The system dynamics community uses simple ODEs to predict or study what-if behavior of markets. The simulation concept has a solid foundation in statistics since it really is just an experimental technique and thus relies on sampling to make inferences.
How many folks in this group use simulation for predictive methods?