ASSIGNMENT 2

Assignment 2

Assignment 2

Introduction

Prediction Markets are exploratory markets which are formed for the function of coming up with future forecasts of events. These are synonymous to stock exchanges, in which securities are traded, except that prediction markets are intangible (virtual). The market prices of these securities which are traded in prediction markets are denoting the probability of occurrence of an event. Application of prediction markets has been quite successful following the accuracy at which it predicted elections results, and given that the significance error was quite low, (amazingly) all around the world (Rhode et al. 2003, p.n.d.).

Google Inc., founded in 1988 by Larry page and Sergey Brin, had began the Google Prediction Markets project in 2003, owing to the efforts of the “Googlers team” which included Doug Banks, Patri Friedman, Ilya Kirnosl, and Piaw Na, the people who were credited for designing the GPM prototype. These people were regular Google employees who used their 20% off time from work each work to come up with ideas to fill this gap that existed in the market. Companies had been relying on intuition and predictions which had no evidence based or analytical support whatsoever. This prototype, therefore, had an already established market that it could be applied to.

The aim of this proposal is to serve as a consultancy report to the C.E.O. of Google Mr. Larry Page, in lieu of the decision being faced by the company, as outlined in the case study. Relevant qualitative and/or quantitative tools for studying the background will also be incorporated. A recommendation would be made and a forecast of the success probability and steps necessary to be taken in order to implement the project would be discussed. The paper will also consider strategic, operational and financial challenges being faced by the industry giant at the moment, apart from discussing how the value chain of Google is configured with the industry drivers.

Literature Review

A qualitative study conducted by in 2006 made comparisons on empirical basis of the logical approaches in order to provide evidence of the most suitable forecasting techniques under predefined criteria. It reviewed the efforts made in the 20th century for minimizing errors in forecasting, and discussed 7 renowned methods in this regard, including Delphi help, causal modelling, structural judgement and judgemental bootstrapping. Data consisted of time series and cross-sections, and trend damping help was included (Graefe & Armstrong 2011, pp.183-195). The recommended methods with prospective results included replicated interaction, moistened causality, organized analogies, and judgement decay for data that was cross-sectional; whereas for time series based data, the suggestions included forecasting based on rules, categorization, and moistened seasonality. Hypothesis testing showed that these methods are not suitable in certain situations. Further hypothesis testing had been suggested to explore other methods (Armstrong 2006, pp.583-598).

Another study conducted in 2009 attempted to define prediction markets as being a medium for trading certain indentures and bonds that would yield revenues based on the results of events that were not certain. Growing evidence has been supporting the theory that prediction ...