Moving Average (MA)

[tomascanosa]

Hi, I'm rather new in econometrics, but I'm trying to take the risk and include some econometrics in my thesis. I´m calculating long term exports and import elasticities. Thus, the log of exports and imports depends on the log of real exchange rate(for both equations), log of local growth (for the elasticity of imports) and log of growth rates of main trade partners (for elasticity of exports). Although the R2 and the values of the seem rather good, I'm having serial autocorrelation. I´m using quarterly data from 1993 to 2013 and I cannot get more data. I could solve the problem including MA(1) to MA(4) to solve it, and only in this case I'm getting good results. Do you think that´s a good idea?
Kind regards,
Tomas