[Nicolas Syrichas]

Dear Joao Santos Silva. I am interested to test formally whether the fixed effects are simply location shifters or if they affect the entire distribution. In this case I should test whether delta_i equals zero(assuming panel data). Do you think a simple F-test will work here and I will appreciate if you could provide an example how to test is in stata

[Nicolas Syrichas]

Further to my email above. I am not an econometrician but I thought about it. Assuming Z=X and errors that are normally distributed then the absolute value of the residual term should be folded normally distributed.

[Joao Santos Silva]

Dear Nicolas Syrichas,

Indeed that is what you need to do. I believe Stata reports such test, so all you have to do it to perform the second step and do the test. You can also go to the ado file and remove the "qui" in line 75; the test should then be automatically reported.

Best wishes,

Joao

[Fadi Ansar]

Dear Joao Santos Silva
I am having some confusions regrading the discussion we had before and I would like to add this to the earlier discussion
a)if my T is less than N (my case T=19 and 56) using xtqreg command will be an issue and if it is a problem which command should i use -In such cases what we will do to run quantile regression
b)Since quantile regressions are used for non normal distribution ,do I need to check the CLRM assumptions


Can you please help me to clear this confusions

Thankyou so much

[Joao Santos Silva]

Dear Fadi Ansar,

a) What is you N? Anyway, with T = 19 or 56 you should be OK.
b) Quantile regression can be used with any distribution, including the normal. You do not need to check the CLRM assumptions, but I suggest you compute the standard errors by bootstrap.

Best wishes,

Joao

[Fadi Ansar]

Thankyou so much Joao Santos Silva .In my case my T=14and N=70 ,so I can do my quantile regression with this data .
with regard to bootstrap regression , I just have to use the appropriate commands nothing else right ,
Once again thanks for clarifying suggestions .

[Fadi Ansar]

I just want to add my question clearly ,Can I use xtqreg command if my T=14 and N =70 ,if not what is the solution

[Joao Santos Silva]

Yes, that is fine.

[Fadi Ansar]

Thanks again Joao Santos Silva

[Fadi Ansar]

Dear Joao Santos Silva .I would like to one more query I encountered while running xtqreg with bootstrap standard errors .
I have used the following commands and got the following results.

bootstrap ,cl( bankname) r(100) id(id) seed(123):xtqreg Zscore MP_callmoneyrate Boone_Ind TotalAsset Liquidi
> tyRatio CBT FinFreedom NIM RI AssetcompositionRatiooffixe GDP INF Ratioofnonintrestoperating MP##C.MP_callmo
> neyrate owmpmr i.Year i.ownership ,quantile (.1 ) i(id)
(running xtqreg on estimation sample)

Bootstrap replications (100)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5 
xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx    50
xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx   100
insufficient observations to compute bootstrap standard errors
no results will be saved
r(2000);

Is this because of data issue or is it because I added year and ownership effects in to the model -the command was working perfectly fine when there is no year dummy and ownership dummy.
Can you please provide your insight on this

Thanks in advance

[Joao Santos Silva]

Dear Fadi Ansar,

It is difficult to advise without further information; please show us the results you get when you estimate without bootstrapping.

Best wishes,

Joao

[Fadi Ansar]

Thankyou Joao Santos Silva .I am attaching my results here
Results when using bootsrap with out i.year and I. Ownership variables

bootstrap ,cl( bankname) r(100) id(id) seed(123):xtqreg Zscore MP_callmoneyrate Boone_Ind TotalAsset Liquidi
> tyRatio CBT FinFreedom NIM RI AssetcompositionRatiooffixe GDP INF Ratioofnonintrestoperating MP##C.MP_callmo
> neyrate owmpmr ,quantile (.9) i(id)
(running xtqreg on estimation sample)

Bootstrap replications (100)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5
................................................x.    50
...x..............................................   100

Bootstrap results                               Number of obs     =        703
                                                Replications      =         98

                                              (Replications based on 83 clusters in bankname)
---------------------------------------------------------------------------------------------
                            |   Observed   Bootstrap                         Normal-based
                     Zscore |      Coef.   Std. Err.      z    P>|z|     [95% Conf. Interval]
----------------------------+----------------------------------------------------------------
           MP_callmoneyrate |   .2275783   .0295831     7.69   0.000     .1695965      .28556
                  Boone_Ind |  -28.36535   2.544262   -11.15   0.000    -33.35201   -23.37869
                 TotalAsset |   6.06e-07   6.61e-07     0.92   0.359    -6.88e-07    1.90e-06
             LiquidityRatio |   .9405847   .8600589     1.09   0.274    -.7450998    2.626269
                        CBT |  -.2851305    .071502    -3.99   0.000    -.4252717   -.1449892
                 FinFreedom |   .0513436   .0121327     4.23   0.000     .0275641    .0751232
                        NIM |   .0446368   .1260745     0.35   0.723    -.2024647    .2917383
                         RI |   .0006666   .0003443     1.94   0.053    -8.08e-06    .0013414
AssetcompositionRatiooffixe |  -11.17763   6.237777    -1.79   0.073    -23.40345    1.048183
                        GDP |   .0327408   .0321081     1.02   0.308    -.0301899    .0956715
                        INF |  -.0023909   .0011611    -2.06   0.039    -.0046666   -.0001152
 Ratioofnonintrestoperating |   .0028424   .1001218     0.03   0.977    -.1933926    .1990775
                            |
                         MP |
                       TMP  |   .5958753   .3987544     1.49   0.135     -.185669     1.37742
           MP_callmoneyrate |          0  (omitted)
                            |
      MP#c.MP_callmoneyrate |
                       TMP  |  -.1334673   .0628047    -2.13   0.034    -.2565622   -.0103724
                            |
                     owmpmr |   .0095196   .0107634     0.88   0.376    -.0115763    .0306155
---------------------------------------------------------------------------------------------
Note: One or more parameters could not be estimated in 2 bootstrap replicates;
      standard-error estimates include only complete replications.

Result without bootstrap standard error

xtqreg Zscore MP_callmoneyrate Boone_Ind TotalAsset LiquidityRatio CBT FinFreedom NIM RI AssetcompositionRat
> iooffixe GDP INF Ratioofnonintrestoperating MP##C.MP_callmoneyrate# ownership owmpmr ,quantile (.1 .2 .3 .4)

-------------------------------------------------------------------------------------------------
| Coef. Std. Err. z P>|z| [95% Conf. Interval]
--------------------------------+----------------------------------------------------------------
MP_callmoneyrate | -.0052286 .1131619 -0.05 0.963 -.2270219 .2165646
Boone_Ind | -21.26151 2.07109 -10.27 0.000 -25.32077 -17.20225
TotalAsset | 1.12e-07 2.91e-07 0.39 0.700 -4.58e-07 6.82e-07
LiquidityRatio | .177064 .337649 0.52 0.600 -.484716 .8388439
CBT | -.3767358 .0689395 -5.46 0.000 -.5118548 -.2416168
FinFreedom | .1693787 .0097297 17.41 0.000 .1503089 .1884485
NIM | .0076053 .0435083 0.17 0.861 -.0776695 .0928801
RI | .0000993 .0001536 0.65 0.518 -.0002018 .0004004
AssetcompositionRatiooffixe | -2.268918 3.057144 -0.74 0.458 -8.26081 3.722975
GDP | .0154013 .0166414 0.93 0.355 -.0172153 .048018
INF | -.0005013 .0011201 -0.45 0.655 -.0026965 .001694
Ratioofnonintrestoperating | .0167296 .0331114 0.51 0.613 -.0481676 .0816268
|
MP#ownership |
EMP#Private Sector | -1.759344 .0129746 -135.60 0.000 -1.784774 -1.733914
EMP#Public Sector | -.9694264 .0211334 -45.87 0.000 -1.010847 -.9280058
TMP#Foreign Bank | 1.571765 .5003696 3.14 0.002 .5910585 2.552471
TMP#Private Sector | 0 .4642727 0.00 1.000 -.9099577 .9099577
TMP#Public Sector | 0 .3857374 0.00 1.000 -.7560314 .7560314
|
ownership#c.MP_callmoneyrate |
Private Sector | -.1339413 .1145656 -1.17 0.242 -.3584857 .0906032
Public Sector | -.2851568 .2189411 -1.30 0.193 -.7142735 .1439598
|
MP#ownership#c.MP_callmoneyrate |
TMP#Foreign Bank | -.3587273 .1237912 -2.90 0.004 -.6013535 -.1161011
TMP#Private Sector | -.5336093 .2115538 -2.52 0.012 -.9482472 -.1189715
TMP#Public Sector | -.5676607 .3179929 -1.79 0.074 -1.190915 .0555939
|
owmpmr | .144381 .1074397 1.34 0.179 -.066197 .354959

-------------------------------------------------------------------------------------------------
Thanks in advance

[Joao Santos Silva]

Please post the results with the model where the bootstrap did not work.

[Fadi Ansar]

  bootstrap ,cl( bankname) r(100) id(id) seed(123):xtqreg Zscore MP_callmoneyrate Boone_Ind TotalAsset Liquidi > tyRatio CBT FinFreedom NIM RI AssetcompositionRatiooffixe GDP INF Ratioofnonintrestoperating MP##C.MP_callmo > neyrate owmpmr i.Year i.ownership ,quantile (.1 ) id(id) (running xtqreg on estimation sample)

Bootstrap replications (100) ----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5  xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx    50 xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx   100 insufficient observations to compute bootstrap standard errors no results will be saved r(2000);

[Fadi Ansar]

here is the codes and model where Bootsraping doesnt work

Dear Joao Santos Silva .I would like to one more query I encountered while running xtqreg with bootstrap standard errors .
I have used the following commands and got the following results.

bootstrap ,cl( bankname) r(100) id(id) seed(123):xtqreg Zscore MP_callmoneyrate Boone_Ind TotalAsset Liquidi
> tyRatio CBT FinFreedom NIM RI AssetcompositionRatiooffixe GDP INF Ratioofnonintrestoperating MP##C.MP_callmo
> neyrate owmpmr i.Year i.ownership ,quantile (.1 ) i(id)
(running xtqreg on estimation sample)

Bootstrap replications (100)
----+--- 1 ---+--- 2 ---+--- 3 ---+--- 4 ---+--- 5
xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx 50
xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx 100
insufficient observations to compute bootstrap standard errors
no results will be saved
r(2000);

Is this because of data issue or is it because I added year and ownership effects in to the model -the command was working perfectly fine when there is no year dummy and ownership dummy.
Can you please provide your insight on this

Thanks in advance