[Gretl-users] gravity model of trade, negative coefficient of GDP

Hey there, I’m now writing a paper about trade relations by means of the gravity model of trade. In my model I decided to include the fixed country-pair effects. However, after having input those country-pair effects, the coefficient of the GDP of importer is negative, whilst the theory itself says increase of any partner’s GDP (exporter or importer) shall have a positive influence on the trade volume. If i exclude the fixed effects, the coefficients are positive as expected. Could anyone give me an explanation to that? Best regards Damian