Stata: tsset command Source: CrunchEconometrix |

Stata now recognises that you are about conducting a time series analysis using quarterly data from 1

^{st}quarter of 1970 to the 4

^{th}quarter of 1991. If you don’t issue this command, Stata will not run your analysis.

Note: if you
are using a yearly data, type the syntax

*tsset year*and if it is a monthly data type*tsset month*.**The Augmented Dickey-Fuller (ADF) Test**

Unit root test for

*lnpce*:
· Go
to

**Statistics**>>**Time series**>>**Tests >> Augmented Dickey-Fuller unit root test**>> dialog box opensStata: Augmented Dickey-Fuller Dialog Box Source: CrunchEconometrix |

· Under

**Variable**, select*lnpce*
·
Under

**Options**, the choice of model is very important since the distribution statistic under the null hypothesis differs across these three cases. Therefore, specify whether to**“suppress constant term”, “include trend term”,**or**“include drift term”**. Thus, our demonstration will involve these options.
· If the

**“display regression table”**box is checked, Stata reports the test statistic together with the estimated test regression.
· Depending on the structure of your data, include the number of

**“lagged differences”****Decision:**The null hypothesis of a unit root is rejected against the one-sided alternative hypothesis if the computed absolute value of the

*tau statistic*exceeds the DF or MacKinnon critical tau values and we conclude that the series is stationary; otherwise (that is, if it is lower), then the series is non-stationary.

**Decision:**Another way of stating this is that in failing to reject the null hypothesis of a unit root, the computed τ value should be

**negative than the critical τ value. Since in general δ is expected to be negative, the estimated τ statistic will have a negative sign. Therefore, a large negative τ value is generally an indication of stationarity.**

__more__**Decision:**On the other hand, using the probability value, we reject the null hypothesis of unit root if the computed probability value is less than the chosen level of statistical significance.

Having specified the

**“suppress constant term”**and checked the**“display regression table”**box, the ADF result is given as:Stata: ADF test, level, "suppress constant" option Source: CrunchEconometrix |

Following similar
procedures, the select

**“include trend term”**for the ADF unit root test yields:Stata: ADF test, level, "include trend term" option Source: CrunchEconometrix |

The result for the “

**include drift term**” option for the ADF unit root test is shown below:Stata: ADF test, level, "include drift term" option Source: CrunchEconometrix |

**Note:**the null hypothesis for the three ADF specifications

__be rejected at the 5% level, confirming that__

*cannot**lnpce*is nonstationary which is a confirmation of the graphical plot. Notice that the interpolated Dickey-Fuller tau statistic differ across all specifications.

Do same for the

*lnpdi*series.
Having confirmed that

*lnpce*is nonstationary, we need to run the tests again using its first difference. So, the next thing to do is to generate the first difference of*lnpce*and run the test across the three specifications.
To generate the difference
variable, the syntax is:

*generate dlnpce=d.lnpce*

·
1

^{st}difference with**“suppress constant”**option result:Stata: ADF test, 1st difference, "suppress constant" option Source: CrunchEconometrix |

1

^{st}difference with**“include trend term”**option result:Stata: ADF test, 1st difference, "include trend term" option Source: CrunchEconometrix |

·
1

^{st}difference with**“include drift term”**option result:Stata: ADF test, 1st difference, "include drift term" option Source: CrunchEconometrix |

Having done the
first-difference analysis, and the trend term is not statistically significant,
we conclude that the null hypotheses of a unit root is rejected and that

*lnpce*series is difference-stationary. That is,**hence, carrying out a***lnpce*is stationary at 1^{st}difference with a constant**“2**test is unnecessary.^{nd}difference”
Again, do same for the

*dlnpdi*series.**How to Perform the Phillips-Perron (PP) Test**

Phillips and Perron use
nonparametric statistical methods to take care of the serial correlation in the
error terms

**adding lagged difference terms. Procedures for testing for unit root using the PP test differs a bit from that of ADF.**__without__
·
Go to

**Statistics**>>**Time series**>>**Tests >> Phillips-Perron unit root test**>> dialog box opens
Fill the details of the
variables in the

**“Variable”**box and indicate which specification to run, then click**OK**. Or, you can use the following codes for the different specifications:
For the

**“level”**specification, the syntax is:
(For

**“Suppress constant term in regression”**)*pperron lnpce, noconstant regress*

(For

**“include trend term in regression”**)*pperron lnpce, trend regress*

For the

**“1**specification, the syntax is:^{st}difference”
(For

**“Suppress constant term in regression”**)*pperron dlnpce, noconstant regress*

(For

**“include trend term in regression”**)*pperron dlnpce, trend regress*

In comparing the results from both procedures, the same conclusion is arrived at. That is,

*lnpce*is difference-stationary.**Note:**The asymptotic distribution of the PP test is the same as the ADF test statistic.

**After unit root testing, what next?**

The outcome of unit root
testing matters for the empirical model to be estimated. The following
scenarios explain the implications of unit root testing for further
analysis.

**Scenario 1: When series under scrutiny are stationary in levels?**

If

*pce*and*pdi*are stationary in levels, that is, they are*I*(0) series (integrated of order zero). In this situation, performing a cointegration test is**necessary. This is because any shock to the system in the short run quickly adjusts to the long run. Consequently, only the long run model should be estimated. That is, the model should be specified as:**__not__*pce*= 𝛂₀ +

_{t }*bpdi*+

_{t}*u*

_{t}
In essence, the estimation
of short run model is not necessary if series are

*I*(0).**Scenario 2: When series are stationary in first differences?**

· Under this scenario, the series are assumed to be non-stationary.

· One special feature of these series is that they are of the same
order of integration.

· Under this scenario, the model in question is not entirely useless
although the variables are unpredictable. To verify further the relevance of
the model, there is need to test for cointegration. That is, can we
assume a long run relationship in the model despite the fact that the series
are drifting apart or trending either upward or downward?

· If there is cointegration, that means the series in question are
related and therefore can be combined in a linear fashion. This implies that,
even if there are shocks in the short run, which may affect movement in the
individual series, they would converge with time (in the long run).

· However, there is no long run if series are not cointegrated. This
implies that, if there are shocks to the system, the model is not likely to
converge in the long run.

· Note that both long run and short run models must be estimated
when there is cointegration.

· The estimation will require the use of vector autoregressive (VAR)
model analysis and VECM models.

· If there is no cointegration, there is no long run and therefore,
only the short run model will be estimated. That is, run only VAR no VECM
analysis!

· There are however, two prominent cointegration tests for

*I*(I) series in the literature. They are Engle-Granger cointegration test and Johansen cointegration test.
· The Engle-Granger test is meant for single equation model while
Johansen is considered when dealing with multiple equations.

**Scenario 3: The series are integrated of different orders?**

· Should in case

*lnpce*and*lnpdi*are integrated of different orders, like the second scenario, cointegration test is also required but the use of either Engle-Granger or Johansen cointegration are no longer valid.
· The appropriate cointegration test to apply is the Bounds test for
cointegration and the estimation technique is the autoregressive distributed
lag (ARDL) model.

· Similar to case 2, if series are not cointegrated based on Bounds
test, we are expected to estimate only the short run. That is run only the ARDL
model.

· However, both the long run and short run models are valid if there
is cointegration. That is run both ARDL and ECM models.

In addition, there are
formal tests that can be carried out to see if despite the behaviour of the
series, there can still be a linear combination or long run relationship or
equilibrium among the series. The existence of the linear combination is what
is known as cointegration. Thus, the regression with

*I*(1) series can either be spurious or cointegrated. The basic single equation cointegration tests are Johansen, Engle-Granger and Bounds cointegration tests. These will be discussed in detail in subsequent tutorials.**[Watch video clip on performing the ADF unit root test in Stata]**

Please
post your comments below….