I develop a new mechanism for risk allocation and information speculation called a dynamic pari-mutuel market (DPM). A DPM acts as hybrid between a pari-mutuel market and a continuous double auction (CDA), inheriting some of the advantages of both. Like a pari-mutuel market, a DPM offers infinite buy-in liquidity and zero risk for the market institution; like a CDA, a DPM can continuously react to new information, dynamically incorporate information into prices, and allow traders to lock in gains or limit losses by selling prior to event resolution. The trader interface can be designed to mimic the familiar double auction format with bid-ask queues, though with an addition variable called the payoff per share. The DPM price function can be viewed as an automated market maker always offering to sell at some price, and moving the price appropriately according to demand. Since the mechanism is pari-mutuel (i.e., redistributive), it is guaranteed to pay out exactly the amount of money taken in. I explore a number of variations on the basic DPM, analyzing the properties of each, and solving in closed form for their respective price functions.

Game sites on the World Wide Web draw people from around the world with specialized interests, skills, and knowledge.

We develop a framework for trading in compound securities: financial instruments that pay off contingent on the outcomes of arbitrary statements in propositional logic. Buying or selling securities -- which can be thought of as betting on or against a particular future outcome -- allows agents both to hedge risk and to profit (in expectation) on subjective predictions. A compound securities market allows agents to place bets on arbitrary boolean combinations of events, enabling them to more closely achieve their optimal risk exposure, and enabling the market as a whole to more closely achieve the social optimum. The tradeoff for allowing such expressivity is in the complexity of the agents' and auctioneer's optimization problems.

According to economic theory, supported by empirical and laboratory evidence, the equilibrium price of a financial security reflects all of the information regarding the security's value. We investigate the dynamics of the computational process on the path toward equilibrium, where information distributed among traders is revealed stepby -step over time and incorporated into the market price. We develop a simplified model of an information market, along with trading strategies, in order to formalize the computational properties of the process. We show that securities whose payoffs cannot be expressed as a weighted threshold function of distributed input bits are not guaranteed to converge to the proper equilibrium predicted by economic theory. On the other hand, securities whose payoffs are threshold functions are guaranteed to converge, for all prior probability distributions. Moreover, these threshold securities converge in at most n rounds, where n is the number of bits of distributed information. We also prove a lower bound, showing a type of threshold security that requires at least n/2 rounds to converge in the worst case.

We report results of an internet experiment designed to test the theory of informational cascades in financial markets. More than 6000 subjects, including a subsample of 267 consultants from an international consulting firm, participated in the experiment. As predicted by theory, we find that the presence of a flexible market price prevents herding. However, the presence of contrarian behavior, which can (partly) be rationalized via error models, distorts prices, and even after 20 decisions convergence to the fundamental value is rare. We also study the effects of transaction costs and the expectations of subjects with respect to future prices. Finally, we look at the behavior of various subsamples of our heterogeneous subject pool.

We develop a model of how information flows into a market, and derive algorithms for automatically detecting and explaining relevant events. We analyze data from twenty-two "political stock markets" (i.e., betting markets on political outcomes) on the Iowa Electronic Market (IEM). We prove that, under certain eciency assumptions, prices in such betting markets will on average approach the correct outcomes over time, and show that IEM data conforms closely to the theory. We present a simple model of a betting market where information is revealed over time, and show a qualitative correspondence between the model and real market data. We also present an algorithm for automatically detecting significant events and generating semantic explanations of their origin. The algorithm operates by discovering significant changes in vocabulary on online news sources (using expected entropy loss) that align with major price spikes in related betting markets.

We analyze the eciency and forecast accuracy of two market games on the World Wide Web: the Hollywood Stock Exchange (HSX) and the Foresight Exchange (FX). We quantify the degree of arbitrage available on HSX, and compare with a real-money market of a similar nature. We show that prices of HSX movie stocks provide good forecasts of actual box oce returns, and that prices of HSX securities in Oscar, Emmy, and Grammy award outcomes constitute accurate assessments of the actual likelihoods that nominees will win. Similar investigations reveal that FX securities prices serve as reliable indicators of uncertain future events. We argue that, in certain circumstances, market simulations can furnish some of the same societal benets as real markets, and can serve as acceptable substitute testbeds for conducting experiments that would otherwise be dicult or impossible.

Prediction markets are used in real life to predict outcomes of interest such as presidential elections. In this work we introduce a mathematical theory for Artificial Prediction Markets for supervised classifier aggregation and probability estimation. We introduce the artificial prediction market as a novel way to aggregate classifiers. We derive the market equations to enforce total budget conservation, show the market price uniqueness and give efficient algorithms for computing it. We show how to train the market participants by updating their budgets using training examples. We introduce classifier specialization as a new differentiating characteristic between classifiers. Finally, we present experiments using random decision rules as specialized classifiers and show that the prediction market consistently outperforms Random Forest on real and synthetic data of varying degrees of difficulty. 1.

The capacity of markets to aggregate information has been conclusively demonstrated but the limitations of that capacityhave still not been fully explored. In this paper, we demonstrate the existence of #information traps". These traps appear to be a sort of equilibrium in which information existing in the market does not become revealed in prices. The foundation for the equilibrium is a pattern of misaligned beliefs in which each person's actions are based upon mistaken beliefs about the information held by others. The mistakes, themselves, haveatype of mutual compatibility and cannot become revealed by the price discovery process because individuals have no incentives or resources to adjust. Attempts to probe the nature of the phenomena involved two period markets with a contingent claim instrument, experienced participants, and unlimited short selling opportunities. # The authors gratefully acknowledge the #nancial support for this research whichwas provided by the Caltech...

We propose a dynamic model of parimutuel betting that addresses the following three empirical regularities: a sizeable fraction of bets is placed early, late bets are more informative than early bets, and proportionally too many bets are placed on longshots. Exploiting a similarity with Cournot oligopoly, we show that bettors have an incentive to bet early when they are large and act on common information. Bettors who are instead privately informed and small have an incentive to bet at the end without access to the information of the others. Longshots (or favorites) are then less (or more) likely to win than indicated by the market odds.