4. Stationarity tests

Almost all the models proposed for time series have stationarity as an assumption. Therefore, a fundamental question is when a time series is stationary. In the Stationarity Tests tool, three tests are available to verify the stationarity of the series.

Example:

We will perform the test for the following data:

---

AMBV3
84.06
83.85
83.56
83.47
83.27
82.81
82.20
82.06
81.62
80.77
81.30
81.92
82.75
82.77
82.84
82.82
82.72
82.29
81.18
80.11
80.27
80.21
79.92
79.96
80.19
80.17
80.17
79.85
81.00
80.44
79.96
79.85
79.82
80.11
80.20
80.31
81.18
80.81
81.15
81.32
81.21
81.40
81.10
81.40
82.27
82.15
81.78
81.69
81.34
81.88

---

We will upload the data to the system.

Configuring as shown in the figure below to we will carry out the test.

Then click Calculate to get the results. You can also generate the analyses and download them in Word format.

The results are:

Augmented Dickey-Fuller test

	Dickey-Fuller
Statistics	-1.90501089436462
P-value	0.612390335192691
Sample size	50
Null hypothesis	There is at least one unit root
Alternative hypothesis	There is no unit root

---

Phillips-Perron test

	Dickey-Fuller Z(alpha)
Statistics	-6.09026378709285
P-value	0.753893665311693
Sample size	50
Null hypothesis	There is at least one unit root
Alternative hypothesis	There is no unit root

---

KPSS Test

	KPSS Level
Statistics	0.540460035842587
P-value	0.0325540459813994
Sample size	50
Null Hypothesis	The time series is stationary.
Alternative Hypothesis	The time series has a unit root.

---