Scribd Logo
Upload

Welcome to Scribd!

  • 47 views

    Example Fta Unit Root Tests: ST ND ST

    Uploaded by

    parcaperu
    The document provides instructions for replicating unit root tests using example data in an EViews workfile. It describes: 1) Opening the existing workfile containing time series data and creating additional logged and transformed series. 2) Selecting the logged FTAPRICE series and setting the sample period before running unit root tests (ADF, PP, KPSS) with different specifications. 3) Interpreting the test statistics and conclusions about whether a unit root can be rejected depending on the critical values. 4) Repeating the unit root tests for other series in levels and differences to determine whether they are stationary or integrated of order 1 or 2.

    Copyright:

    © All Rights Reserved
    Download as DOC, PDF, TXT or read online from Scribd

    Example Fta Unit Root Tests: ST ND ST

    Uploaded by

    parcaperu
    47 views1 page
    The document provides instructions for replicating unit root tests using example data in an EViews workfile. It describes: 1) Opening the existing workfile containing time series data and creating additional logged and transformed series. 2) Selecting the logged FTAPRICE series and setting the sample period before running unit root tests (ADF, PP, KPSS) with different specifications. 3) Interpreting the test statistics and conclusions about whether a unit root can be rejected depending on the critical values. 4) Repeating the unit root tests for other series in levels and differences to determine whether they are stationary or integrated of order 1 or 2.

    Original Description:

    raiz unitaria

    Original Title

    Ex1 Unit Root

    Copyright

    © © All Rights Reserved

    Available Formats

    DOC, PDF, TXT or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    The document provides instructions for replicating unit root tests using example data in an EViews workfile. It describes: 1) Opening the existing workfile containing time series data and creating additional logged and transformed series. 2) Selecting the logged FTAPRICE series and setting the sample period before running unit root tests (ADF, PP, KPSS) with different specifications. 3) Interpreting the test statistics and conclusions about whether a unit root can be rejected depending on the critical values. 4) Repeating the unit root tests for other series in levels and differences to determine whether they are stationary or integrated of order 1 or 2.

    Copyright:

    © All Rights Reserved
    Download as DOC, PDF, TXT or read online from Scribd
    Download as doc, pdf, or txt
    47 views1 page

    Example Fta Unit Root Tests: ST ND ST

    Uploaded by

    parcaperu
    The document provides instructions for replicating unit root tests using example data in an EViews workfile. It describes: 1) Opening the existing workfile containing time series data and creating additional logged and transformed series. 2) Selecting the logged FTAPRICE series and setting the sample period before running unit root tests (ADF, PP, KPSS) with different specifications. 3) Interpreting the test statistics and conclusions about whether a unit root can be rejected depending on the critical values. 4) Repeating the unit root tests for other series in levels and differences to determine whether they are stationary or integrated of order 1 or 2.

    Copyright:

    © All Rights Reserved
    Download as DOC, PDF, TXT or read online from Scribd
    Download as doc, pdf, or txt
    You are on page 1of 1

    EXAMPLE FTA UNIT ROOT TESTS

    This small guide shows you how to recreate the results for the example used in class. For the corresponding
    graphs and tables see the sheets/handouts on this topic. I used EViews5.

    The following steps are deliberately kept easy, using the EViews point-and-click approach. However, I assume
    you have familiarized yourself with the operation of EViews and understand the various routes one can take to
    obtain the results described below.
    There is an EViews program (list of commands executed automatically) associated with this example that shows
    you how to use the appropriate command statements and do some programming.

    START EVIEWS

    Select Menu options File/Open/EViewsWorkfile


    Select the existing workfile FTA.wf1 from the directory where you saved it.
    The workfile contains the time series data for the FT-All Share price index, FT-All Share dividend index, yield
    on 20 year UK government bonds (Gilts), rate on 91 day UK Treasury bills. Monthly observations, January 1965
    (1965:01 or 1965M01) to December 1995 (1995:12 or 1995M12).

    DATA TRANSFORMATIONS
    Additional series from data transformations are created in the command window at the top as
    smpl @all ‘ reset the sample period for calculations to the total sample used in the workfile
    ‘ use all available observations
    series lprice = log(ftaprice) ‘ create (natural) log version of the variable

    UNIT ROOT TESTS


    Select the series lnftaprice
    Set the sample period. In this case use the full sample period: smpl @all. Note that EViews will adjust the
    sample for the lags used in the unit root tests. If you want/need to avoid this, think about where the correct
    sample period should start and set the smpl command accordingly.
    Select View/Unit Root Test,
    Select the desired test: ADF, PP, KPSS (or one of the other methods, which we will not discuss)
    Test for unit root in level
    Select the desired options on inclusion/exclusion of constant term and trend variable
    Select the option on lag length determination: AIC, SIC, HIC, or set the lags to some predetermined value
    Note: For PP and KPSS test statistics, we do not go into details about the selection of Spectral estimation and
    Bandwith selection methods. Rely on the EViews default settings for these tests.

    Normally the unit root tests are performed for levels, 1st differences, 2nd differences of the variables. If we cannot
    reject a unit root in levels, but do reject a unit root in 1st differences the variable in levels is called I(1), i.e.
    needing one time differencing to become stationary. The series in first-differences is stationary and called I(0). If
    we cannot reject a unit root in 1st differences, but do reject a unit root in 2nd differences the variable in levels is
    called I(2), i.e. needing two times differencing to become stationary.
    Do we reject the null hypothesis of a unit root (ADF, PP) or null hypothesis of stationarity in the case of the
    KPSS test? Look up the correct test-statistic and compare its value to the EViews reported critical values and/or
    look up the corresponding probability value to determine the significance level.

    Repeat the procedure for other variables


    ldiv = log(ftadiv)
    r20
    rs

    You might also like