Need help with my module

[Kosebaja]

Hello

Im currently doing some study on the real estate market (observation are from 1985-2011), but I have a big problem getting significant variables. Even thou i follow economic theory when choosing the variables.

Square meter prices = c + population growth + housing stock growth + lending rate + enemployment rate

My module are not stationary in level, but its stationary in first difference.

When I test for cointegration I get : "Trace test indicates 4 cointegrating eqn(s) at the 0.05 level" and the long run equation I get dont match economic theory.

After the cointegration test I use a VECM model, with 2 lags, but get that all my variables are not significant. Except the "c".

Can anyone here give me some guidelines on what I should do different. Been working on these numbers for some weeks now and at the moment I ran out if ideas.

Thanks