Hello everyone, I am currently doing my thesis about the determinants of corporate cash holding in a specific country. I will study the effect of firms specific factors and macroeconomic factors on the dependant variable ( cash ). My panel is unbalanced and the VIF test gave me a fine result for as you can see:
------------------------------------------
Variable VIF 1/VIF

bd 5.19 0.192749
inflation 3.11 0.321346
oilpricech~e 2.24 0.446424
stockmarke~e 1.87 0.535713
size_w 1.76 0.567209
sibor3months 1.73 0.576907
lev_w 1.63 0.614210
cf_w 1.51 0.662620
div 1.46 0.685949
mbv_w 1.26 0.793562
nwc_w 1.24 0.805450
capexve_w 1.14 0.880446

Mean VIF 2.01

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





After that I found fixed effect model is better than Pooled OLS, and then I did Hausman test and fixed effect model is preferred. So, I run Fixed Effect Model for the relationship with each macroeconomic variable and firms specific factors as control factors. Unfortunately, I suffer from omitted variable ( year 2018 is omitted because of collinearity) as you can see when I add the macro factors. the tables below show examples of the main problem.
note: when I run fixed effect model for firms specific factors only, I have no problems.





table 1: adding (inflation of the country) caused year omitted
-----------------------------------------------
xtreg LogCash lev_w mbv_w nwc_w size_w capexve_w div cf_w inflation i.year, fe
note: 2018.year omitted because of collinearity

Fixed-effects (within) regression Number of obs = 1,119
Group variable: company Number of groups = 108

R-sq: Obs per group:
within = 0.1634 min = 4
between = 0.1064 avg = 10.4
overall = 0.1009 max = 11

F(17,994) = 11.42
corr(u_i, Xb) = -0.4653 Prob > F = 0.0000


LogCash Coef. Std. Err. t P>t [95% Conf. Interval]

lev_w -2.310146 .2577364 -8.96 0.000 -2.815916 -1.804376
mbv_w -.0193688 .0321984 -0.60 0.548 -.0825535 .0438159
nwc_w -1.664048 .2695891 -6.17 0.000 -2.193077 -1.135019
size_w .4467487 .0978057 4.57 0.000 .2548192 .6386781
capexve_w .2138972 .3204357 0.67 0.505 -.4149108 .8427052
div .0870217 .0639039 1.36 0.174 -.0383804 .2124238
cf_w 1.711467 .4643174 3.69 0.000 .800312 2.622622
inflation 16.28229 3.051657 5.34 0.000 10.29386 22.27072

year
2009 .3553394 .0879908 4.04 0.000 .1826704 .5280084
2010 .3009735 .0872354 3.45 0.001 .1297868 .4721602
2011 .241402 .0863266 2.80 0.005 .0719987 .4108054
2012 .3650693 .092904 3.93 0.000 .1827588 .5473797
2013 .1822877 .0864988 2.11 0.035 .0125464 .352029
2014 .3792166 .1011391 3.75 0.000 .1807459 .5776873
2015 .4767654 .1170848 4.07 0.000 .2470037 .7065271
2016 .2970261 .0994 2.99 0.003 .1019682 .4920839
2017 .6343364 .164206 3.86 0.000 .3121062 .9565666
2018 0 (omitted)

_cons -6.570472 .7631556 -8.61 0.000 -8.068053 -5.072891

sigma_u .88489799
sigma_e .66656809
rho .63799217 (fraction of variance due to u_i)

F test that all u_i=0: F(107, 994) = 11.93 Prob > F = 0.0000


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





table 2: adding all the macro variables caused more omitted variables:

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

xtreg LogCash lev_w mbv_w nwc_w size_w capexve_w div cf_w inflation oilpricechange stockmarketprice sibor3months bd i.year, fe
note: 2014.year omitted because of collinearity
note: 2015.year omitted because of collinearity
note: 2016.year omitted because of collinearity
note: 2017.year omitted because of collinearity
note: 2018.year omitted because of collinearity

Fixed-effects (within) regression Number of obs = 1,119
Group variable: company Number of groups = 108

R-sq: Obs per group:
within = 0.1634 min = 4
between = 0.1064 avg = 10.4
overall = 0.1009 max = 11

F(17,994) = 11.42
corr(u_i, Xb) = -0.4653 Prob > F = 0.0000


LogCash Coef. Std. Err. t P>t [95% Conf. Interval]

lev_w -2.310146 .2577364 -8.96 0.000 -2.815916 -1.804376
mbv_w -.0193688 .0321984 -0.60 0.548 -.0825535 .0438159
nwc_w -1.664048 .2695891 -6.17 0.000 -2.193077 -1.135019
size_w .4467487 .0978057 4.57 0.000 .2548192 .6386781
capexve_w .2138972 .3204357 0.67 0.505 -.4149108 .8427052
div .0870217 .0639039 1.36 0.174 -.0383804 .2124238
cf_w 1.711467 .4643174 3.69 0.000 .800312 2.622622
inflation -6.190001 4.811288 -1.29 0.199 -15.63145 3.251446
oilpricechange -1.049068 .4558061 -2.30 0.022 -1.94352 -.154615
stockmarketprice .5358696 .5211199 1.03 0.304 -.4867518 1.558491
sibor3months 7.064984 9.291642 0.76 0.447 -11.1685 25.29847
bd 3.140616 1.335088 2.35 0.019 .5207012 5.760531

year
2009 .0664926 .2185342 0.30 0.761 -.3623487 .4953339
2010 .3630901 .173711 2.09 0.037 .0222078 .7039725
2011 .2763219 .1912937 1.44 0.149 -.099064 .6517078
2012 -.2925965 .2286194 -1.28 0.201 -.7412285 .1560356
2013 -.2622535 .2363012 -1.11 0.267 -.7259599 .201453
2014 0 (omitted)
2015 0 (omitted)
2016 0 (omitted)
2017 0 (omitted)
2018 0 (omitted)

_cons -5.745039 .7408075 -7.76 0.000 -7.198765 -4.291313

sigma_u .88489799
sigma_e .66656809
rho .63799217 (fraction of variance due to u_i)

F test that all u_i=0: F(107, 994) = 11.93 Prob > F = 0.0000

.
----------------------------------------------------------



The final qustion is when I use ' eststo ' command, the ommited variables become the macroeconomic factors ( see table 3 below ), Can any one explain the problem for me please? Thank you very much for your help.

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

eststo: xtreg LogCash lev_w mbv_w nwc_w size_w capexve_w div cf_w inflation oilpricechange stockmarketprice si
> bor3months bd i.year, fe
note: inflation omitted because of collinearity
note: oilpricechange omitted because of collinearity
note: stockmarketprice omitted because of collinearity
note: sibor3months omitted because of collinearity
note: bd omitted because of collinearity

Fixed-effects (within) regression Number of obs = 1119
Group variable: company Number of groups = 108

R-sq: within = 0.1634 Obs per group: min = 4
between = 0.1064 avg = 10.4
overall = 0.1009 max = 11

F(17,994) = 11.42
corr(u_i, Xb) = -0.4653 Prob > F = 0.0000

----------------------------------------------------------------------------------
LogCash | Coef. Std. Err. t P>|t| [95% Conf. Interval]
-----------------+----------------------------------------------------------------
lev_w | -2.310146 .2577364 -8.96 0.000 -2.815916 -1.804376
mbv_w | -.0193688 .0321984 -0.60 0.548 -.0825535 .0438159
nwc_w | -1.664048 .2695891 -6.17 0.000 -2.193077 -1.135019
size_w | .4467487 .0978057 4.57 0.000 .2548192 .6386781
capexve_w | .2138972 .3204357 0.67 0.505 -.4149108 .8427052
div | .0870217 .0639039 1.36 0.174 -.0383804 .2124238
cf_w | 1.711467 .4643174 3.69 0.000 .800312 2.622622
inflation | 0 (omitted)
oilpricechange | 0 (omitted)
stockmarketprice | 0 (omitted)
sibor3months | 0 (omitted)
bd | 0 (omitted)
|
year |
2009 | .0456503 .1028826 0.44 0.657 -.1562417 .2475424
2010 | -.0735191 .1018148 -0.72 0.470 -.2733156 .1262775
2011 | -.1374868 .1026184 -1.34 0.181 -.3388603 .0638867
2012 | -.1574293 .1035933 -1.52 0.129 -.3607159 .0458573
2013 | -.2351901 .1058544 -2.22 0.027 -.4429139 -.0274663
2014 | -.2531874 .107252 -2.36 0.018 -.4636537 -.0427212
2015 | -.3067382 .1063209 -2.89 0.004 -.5153774 -.0980991
2016 | -.3632207 .1069679 -3.40 0.001 -.5731295 -.1533118
2017 | -.495003 .1077467 -4.59 0.000 -.7064401 -.2835659
2018 | -.5863251 .1098902 -5.34 0.000 -.8019685 -.3706818
|
_cons | -5.579858 .7012164 -7.96 0.000 -6.955892 -4.203823
-----------------+----------------------------------------------------------------
sigma_u | .88489799
sigma_e | .66656809
rho | .63799217 (fraction of variance due to u_i)
----------------------------------------------------------------------------------
F test that all u_i=0: F(107, 994) = 11.93 Prob > F = 0.0000
(est1 stored)


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