Announcement

Collapse
No announcement yet.
X
  • Filter
  • Time
  • Show
Clear All
new posts

  • Hausman - the value of prob>chi2 doesn't appear

    Data Description

    I'm currently working on my thesis and I encountered difficulties when processing the data. So I tested several factors that can influence bad credit at the group of banks.
    I have 1 dependent variable named Non-Performing Loan (NPL) which is a loan that hasn't been paid in a certain period of time. The independent variable are:
    1. Load to Deposit Ratio (LDR) which is the ratio used to measure how much credit is disbursed compared to the funds obtained
    2. Return on Asset (ROA) which is ratio that measure how well the bank using its asset to produce profit
    3. Inflation
    4. Gross Domestic Product (GDP)

    The Problem

    The result shows that the prob>chi2 value at Hausman test does not appear and there is an explanation V_b - v_B not positive definite (as you can see in the picture)

    Click image for larger version

Name:	xttest.jpeg
Views:	1
Size:	10.9 KB
ID:	1716261





    fe result
    Click image for larger version

Name:	fe.jpeg
Views:	1
Size:	66.9 KB
ID:	1716262





    re result
    Click image for larger version

Name:	re.jpeg
Views:	1
Size:	64.9 KB
ID:	1716263





    hausman
    Click image for larger version

Name:	hausman.jpeg
Views:	1
Size:	50.9 KB
ID:	1716264





    I give you link to my excel data so you can see the data distribution. https://docs.google.com/spreadsheets...f=true&sd=true

    Based on the result above, hausman doesn't give me the value of prob>chi2 so I can't choose weather FE or RE are the best model.
    Would you like to explain what happened with this data and help me solve this problem without eliminating the independent variable? Thank you in advance!
    Last edited by Luthfi Syarhan; 06 Jun 2023, 12:14.

  • #2
    It appears that you stored the RE estimates twice.

    Code:
    xtset company year
    xtreg npl ldr roa inflation gdp, fe
    est sto fe
    xtreg npl ldr roa inflation gdp, re
    hausman fe .

    Comment


    • #3
      Thank you so much for your help! I think i missed some steps last night.

      Comment

      Working...
      X