Aggregate and disaggregated fluctuations

In the usual version of the neoclassical growth model used to identify neutral (N-Shock) and investment shocks (I-Shock), a linear transformation frontier between consumption and investment goods is assumed. This paper extends the original framework, allowing for curvature in the transformation frontier, and studies how this affects the relative price of investment goods and hence the identification of investment shocks. A concave frontier allows a substantial improvement in the prediction of the saving rate. Furthermore, a concave frontier induces short-run aggregate effects of relative demand shifts, thereby fostering the propagation of the shocks under consideration, which overall account for 86% of the aggregate fluctuations. When I identify shocks with curvature, the N-shock appears to be stationary while the I-shock is a unit root. This leads the N-shock to play a major role: 91% of the fluctuations explained are due to the N- shock