Thesis (M. Eng. in Logistics)--Massachusetts Institute of Technology, Engineering Systems Division, 2009. / Includes bibliographical references (leaves 105-106). / Prediction Markets hold the promise of improving the forecasting process. Research has shown that Prediction Markets can develop more accurate forecasts than polls or experts. Our research concentrated on analyzing Prediction Markets for business decision-making. We configured a Prediction Market to gather primary data, sent out surveys to gauge participant views and conducted in-depth interviews to explain trader behavior. Our research was conducted with 169 employees from General Mills who participated in Prediction Markets that lasted from two to ten weeks. Our research indicates that short term forecasting Prediction Markets are no more accurate than conventional forecasting methods. It also presents and addresses three interesting contradictions. First, the Sales Organization won the majority of the Prediction Markets, yet the overall performance of Sales as a group was worse than that of other groups. Second, Prediction Markets were able to gain access to more information than General Mills' current process, yet the impact on forecast accuracy was not significant. Third, with a MAPE of 11% for promotional Prediction Markets, it would seem that promotional demand was well understood up-front, yet when we dissected the promotional forecasts we discovered that participants changed their minds over time degrading overall forecast accuracy. We believe that we have extended the current body of work on Prediction Markets in ways that will increase the utilization in business environments. / by Kai Trepte and Rajaram Narayanaswamy. / M.Eng.in Logistics
| Identifer | oai:union.ndltd.org:MIT/oai:dspace.mit.edu:1721.1/53546 |
| Date | January 2009 |
| Creators | Trepte, Kai, Narayanaswamy, Rajaram |
| Contributors | Larry Lapide., Massachusetts Institute of Technology. Engineering Systems Division., Massachusetts Institute of Technology. Engineering Systems Division. |
| Publisher | Massachusetts Institute of Technology |
| Source Sets | M.I.T. Theses and Dissertation |
| Language | English |
| Detected Language | English |
| Type | Thesis |
| Format | 106 leaves, application/pdf |
| Rights | M.I.T. theses are protected by copyright. They may be viewed from this source for any purpose, but reproduction or distribution in any format is prohibited without written permission. See provided URL for inquiries about permission., http://dspace.mit.edu/handle/1721.1/7582 |
Page generated in 0.0018 seconds