Seasonal demand forecasting and incentivizing information sharing
研究零售商季节性需求预测下，信息共享对供应商的价值，发现仅在强季节性时才有价值，且供应商可通过调整补货周期来激励共享。
We study the value of information sharing in a two-player supply chain where the retailer observes external demand that is generated by a seasonal process that is either AR(1), SARIMA (0,0,0)X(0,1,0)S or SARIMA (0,0,0)X(1,0,0)S. Similar to previous research, we assume that the retailer constructs its best linear forecast of leadtime demand and uses it to determine the order quantity via a periodic review myopic order-up-to policy, which it places with the supplier. We study whether a given seasonal model of retailer’s demand and a given retailer’s leadtime result in information sharing being valuable to the supplier. We find that in the cases of AR(1) demand and SARIMA (0,0,0)X(1,0,0)S demand, information sharing is valuable only in the presence of strong seasonality in the demand observed by the retailer. In all demand cases studied, we show there is no value to information sharing when the retailer’s forecasting leadtime is a multiple of the seasonal period corresponding to the retailer’s demand process. Since the value of information sharing is dependent on the retailer’s forecasting leadtime, we show that the supplier may have an incentive to adjust its replenishment leadtime to the retailer. Additionally, in the case of SARIMA (0,0,0)X(0,1,0)S demand, when the retailer’s forecasting leadtime is not a multiple of the seasonal period S, we show that the retailer’s current order can be well approximated by a linear combination of the retailer’s past demand (leading to positive value of information sharing).