An Empirical Analysis of Forecast Sharing in the Semiconductor Equipment Supply Chain
基于半导体设备供应链的大规模数据，研究买方预测行为（修订频率、幅度和虚报比例）如何影响供应商的交货表现，发现双方会因对方不可靠行为而互相惩罚。
We study the demand forecast-sharing process between a buyer of customized production equipment and a set of equipment suppliers. Based on a large data collection we undertook in the semiconductor equipment supply chain, we empirically investigate the relationship between the buyer's forecasting behavior and the supplier's delivery performance. The buyer's forecasting behavior is characterized by the frequency and magnitude of forecast revisions it requests (forecast volatility) as well as by the fraction of orders that were forecasted but never actually purchased (forecast inflation). The supplier's delivery performance is measured by its ability to meet delivery dates requested by the customers. Based on a duration analysis, we are able to show that suppliers penalize buyers for unreliable forecasts by providing lower service levels. Vice versa, we also show that buyers penalize suppliers that have a history of poor service by providing them with overly inflated forecasts.