Scribd
Upload
90 views8 pages

Econometric Modeler Results Summary

The document summarizes the results of time series analysis on the sales1 time series data. [1] It plots the raw sales1 data and performs an Augmented Dickey-Fuller test, failing to reject the null hypothesis that the data contains a unit root. [2] It then takes the seasonal difference of sales1 and plots it, before [3] taking the difference of that seasonal differenced series and plotting and testing it, rejecting the null hypothesis that this final series contains a unit root, indicating the data is now stationary.

Uploaded by

Omi
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
90 views8 pages

Econometric Modeler Results Summary

The document summarizes the results of time series analysis on the sales1 time series data. [1] It plots the raw sales1 data and performs an Augmented Dickey-Fuller test, failing to reject the null hypothesis that the data contains a unit root. [2] It then takes the seasonal difference of sales1 and plots it, before [3] taking the difference of that seasonal differenced series and plotting and testing it, rejecting the null hypothesis that this final series contains a unit root, indicating the data is now stationary.

Uploaded by

Omi
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd

Econometric Modeler Analysis

Summary of results from the Econometric


Modeler App
Econometrics Toolbox Version 5.6 (R2021a)

14-Aug-2021
Table of Contents
1. Time Series: sales1 ...................................................................................................................................... 2
1.1. Time Series Plot ............................................................................................................................... 2
1.2. Augmented Dickey-Fuller Test ....................................................................................................... 3
2. Time Series: sales1SeasonalDiff................................................................................................................. 4
2.1. Time Series Plot ............................................................................................................................... 4
3. Time Series: sales1SeasonalDiffDiff .......................................................................................................... 5
3.1. Time Series Plot ............................................................................................................................... 5
3.2. Augmented Dickey-Fuller Test ....................................................................................................... 6

i
1. Time Series: sales1
1.1. Time Series Plot

Time Series Plot


1100
sales1
1000

900

800

700

600

500

400

300

200
10 20 30 40 50 60 70 80 90 100 110 120
Index

Figure 1.1. Time Series Plot of sales1

2
1. Time Series: sales1

1.2. Augmented Dickey-Fuller Test


Null Hypothesis: sales1 contains a unit root

Table 1.1. Test Parameters


Lags Model Test Statistic Significance Level
1 0 AR t1 0.05

Table 1.2. Test Results


Null Rejected P-Value Test Statistic Critical Value
1 false 0.2363 -1.13 -1.9437

3
2. Time Series: sales1SeasonalDiff
2.1. Time Series Plot

Time Series Plot


200
sales1SeasonalDiff

150

100

50

-50

-100
10 20 30 40 50 60 70 80 90 100 110 120
Index

Figure 2.1. Time Series Plot of sales1SeasonalDiff

4
3. Time Series: sales1SeasonalDiffDiff
Time series sales1SeasonalDiffDiff is the first-order difference of time series
sales1SeasonalDiff.

3.1. Time Series Plot

Time Series Plot


250
sales1SeasonalDiffDiff
200

150

100

50

-50

-100

-150

-200
10 20 30 40 50 60 70 80 90 100 110 120
Index

Figure 3.1. Time Series Plot of sales1SeasonalDiffDiff

5
3. Time Series: sales1SeasonalDiffDiff

3.2. Augmented Dickey-Fuller Test


Null Hypothesis: sales1SeasonalDiffDiff contains a unit root

Table 3.1. Test Parameters


Lags Model Test Statistic Significance Level
1 0 AR t1 0.05

Table 3.2. Test Results


Null Rejected P-Value Test Statistic Critical Value
1 true 0.001 -19.0973 -1.9442

6
3. Time Series: sales1SeasonalDiffDiff

Common questions

Powered by AI

The first-order differencing of sales1SeasonalDiff likely results in stationarity for sales1SeasonalDiffDiff because differencing is a common method to remove non-stationary characteristics such as trends and seasonality in a time series. By further differencing sales1SeasonalDiff, the remaining seasonal and trend components in the original series are eliminated, ensuring stationarity as confirmed by the Augmented Dickey-Fuller Test .

The sales1 time series contains a significant trend component, as indicated by the non-rejection of the null hypothesis for unit root based on the Augmented Dickey-Fuller Test. The differenced series, sales1SeasonalDiffDiff, becomes stationary, which implies that differencing was necessary to remove the trend component .

Transforming sales1 into sales1SeasonalDiff does not immediately imply stationarity, but the transformation of sales1SeasonalDiff into sales1SeasonalDiffDiff ensures stationarity. The Augmented Dickey-Fuller Test results show that the null hypothesis of a unit root is rejected for sales1SeasonalDiffDiff, with a test statistic of -19.0973 significantly lower than the critical value of -1.9442, and a p-value of 0.001 .

Further econometric modeling of sales1 should proceed using the sales1SeasonalDiffDiff series, as this series has been successfully transformed into a stationary form. This transformation involves applying differencing techniques to remove trends and seasonality features highlighted by the non-stationarity observed in the original sales1 series, thereby providing a reliable foundation for predictive modeling or any econometric analysis .

The document suggests using differentiation techniques to mitigate the effects of trends and seasonal patterns inherent in time series data, facilitating the transformation into a stationary series. By first differencing sales1 to get sales1SeasonalDiff and then further differencing to get sales1SeasonalDiffDiff, the series is effectively processed to eliminate trends and achieve stationarity, essential for robust econometric modeling .

The p-value of 0.001 for sales1SeasonalDiffDiff in the Augmented Dickey-Fuller Test indicates that there is a very low probability of incorrectly rejecting the null hypothesis that this series contains a unit root. This strong statistical result suggests that any trends or seasonality present in the raw data have been effectively removed through differencing .

Achieving stationarity is crucial for accurate econometric modeling because non-stationary data can lead to spurious regression results, making predictions unreliable. The document reflects this process through the application of the Augmented Dickey-Fuller Test and subsequent conversions from sales1 to sales1SeasonalDiffDiff, demonstrating the necessary steps to ensure the data's appropriateness for modeling purposes .

The time series plot of sales1 shows apparent trends, indicating non-stationarity. In contrast, the plot for sales1SeasonalDiff shows reduced overall trends and flatter segments compared to sales1, but does not outright ensure stationarity. Finally, the plot of sales1SeasonalDiffDiff displays oscillating behavior around a constant mean, reflecting the absence of any discernible trend or seasonal component, thus suggesting stability .

The test statistic for sales1SeasonalDiffDiff is -19.0973, significantly lower than the critical value of -1.9442, demonstrating strong evidence against the presence of a unit root. This result indicates stationarity, confirming that the time series no longer contains unit root characteristics after differencing, which is vital for subsequent time series analysis and model accuracy .

The Augmented Dickey-Fuller Test indicates that the time series sales1 is non-stationary as the null hypothesis, which states that sales1 contains a unit root, cannot be rejected. The p-value is 0.2363, which is higher than the significance level of 0.05, and the test statistic is -1.13, which is greater than the critical value of -1.9437 .

You might also like