On the mathematics and economic assumptions of continuous-time models