Many CGE models analyzing the cost of climate policies are recursive dynamic in order to take into account much sectoral and regional detail. Introducing forward looking behavior might be preferable from a theoretical perspective because it allows agents to endogenously adjust intertemporal investment and trade decisions, however, this comes at computational cost. Here we present three versions of the same model (forward looking, recursive dynamic, and an intermediate case which only restricts intertemporal trade decisions) and compare outcomes for exemplary climate policies. Limiting intertemporal adjustments makes climate policy more costly. Preliminary results suggest that the intertemporal adjustments via trade are more important than adjustments via investment decisions if the baseline starts from a first best capital path, but that might depend on the type of policy to be implemented.