Dear Statalists

I'm having a bit of trouble interpreting the following output:

Code:
------------------------------------------------------------------------------
     CAR_win |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
         UAI |    .000398   .0002983     1.33   0.184    -.0001909     .000987
      LnSize |   -.004797   .0023237    -2.06   0.041    -.0093846   -.0002094
    Leverage |  -.0165098   .0201235    -0.82   0.413     -.056239    .0232195
 LnDealValue |   .0049224   .0024342     2.02   0.045     .0001167    .0097282
    Domestic |  -.0060647   .0094017    -0.65   0.520    -.0246263    .0124969
  Horizontal |  -.0071151   .0072168    -0.99   0.326     -.021363    .0071329
       Japan |  -.0226577   .0123482    -1.83   0.068    -.0470364    .0017211
          US |  -.0255653   .0130358    -1.96   0.052    -.0513015    .0001709
       _cons |   .0483734   .0301197     1.61   0.110    -.0110911    .1078378
------------------------------------------------------------------------------
CAR = a ratio which serves as a %. I know that if I interpret the Logarithm of Size, I can say that by an increase of 1% in Size, the CAR decreases in ...%.
However I'm not sure about the coëfficiënt. Do I say that the CAR decreases by -0,004797% or by 0,4797%?

Same with Japan & US, they are dummies. So if their value is equal to 1, is there a decrease of 22% or of 0,22%?

Kind regards! Just a rookie question.