Intelligent Algorithms to SupportBuyer Intuition


CALL: 2019

DOMAIN: LE - Law, Economics, Finance


LAST NAME: Löhndorf


INDUSTRY / PPP PARTNER: ArcelorMittal Sourcing

HOST INSTITUTION: University of Luxembourg

KEYWORDS: raw material procurement; strategic sourcing; market price volatility; decision-making under uncertainty; machine learning; generative models; stochastic-dynamic optimization; supplier gaming

START: 2019-10-01

END: 2022-09-30


Submitted Abstract

ArcelorMittal Sourcing purchases 5 million tons of special quality coking coals per year with a market value of about $650 million. Purchases are based on long-term contracts with a small number of suppliers, where annual volumes must move in a fixed band, but prices are negotiated on a quarterly basis. To ensure the security of supplies, volumes can be allocated between suppliers, but the total volume must meet the total demand of the steel mills. Before the beginning of a quarter, the buyer must decide about the allocation of volumes to suppliers and to months and decide whether to accept a fixed price from a supplier or whether to peg supplies from that supplier to the market price.To secure its competitiveness, the long-term objective of the company is to negotiate fixed prices below the future market price. As market prices are uncertain when a decision is made, the buyer faces the risk of either having purchased above market price or above the negotiated fixed price. Currently, buyers manage this uncertainty manually and based on their own intuition. Intelligent algorithms that collect, process, and predict the risk of available sourcing options to support the intuition of human buyers are the key missing element for establishing an efficient strategic sourcing system.IntuitBuy intends to fill this gap by providing buyers with an intelligent system that processes structured market data along with unstructured data, like news flashes or satellite images, to help buyers allocate volumes to suppliers and inform them which pricing option to choose. As buyer decisions may affect the market price as well as influence the behaviour of suppliers in the future, an important part of IntuitBuy will be dedicated to understanding whether it is possible to use the data to anticipate supplier gaming. This would not only help buyers of coking coal, but also provide a blueprint for other categories, where either the company or its supplier have larger negotiation power.

This site uses cookies. By continuing to use this site, you agree to the use of cookies for analytics purposes. Find out more in our Privacy Statement