Fixed effect model output

[Vasilis Zagkreos]

Dear Stata experts,

I would really appreciate it if you could please assist me on the following.

I use panel data for the period of 2006 to 2020. My regression result shows that for a certain year of the variable "Time" has no value and next to that specific year there is a the letter "b" (i.e. refer to the year of 2006 of the attached table below). Could you please help me on what this letter "b" next to 2006 mean? Also, why the regression result has no value for that specific year (i.e. no value for 2006)?

Regression result

Y	Coef.	St.Err.	t-value	p-value	[95% Conf	Interval]
X1	1.086	.221	4.91	0	.65	1.521
X2	-.719	.342	-2.10	.037	-1.393	-.046
2006b	0	.	.	.	.	.
2007	-.03	.026	-1.16	.246	-.081	.021
2008	-.354	.039	-9.01	0	-.432	-.277

.

Moreover, I have expanded my model using more variables. However, my expanded model result shows that the variable "Time" begins from 2009 up to 2020 (in this case I have the same issue as above but for the year of 2009), Considering that my data are from 2006 to 2020, why does my expanded model output under variable "Time" begins from 2009 instead of 2006 (refer to the attached table below)?

Regression result

Y	Coef.	St.Err.	t-value	p-value	[95% Conf	Interval]
X1	1.001	.662	1.51	.137	-.329	2.331
X2	-1.784	.691	-2.58	.013	-3.172	-.397
X3	-.068	.071	-0.96	.341	-.211	.075
X4	-.017	.009	-1.90	.063	-.036	.001
X5	-1.201	1.41	-0.85	.398	-4.033	1.63
X6	-.834	1.027	-0.81	.421	-2.896	1.229
X7	1.582	2.063	0.77	.447	-2.563	5.726
X8	2.552	2.822	0.90	.37	-3.115	8.22
X9	.006	.764	0.01	.994	-1.528	1.54
X10	0	0	0.40	.689	0	0
X11	1	.505	1.98	.053	-.014	2.014
X12	47.565	99.489	0.48	.635	-152.263	247.394
X13	.001	.001	0.69	.491	-.002	.003
X14	.001	.002	0.59	.558	-.002	.004
2009b	0	.	.	.	.	.
2010	.038	.115	0.33	.742	-.193	.27
2011	.093	.143	0.65	.52	-.195	.381
2012	-.022	.077	-0.28	.777	-.177	.133
2013	.329	.118	2.79	.007	.092	.566

Thank you in advance.

Kind regards
Vasilis Zagkreos

[Carlo Lazzaro]

Vasilis:
welcome to this forum.
As per -fvvarlist- notation, the -b.- prefix means -base- or reference category, that Stata automatically omits (-o-) to shelter you from the so called dummy trap (https://en.wikipedia.org/wiki/Dummy_...le_(statistics)).

[Vasilis Zagkreos]

Dear Carlo,

Thank you very much for your swift response.

Regarding my expanded model attached on my previous post, do you know why the time variable begins with 2009 instead of 2006? (Note that my data are from 2006 until 2020)

Thank you.
Vasilis

[Carlo Lazzaro]

Vasilis:
without firther details from your end (that is, posting exactly what you typed and what Stata gave you back, as recommended by the FAQ), my guess-work is that 2006-2008 were omitted due to perfect collinearity with the fixed effect.

[Vasilis Zagkreos]

Carlo,

Please find attached below the command used and the results of Stata as an additional info to get exactly what I did and therefore help you on my post #2.

xtreg Price_BookValue_Ratio_n_w Return_on_Equity_w Price_Volatility_w LnTA_w INT_w NPL_over_TotalLoans_w TRAD_w RealGDPGrowth_w LongInterestRates10Year_w i.Time, re vce(cluster CompanyID)


Regression results

Price_BookValue_Ra~w	Coef.		St.Err.	t-value		p-value	[95% Conf		Interval]		Sig
Return_on_Equity_w	1.273		.49	2.60		.009	.314		2.233		***
Price_Volatility_w	-.587		.315	-1.86		.062	-1.204		.03		*
LnTA_w	.016		.028	0.57		.572	-.038		.07
INT_w	-.016		.006	-2.80		.005	-.028		-.005		***
NPL_over_TotalLoan~w	-.767		.851	-0.90		.368	-2.435		.901
TRAD_w	-.103		.56	-0.18		.854	-1.201		.995
RealGDPGrowth_w	.749		1.363	0.55		.582	-1.921		3.42
LongInterestRates1~w	8.476		1.956	4.33		0	4.642		12.31		***
2009b	0		.	.		.	.		.
2010	.046		.075	0.61		.541	-.101		.192
2011	.014		.107	0.13		.897	-.195		.223
2012	-.069		.054	-1.29		.197	-.174		.036
2013	.203		.078	2.60		.009	.05		.355		***
2014	.327		.081	4.05		0	.168		.485		***
2015	.461		.124	3.73		0	.218		.703		***
2016	.129		.085	1.52		.128	-.037		.294
2017	.18		.095	1.88		.06	-.007		.367		*
2018	.133		.09	1.48		.139	-.043		.309
2019	.075		.092	0.81		.419	-.107		.256
2020	.119		.094	1.27		.204	-.065		.302
Constant	.454		.424	1.07		.284	-.377		1.285
Mean dependent var		0.861			SD dependent var			0.554
Overall r-squared		0.241			Number of obs			847
Chi-square		322.447			Prob > chi2			0.000
R-squared within		0.191			R-squared between			0.235
*** p<.01, ** p<.05, * p<.1

If it has to do with collinearity is there any way to deal with this?


Thank you.
Vasilis

[Carlo Lazzaro]

Vasilis:
thanks for clarifying.
Sorry to appear pedantic, but you should use CODE delimiters to share what you typed and what Stata gave you back (as per FAQ again). Thanks.
That said:
1) if the years left unrepoted were omitted fue to perfect collinearity, Stata should have reported it;
2) another cause might be missing values (less likely, though).