Repeated Transition Method and the Nonlinear Business Cycle with the Corporate Saving Glut

This paper develops a novel methodology to globally solve nonlinear dynamic stochastic general equilibrium models with high accuracy. The algorithm is based on the ergodic theorem: if a simulated path of the aggregate shock is long enough, all the possible equilibrium allocations are realized, enabling a complete characterization of the rationally expected future outcomes at each point on the path. The algorithm is applied to a heterogeneous-firm business cycle model where firms hoard cash as a buffer stock. Using the model, I analyze the state-dependent shock sensitivity of consumption over corporate cash stocks and provide empirical evidence.