I have a database in quarterly form containing quantities (stdunits), price and recalls status of a product whose id is idproduct:
Code:
* Example generated by -dataex-. To install: ssc install dataex clear input float(idproduct quarter) double stdunits float price str33 atc4 float recalls 201 216 174300 2.1964164 "A11A1" 0 201 217 451080 2.2246852 "A11A1" 0 201 218 285600 2.450323 "A11A1" 0 201 219 284340 2.382636 "A11A1" 0 201 220 307860 2.447608 "A11A1" 0 201 221 334320 2.672332 "A11A1" 0 201 222 475020 2.7033916 "A11A1" 0 201 223 452760 2.6765375 "A11A1" 0 260 197 916800 .28061396 "A11A1" 0 260 198 2044800 .29431167 "A11A1" 0 260 199 2313600 .2754174 "A11A1" 0 260 200 1993200 .27755743 "A11A1" 0 260 201 1194000 .28332245 "A11A1" 0 260 202 1599600 .2791442 "A11A1" 0 260 203 1368000 .27613425 "A11A1" 0 260 204 1323600 .27033246 "A11A1" 0 260 205 1236000 .27094033 "A11A1" 0 260 206 1190400 .27203014 "A11A1" 0 260 207 1033200 .27012616 "A11A1" 0 260 208 1125600 .2658686 "A11A1" 0 260 209 943200 .252326 "A11A1" 0 260 210 903600 .24691777 "A11A1" 0 260 211 770400 .25007698 "A11A1" 0 260 212 914400 .2402639 "A11A1" 0 260 213 812400 .25345165 "A11A1" 0 260 214 813600 .2501289 "A11A1" 0 260 215 736800 .23505853 "A11A1" 0 260 216 516000 .23366822 "A11A1" 0 260 217 232800 .2281709 "A11A1" 0 260 218 423600 .2368929 "A11A1" 0 260 219 476400 .2225063 "A11A1" 0 260 220 330000 .22424436 "A11A1" 0 260 221 308400 .22906102 "A11A1" 0 260 222 208800 .2355502 "A11A1" 0 260 223 30000 .2216218 "A11A1" 0 262 196 344640 .804536 "A11A1" 0 262 197 4250400 .721491 "A11A1" 0 262 198 1444800 .6426724 "A11A1" 0 262 199 900480 .6522723 "A11A1" 0 262 200 660480 .6813239 "A11A1" 0 262 201 238080 .772738 "A11A1" 0 262 202 118080 .7668843 "A11A1" 0 262 203 21120 .8730164 "A11A1" 0 264 176 2647980 .19712014 "A11A1" 0 264 177 2199870 .204385 "A11A1" 0 264 178 2317680 .19413882 "A11A1" 0 264 179 1916100 .17797776 "A11A1" 0 264 180 1808550 .17485006 "A11A1" 0 264 181 1683900 .1725303 "A11A1" 0 264 182 1722780 .1724692 "A11A1" 0 264 183 1790010 .15955125 "A11A1" 0 264 184 1828350 .1580459 "A11A1" 0 264 185 1678050 .1522747 "A11A1" 0 264 186 1589400 .1528636 "A11A1" 0 264 187 1528920 .1497916 "A11A1" 0 264 188 1697400 .15005007 "A11A1" 0 264 189 1523430 .14358002 "A11A1" 0 264 190 1638000 .13797528 "A11A1" 0 264 191 1427580 .1390256 "A11A1" 0 264 192 1557630 .1336188 "A11A1" 0 264 193 1578600 .12870497 "A11A1" 0 264 194 1682280 .12747423 "A11A1" 0 264 195 93420 .13353658 "A11A1" 0 264 196 2700 .2186262 "A11A1" 0 264 197 90 .11101135 "A11A1" 0 264 198 2520 .1048232 "A11A1" 0 264 199 270 .10111975 "A11A1" 0 265 176 1381500 .26520652 "A11A1" 0 265 177 1766880 .27631393 "A11A1" 0 265 178 1565640 .26443782 "A11A1" 0 265 179 1423350 .2506607 "A11A1" 0 265 180 1399140 .25308436 "A11A1" 0 265 181 1260720 .2490079 "A11A1" 0 265 182 1199250 .25277126 "A11A1" 0 265 183 1124640 .2348305 "A11A1" 0 265 184 981630 .23415777 "A11A1" 0 265 185 893610 .2343847 "A11A1" 0 265 186 787590 .2215974 "A11A1" 0 265 187 733140 .2230178 "A11A1" 0 265 188 763470 .22647485 "A11A1" 0 265 189 739080 .213277 "A11A1" 0 265 190 673290 .2160663 "A11A1" 0 265 191 600210 .2170518 "A11A1" 0 265 192 582840 .21356922 "A11A1" 0 265 193 547470 .21118917 "A11A1" 0 265 194 561870 .2110842 "A11A1" 0 265 195 953910 .2011891 "A11A1" 0 265 196 11160 .23798816 "A11A1" 0 265 197 2520 .13832366 "A11A1" 0 265 198 450 .1638736 "A11A1" 0 265 199 360 .1668476 "A11A1" 0 265 200 90 .16961233 "A11A1" 0 851 176 131100 .25233778 "A11A1" 0 851 177 123900 .25428304 "A11A1" 0 851 178 107200 .25201392 "A11A1" 0 851 179 80800 .2484109 "A11A1" 0 851 180 58000 .24146485 "A11A1" 0 851 181 55600 .2420958 "A11A1" 0 851 182 25100 .2468457 "A11A1" 0 851 183 5100 .22528493 "A11A1" 0 end format %tq quarter
The result, ideally, should be a square matrix (18 x 18) of coefficients representing the own elasticities on the main diagonal and the cross elasticities outside diagonal. I don't know if it is possible. However this is equivalent to estimate the following system of equations:
q_1 = b1*p_1 + b2*p_2+b3*p_3+...+bN*p_N
q_2 = b2*p_2 + b1*p_1+b3*p_3+...+bN*p_N
q_3 = b3*p_3 + b1*p_1+b2*p_2+b4*p_4+...+bN*p_N
....
q_N = bN*p_N + b1*p_1+b2*p_2+...+bN-1*p_N-1
This is a system of equations for panel (hence my difficulty in finding a correct method. I do not think that SEM or SUR are available for panel and in case I don't know how I can implement on panels such methodologies). Is there a way to estimate own and cross elasticities for panel?
Thank you,
Federico
0 Response to Cross Elasticities
Post a Comment