[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

From |
Daniel Wilde <dgw24@bath.ac.uk> |

To |
statalist@hsphsun2.harvard.edu |

Subject |
st: XTABOND2 tests of autocorrelation and restrictions (fwd) |

Date |
Thu, 22 Nov 2007 16:32:46 +0000 |

Hi, All

I originally posted this email a few days ago, but as I didn't get any replies I thought I would try again. I have also added two more questions. I know the questions may appear simple to some people, but I'd greatly appreciate any help.

Using the XTABOND2 command I get significant and expected results. But I do

not understand how to read the output for the autocorrelation test or the

Hansen test. More specifically:

1) Is the Arellano-Bond test for AR(1) in first differences testing for

autocorrelation in the full error term Eit or is it testing for

autocorrelation in the Vit component of the error term.

2) What are the null and alternative hypothesizes in the Arellano Bond

test. Does a P-value of 0.32 indicate that there is or is not

autocorrelation?

3) If the P ? Value is showing that autocorrelation is occurring in Vit ?

how can this be dealt with in the XTABOND2 command?

4) What are the null and alternative hypotheses in the Hansen test? Does a

P-value of 0.271 indicate that I have used too many instruments?

5) Is it acceptable to use a transformation of the dependent variable as a lag. E.G if the dependent variable is a log of data75 can we use data75 as the instrument.

6) If I add additional instruments that I have not included in the original regression equation, should these be in addition to, or instead of, the instruments which are lagged versions of the independent variables?

Any help would be greatly appreciated.

Thanks

Daniel Wilde

FULL RESULTS ARE:

. xi: xtabond2 lngdata75 l.lngdata75 lngdata36 lngdata38 lngdata99 i.time5,

gmm(l.data75) iv(lngdata36 lngdata38 lngdata99 i.time5) level robust

i.time5 _Itime5_1-10 (naturally coded; _Itime5_1 omitted)

_Itime5_2 dropped because of collinearity.

_Itime5_3 dropped because of collinearity.

_Itime5_4 dropped because of collinearity.

_Itime5_5 dropped because of collinearity.

_Itime5_6 dropped because of collinearity.

_Itime5_7 dropped because of collinearity.

_Itime5_10 dropped because of collinearity.

Building GMM instruments..

7 instrument(s) dropped because of collinearity.

Estimating.

Performing specification tests.

Dynamic panel-data estimation, one-step system GMM

---------------------------------------------------------------------------

---

Group variable: unit_id Number of obs =

182

Time variable : time5 Number of groups =

103

Number of instruments = 26 Obs per group: min =

1

Wald chi2(5) = 9856.70 avg =

1.77

Prob > chi2 = 0.000 max =

3

---------------------------------------------------------------------------

---

| Robust

lngdata75 | Coef. Std. Err. z P>|z| [95% Conf.

Interval]

-------------+-------------------------------------------------------------

---

lngdata75 |

L1 | .7644667 .0506732 15.09 0.000 .6651489

.8637844

lngdata36 | -.2266392 .0425735 -5.32 0.000 -.3100817

-.1431967

lngdata38 | .1954291 .0653342 2.99 0.003 .0673764

.3234818

lngdata99 | -.2925302 .1153542 -2.54 0.011 -.5186202

-.0664401

_Itime5_8 | .0324644 .0234058 1.39 0.165 -.01341

.0783389

_Itime5_9 | -.0185157 .0244261 -0.76 0.448 -.06639

.0293586

_cons | 3.23673 .8508868 3.80 0.000 1.569023

4.904438

---------------------------------------------------------------------------

---

Instruments for first differences equation

Standard

lngdata36 lngdata38 lngdata99 _Itime5_2 _Itime5_3 _Itime5_4 _Itime5_5

_Itime5_6 _Itime5_7 _Itime5_8 _Itime5_9 _Itime5_10

GMM-type (separate instruments for each period)

L(1/.).L.data75

Instruments for levels equation

Standard

_cons

lngdata36 lngdata38 lngdata99 _Itime5_2 _Itime5_3 _Itime5_4 _Itime5_5

_Itime5_6 _Itime5_7 _Itime5_8 _Itime5_9 _Itime5_10

GMM-type (separate instruments for each period)

D.L.data75

(Instrument count reported above excludes 7 of these as collinear.)

---------------------------------------------------------------------------

---

Arellano-Bond test for AR(1) in first differences: z = -0.99 Pr > z =

0.321

Arellano-Bond test for AR(2) in first differences: z = . Pr > z =

.

---------------------------------------------------------------------------

---

Sargan test of overid. restrictions: chi2(19) = 114.77 Prob > chi2 =

0.000

(Not robust, but not weakened by many instruments.)

Hansen test of overid. restrictions: chi2(19) = 22.26 Prob > chi2 =

0.271

(Robust, but can be weakened by many instruments.)

.

---------- End Forwarded Message ----------

*

* For searches and help try:

* http://www.stata.com/support/faqs/res/findit.html

* http://www.stata.com/support/statalist/faq

* http://www.ats.ucla.edu/stat/stata/

- Prev by Date:
**Re: st: Logit/probit across different groups** - Next by Date:
**st: More subtle use of by-plots** - Previous by thread:
**st: How do I get more than 244 characters into my return local macro?** - Next by thread:
**st: More subtle use of by-plots** - Index(es):

© Copyright 1996–2015 StataCorp LP | Terms of use | Privacy | Contact us | What's new | Site index |