Re: Some economics but mostly maths, please help!
A first guess would be that your risk premium is related to your independent variables as a ratio some thing like this;
risk premium = parameter * trading days / market size
so the bigger the marker, the lower the risk premium
and as trading days decreases, the risk premium decrease
so for city K, the initial condition is
0.02 = parameter * 40 / 20
so your parameter is valued 0.01, and therefore for city B you have this;
city B = 0.01 * (0.25*40)/(2*20) = 0.0025
so the risk premium is .25% using that model
Obviously there could be some much more complicated relationship between those variates, and you would need more information to either suggest a model or test the fitted model.