Bayesian Methods in Bidding Processes

• Activity Number: 178
• Type: Topic Contributed
• Date/Time: Monday, August 8, 2005 : 2:00 PM to 3:50 PM
• Sponsor: Section on Bayesian Statistical Science
• Author(s): Jesus Palomo*+
• Companies: SAMSI/Duke University
• Keywords: Bayesian inference ; Project management ; Decision analysis ; Risk management

Abstract

In this paper, I propose a comprehensive framework, extending the traditional project management methods to cope with the whole bidding process in procurement contract auctions. I use a Bayesian approach to model, from a project life cycle perspective, the uncertainties of such a risky market mechanism. For a given auction, the framework proposed is conceived as a sequential process, modeling one uncertainty type at each step in turn. I start with the estimation of the forecasting project cost distribution with a dynamic model, allowing for additional input from an expert. Then, as several disruptive events could arise during the development of the contract, entailing additional project costs, I provide models to forecast the probability of such events and their combined effects. Once with an appropriate forecast, the company is ready to submit a bid. As the company needs to win the auction to execute the project, I formulate a bidding strategy that maximizes the expected utility of the bidder in an asymmetric first-price, sealed-bid auction. Finally, once actual data arrive, posterior distributions are computed to incorporate the knowledge for future proposal processes.