The 20th International Symposium on Inventories organised by the International Society for Inventory Research (ISIR) was held at Budapest. Three members of the Centre for Marketing Analytics and Forecasting contributed by presenting their latest research results:
John Boylan presented a co-work with Mohamed Zied Babai on “estimating discrete cumulative distributions by resampling”. Since slow-moving demand is highly discrete, it does not always conform to standard probability distributions which are required for stock-control purposes. In their work, they analyse some of the properties of a popular non-parametric resampling method with replacement as an estimator of the cumulative distribution function. They show that those estimators are biased, for lead-times of two periods or more. This finding allows identifying the conditions under which the maximum bias is attained and to compare the resulting variances in cumulative distributions of resampling methods with and without replacement. Corrections for bias are available in certain cases.
Nikolaos Kourentzes elaborated with Centre affiliates Devon Barrow and Juan Trapero new ways to “optimise forecasting models for inventory planning”. In the past, the forecasting literature often focused on providing optimal models regarding the likelihood and various accuracy metrics. However, this does not always lead to better inventory performance, as often the translation between forecast errors and inventory results is not linear. In their study Nikos and his co-authors propose a way to linearise the competing multiple inventory objectives, i.e. meeting demand, while eliminating excessive stock. They then evaluate the proposed parametrisation against established alternatives to demonstrate its performance and explore the connection between forecast accuracy and inventory performance.
Patrick Saoud and his supervisors Nikolaos Kourentzes and John Boylan investigate on the topic of “using demand uncertainty as a determinant for the bullwhip effect”. One of the key issues is to correctly quantify the effect itself. Currently, the metric adopted by most of the papers measures the propagation of the variability of demand. The research of Patrick, however, argues that demand uncertainty, rather than variability, should be gauged as a determinant of the Bullwhip since poor demand forecasting is one of the original causes of the phenomenon. His presentation highlights the limitations of the current metric and inspects different scenarios where his proposed metric is compared regarding cost performance and managerial actions to manage the bullwhip effect.
Furthermore, John Boylan began his term of office as President of ISIR. Speaking at the banquet held at the Gundel Restaurant, John thanked the members of the society for placing their trust in him as their president. Previous incumbents have included such well-known names in inventory research as Edward Silver and Sven Axsäter. John explained (with tongue very firmly in cheek) that he would model his presidency on Donald Trump, spewing out at least ten tweets a day, resisting any attempts at impeachment, and ‘building a wall’ to prevent any new members joining ISIR! Reverting to more serious comments, John thanked all those who had been involved in organising an excellent symposium, and particularly Attila Chikán, ISIR’s founder and Executive Vice-President.
If you require accessible versions of the slides or documents within this news story, please contact us.