Euro trade effect

Currently I am doing a regression about the effect of the Euro on trade. Now in my first regression I used as a dependent variable export and as independent variable gdp, gdppc, eu and euro. Now my euro estimate was slightly positive with 0.02, but not significant. When I did the same regression with dependent variable trade (import + export), my euro estimate increased to 0.07 and became highly significant. Who has a suggestion why my Euro estimate increased when using trade instead of export?