An Empirical Signal Processing Model of the Supply Chain

We present a structural econometric framework of supply chain information flows. The framework analyzes supply chain black boxes by measuring how input demand signals transform into output order signals. It characterizes inventory policies with impulse response functions (IRFs). We develop an algorithm to estimate these IRFs, and the structural parameters underlying them, with single firm sales and production data. We demonstrate our methodology with two studies of the automotive manufacturing sector. First, we decompose the automobile bullwhip effect by driver. Second, we measure the degree to which sharing information and improving demand forecasts reduces supply chain uncertainty.