One of the most important conferences, possibly the most important, about forecasting is The International Symposium on Forecasting. This year was celebrated in Tesalonika (Greece) and I had the opportunity to share with the rest of colleagues our latest advances in supply chain forecasting. This year, we presented different approaches (parametric and non-parametric) to better estimate the safety stock. Recall that, the safety stock may imply a significant amount of money for many companies. Among the non-parametric approaches, we employed Kernel Density Estimators and the parametric counterpart was modelled with GARCH models. Both of them work well under several circumnstances, although GARCH models outperform the rest, when the lead time was large. The article with more details is available in:
- Juan R. Trapero, Manuel Cardós, Nikolaos Kourentzes, (2019) “Empirical safety stock estimation based on Kernel and GARCH models”, OMEGA, The International Journal of Management , Volume 84, pp. 199-211.
