I have a problem regarding a time-series regression for several industries. I am using stata13.
My data looks as follows:
Code:
clear input float(Date mktrf smb hml rf agric mines oil) -402 2.96 -2.3 -2.87 .22 2.1499999 2.42 -2.91 -401 2.64 -1.4 4.19 .25 1.98 .42 3.53 -400 .36 -1.32 .01 .23 -.8 -.45 -6.15 -399 -3.24 .04 .51 .32 -.78 -.66 -2.3 -398 2.53 -.2 -.35 .31 6.44 1.27 .6 -397 2.62 -.04 -.02 .28 -3.55 2 3.17 -396 -.06 -.56 4.83 .25 -3.91 -1.41 5.91 -395 4.18 -.1 3.17 .26 7.39 4.62 8.96 -394 .13 -1.6 -2.67 .3 -.9 3.21 -13.65 -393 .46 .43 .6 .25 5.14 1.06 -2.74 -392 5.44 1.41 4.93 .3 3.82 1.1099999 11.14 -391 -2.34 .47 -1.53 .26 -3.66 -5.46 -1.13 -390 7.26 -3.23 -1.16 .3 5.75 5.84 2.0900002 -389 1.97 -.72 -3.69 .28 1.71 1.62 -11.75 -388 4.76 -3.57 -.71 .21 7.12 2.63 11.75 -387 -4.31 2.13 -4.33 .25 -6.05 -4.78 -7.51 -386 6.58 2.76 -.31 .21 6.3 8.61 10.17 -385 2.09 .93 -1.06 .22 -1.04 6.71 3.93 -384 -.68 4.25 -.72 .25 -2.1399999 -2.65 -.17 -383 -1.7 -2.03 -.69 .33 .1 -4.2400002 -3.58 -382 8.81 -.26 -1.2 .29 1.42 5.86 6.34 -381 4.23 3.82 3.67 .22 -.5 .9 11.94 -380 1.52 2.98 -3.46 .32 -1.0999999 2.44 -.21 -379 -4.85 -3.5 -.06 .31 -2.44 -4.92 -8.18 -378 .62 -1.35 -.47 .32 -.61 -1.87 -.72 -377 6.68 -2.07 -2.11 .32 1.36 5.62 1.77 -376 2.88 2.18 .76 .27 -1.17 2.97 8.24 -375 1.33 2.27 -2.26 .41 4.32 3.34 -1.3 -374 11.81 -1.81 2.8 .38 1.26 12.89 10.54 -373 .36 -.85 -.6 .06 -1.34 .65 -14.78 end format %tm Date
1. I want to do a time-series regression with a lag of 12 months for every industry (agric, mines, oil, etc.) and every month.
2. I want to summarize the coefficients and save them to make further calculations.
At the moment i am just able to do the regression for every industry once with a lag period of 12-month.
The code looks like this:
Code:
foreach x of varlist agric-oil { regress L[(1/60).`x' L(1/60).mktrf L(1/60).smb L(1/60).hml }
Code:
Source | SS df MS Number of obs = 1050 -------------+------------------------------ F(180, 869) = 0.55 Model | 28565.7057 180 158.698365 Prob > F = 1.0000 Residual | 252514.675 869 290.580754 R-squared = 0.1016 -------------+------------------------------ Adj R-squared = -0.0845 Total | 281080.381 1049 267.950792 Root MSE = 17.046 ------------------------------------------------------------------------------ agric | Coef. Std. Err. t P>|t| [95% Conf. Interval] -------------+---------------------------------------------------------------- mktrf | L1. | .0844818 .1218709 0.69 0.488 -.154714 .3236775 L2. | -.0579726 .1250075 -0.46 0.643 -.3033246 .1873794 L3. | -.0551239 .1249082 -0.44 0.659 -.3002809 .190033 L4. | .0614014 .1247025 0.49 0.623 -.1833518 .3061546 L5. | .1953326 .1252396 1.56 0.119 -.0504748 .4411401 L6. | -.062288 .1272491 -0.49 0.625 -.3120395 .1874635 L7. | -.1434886 .1274496 -1.13 0.261 -.3936337 .1066565 L8. | -.0042542 .1274069 -0.03 0.973 -.2543154 .2458071 L9. | .0283648 .127659 0.22 0.824 -.2221912 .2789208
The final output should look somehow like this:
Code:
input float(Date agric_const agric_mktrf agric_smb agric_hml mines_const mines_mktrf mines_smb mines_hml) -376 0.84 0.93 2.45 3.10 0.33 0.42 0.82 0.96 -375 0.92 0.91 2.30 1.72 0.38 0.40 0.80 0.90
I hope the problem gets clear.
Thank you very much in advance.
Kind regards,
Andreas
0 Response to Store results for regression coefficients for a time-series regression over multiple industries
Post a Comment