determine the impact of announcements or news on the forecasted capitalization during the course of the prediction market;
learn about how and when the price formation process aggregated information for these markets; and
(5) estimate the degree of ex ante uncertainty about the post-IPO market price and show how it varied through time and surrounding events.
In addition, the combination of the two markets we conducted allows us to generate two different
forecasts for Google’s market capitalization, compare them and analyze whether contract structure
matters for prediction markets.
i. The Google Linear Market
The Google Linear market opened on June 29, 2004 with two contracts.16
values were determined as follows:
Contract Liquidation Values = $0 if the IPO does not take place by March 31, 2005; = (Market Cap.)/$100 billion if $0 bil. < Market Cap. <= $100 bil; = $1 if Market Cap. > $100 bil. = $1 if the IPO does not take place by March 31, 2005; = ($100 bil.-Market Cap.)/$100 billion if $0 bil. < Market Cap. <= $100 bil; = $0 if Market Cap. > $100 bil.
In the absence of hedging demand, prices should equal expected values in this market.17 Thus, the price
of IPO_UP times $100 billion is the IEM’s forecast of the market capitalization of Google stock after the
The appendix contains the prospectus for this market. This argument can be made in numerous ways. For example, modern option pricing theory implies that prices should equal expected liquidation values according to the risk neutral distribution discounted back at the risk free rate. Risk neutral probabilities are driven away from true probabilities by imbalances in hedging demand. The small size of these markets along with evidence on behavior and prices in political markets (e.g., Forsythe, Nelson, Neumann and Wright, 1992; Forsythe, Rietz and Ross, 1999; and Oliven and Rietz, 2004) suggest that hedging demands are not significant factors in determining prices. The risk free rate in these markets is zero because contract bundles (one risk free asset) and cash (the other risk free asset) both earn a zero return and can be freely exchanged for one another. As a result of these two factors, option pricing theory implies that prices should equal actual expected liquidation values at each point in time. Similar arguments (using the absence of systematic risk factors and a zero risk free rate) can be made using CAPM, APT or general equilibrium theory to get the same result. Whether prices actually reflect expected values is an empirical matter and the evidence suggests that they do in IEM markets in general (see, for example, Berg, Forsythe, Nelson and Rietz, 2003; and Berg, Nelson and Rietz, 2003).