I am new to Stata and this is my first post.
I want to estimate a translog cost function and its partial derivative (marginal cost). I need the estimate of marginal cost for every year for every bank to estimate my Lerner Index. I use panel data for over 2000 banks over 8 periods (ranging from 2011 to 2018).
I am not sure how to estimate the things described above, cause Stata ends the regression of my translog cost function due to "insufficient observations". Thats possibly because I try to estimate a single line of my data - One bank at a specific year. I do not know how to proceed... But I need the marginal costs for each bank in each year. For any advice I will be more than glad. I've read a paper, where they use OLS, so I want to do the same.
Here is a very short overview of my data
Code:
* Example generated by -dataex-. To install: ssc install dataex clear input byte id_bank int year double(tc q1) float(q2 w1 w2 w3) 1 2011 55000 2910016 3064633 .019 .017 .00099 1 2012 48000 3291208 3417275 .014 .018 .0012 1 2013 96000 4693980 4880171 .013 .016 .0014 1 2014 91000 3866875 3808000 .0098 .022 .0017 1 2015 266800 7114940 8787466 .0067 .019 .0078 1 2016 242500 6990195 8525062 .004 .016 .0058 1 2017 290000 7905072 10314923 .0036 .019 .003 1 2018 286000 8562460.9 10845030 .0037 .017 .0025 2 2011 . . . . . . 2 2012 . . . . . . 2 2013 261510.46 12342217 13048772 .014 .013 .0021 2 2014 202784 10891645 11083400 .011 .014 .00077 2 2015 207959.1 9841932 9840160 .0066 .015 .0011 2 2016 201420 9770153 9562544 .0048 .014 .0013 2 2017 246413 11439694 10876763 .0037 .017 .002 2 2018 243999 11253174 11242116 .0068 .018 .0012 3 2011 119539 8882993 7307546 .04 .009 .00043 3 2012 126488 9735650 8199895 .028 .0087 .00046 3 2013 143777 10110711 8207365 .029 .009 .00044 3 2014 143369 8192736 7233554 .027 .013 .00025 3 2015 125294 7752964 6986139 .02 .01 .0006 3 2016 116501 8372932 7324262 .019 .0092 .00054 3 2017 133004 9060156 8461524 .017 .0095 .00048 3 2018 141117 9152467 8232791 .0057 .0097 .00074 end
Code:
tsset id_bank year, yearly
Code:
*\\ TAKING THE LOGS OF VARIABLES foreach var in tc q1 q2 w1 w2 w3 { gen double ln_`var'= ln(`var') }
Code:
*\\ MULTIPLY SQUARED TERMS BY 0.5 foreach var in q1 q2 w1 w2 w3 { gen double ln_`var'2 = 0.5*(ln_`var')^2 }
Code:
forvalues y = 2011(1) 2018 { forvalues i = 1(1) 3 { display `y' display `i' capture reg ln_tc ln_q1 ln_q2 ln_q12 ln_q22 c.ln_q1#c.ln_q2 ln_w1 ln_w2 ln_w3 ln_w12 c.ln_w1#c.ln_w2 c.ln_w1#c.ln_w3 ln_w22 c.ln_w2#c.ln_w3 ln_w32 c.ln_q2#c.ln_w1 c.ln_q2#c.ln_w2 c.ln_q2#c.ln_w3 c.ln_q1#c.ln_w1 c.ln_q1#c.ln_w2 c.ln_q1#c.ln_w3 if `y' == year& `i'== id_bank } }
Thanks for any help.
0 Response to Problems with Estimation of Translog Cost Function and its derivative - Error Insufficient Observation
Post a Comment