The bullwhip effect and the Great Trade Collapse

This paper demonstrates the bullwhip effect in a simple framework and tests its predictions using US industry level import data. I show that after final goods suffer a demand shock, upstream suppliers face a greater volatility of sales than their downstream counterparts and might even lose sales temporarily. The effect of the shock is magnified when the inventory-to-sales ratio of the industry is high. The impact can turn non-monotonic, that is, the volatility of a downstream production stage might exceed the volatility of its suppliers, if the upstream producers operate in networks of chains with uncorrelated demands. I show empirically that, in line with the bullwhip effect, the volatility of US imports after the Lehman shock is higher for upstream industries. In addition, upstream products are more likely to drop out of trading completely. Most of the dropouts, however, are temporary and about 90 per cent of products return to trading within two years after the shock. Those imports that do not return are more likely to have been traded for a shorter period and in smaller quantities pre-crisis.