From newsstands to airlines: new applications for the newsvendor model
Author: Ideas with Impact
Posted on Dec 12, 2017
Category: Faculty
To learn about managing inventories, business students are introduced to the “newsvendor model” in their operations management courses. The newsvendor model is a mathematical model used to manage inventories in as lean and efficient a manner as possible. While the model has been around for a long time, Dr. Donglei Du, professor of quantitative methods at the University of New Brunswick’s faculty of business administration, has been exploring new ways it can be used in different industries to make better business decisions.
First introduced by the economist Francis Ysidro Edgeworth (1845-1926), the newsvendor model takes its name from the challenge newsvendors face each morning when they decide how many newspapers they will need for the day. If they buy too many, they incur overage cost; if they buy too few, they incur underage cost. To be successful, they have to find the sweet spot where they can purchase the smallest amount possible that will at the same time fulfill all the demands they will have; and the level of demand is almost always unknown at the time of ordering.
Du and his colleagues have been studying the properties of various newsvendor models and analyzing historical data to discover how these models can be applied in different areas and for different purposes. One area where the model could be used to make better business decisions Du explained is in the airline industry. “Airlines overbook their flights all the time knowing that a certain percentage of travelers will cancel their reservations at the last minute or just not show up. By overbooking seats, airlines improve their revenue . Of course, if they miscalculate the number of seats they oversell, they end up in a situation where they have to bump customers, and this is not good for business. Therefore, we have an obvious balancing decision here and newsvendor models are perfect for this type of situation.”
Using the newsvendor model to figure out by how much they can overbook their seats, would be good for airlines’ profits and for their customers. Du explains further that newsvendor models could be used similarly to help car rental agencies and hotels manage their bookings more efficiently, and even to help university admissions offices in managing their programs admissions.
“While the model has been around for a long time, the newsvendor model pops up like mushrooms after the rain in so many applications,” says Du. “The beauty of doing theoretical work like this project is that you find unexpected applications.”
The results of Du’s research have been published in 2014 in Operations Research, a journal included in the Financial Times Top 50 Journals, which is used to rank the research output of business school around the world. “A risk- and ambiguity-averse extension of the max-min newsvendor order formula”, was co-written with Qiaming Han and Luis Zulauaga.
Du and his coauthors extend the classical newsvendor problem to include both risk and ambiguity factors to overcome the limitations of the classical newsvendor problem and hence propose a new model to be more suitable for practical applications. Du introduces the newsvendor model to students in his operations management classes to help them appreciate how it can be used to manage inventories efficiently. Next year he will be teaching them how to use it in the alternate financial sector with options trading, financial derivatives and market-making in a new course he will be teaching on algorithmic trading.
For more information, contact Liz Lemon-Mitchell with UNB’s faculty of business administration.
Learn more about our business programs.