Developing and implementing a stochastic decision-support model within an organizational context : The experience

Hoyland, Kjetil and Ranberg, Erik and Wallace, Stein W (2005) Developing and implementing a stochastic decision-support model within an organizational context : The experience. The Journal of Risk Finance, 6 (1). pp. 40-46. ISSN 1526-5943

Abstract

Purpose – Discusses why it is necessary to align a mathematical model with the organization in order to achieve the desired results. The structure of a model's input must fit with the structure of data collection in the firm, and the output must be consistent with the decision structure. Otherwise, data collection will not be properly taken care of and the results of a model will not find their way to where decisions are made. Five years passed from the cooperation first started with the university until the model came on-line. Design/methodology/approach – Parts 1 and 2 of this series of papers discussed the stochastic programming model itself and the relationship between the model and the organization. The results are now reported. Findings – Reports on both organizational and financial results. Practical implications – Shows that, although a lot of work is needed to implement a complicated stochastic programming model within an organization, it can be done and can lead to good results. However, it takes time to adjust an organization to a new way of thinking, as a model like this can never work unless supported by upper management as well as those who use the model daily. Originality/value – The main value is to demonstrate in rather large detail how a somewhat complicated model can be developed and used in an organization, which, originally, was not well aligned with the ideas of the modellers.