We present an experimental and simulated model of a multi-agent stock market driven by a double auction order matching mechanism. Studying the effect of cumulative information on the performance of traders, we find a non monotonic relationship of net returns of traders as a function of information levels, both in the experiments and in the simulations. Particularly, averagely informed traders perform worse than the non informed and only traders with high levels of information (insiders) are able to beat the market. The simulations and the experiments reproduce many stylized facts of stock markets, such as fast decay of autocorrelation of returns, volatility clustering and fat-tailed distribution of returns. These results have an important message for everyday life. They can give a possible explanation why, on average, professional fund managers perform worse than the market index.
The average economic agent is often used to model the dynamics of simple markets, based on the assumption that the dynamics of many agents can be averaged over in time and space. A popular idea that is based on this seemingly intuitive notion is to dampen electric power fluctuations from fluctuating sources (as e.g. wind or solar) via a market mechanism, namely by variable power prices that adapt demand to supply. The standard model of an average economic agent predicts that fluctuations are reduced by such an adaptive pricing mechanism. However, the underlying assumption that the actions of a…
We present an analysis of the price impact associated with trades effected by different financial firms. Using data from the Spanish Stock Market, we find a high degree of heterogeneity across different market members, both in the instantaneous impact functions and in the time-dependent market response to trades by individual members. This heterogeneity is statistically incompatible with the existence of market-wide universal impact dynamics which apply uniformly to all trades and suggests that rather, market dynamics emerge from the complex interaction of different behaviors of market partici…
In this paper we present an interacting-agent model of stock markets. We describe a stock market through an Ising-like model in order to formulate the tendency of traders getting to be influenced by the other traders' investment attitudes [1], and formulate the traders' decision-making regarding investment as the maximum entropy principle for nonextensive entropy. We demonstrate that the equilibrium probability distribution function of the traders' investment attitude is the {\it q-exponential distribution}. We also show that the power-law distribution of the volatility of price fluctuations, …
We present results on simulations of a stock market with heterogeneous, cumulative information setup. We find a non-monotonic behaviour of traders' returns as a function of their information level. Particularly, the average informed agents underperform random traders; only the most informed agents are able to beat the market. We also study the effect of a strategy updating mechanism, when traders have the possibility of using other pieces of information than the fundamental value. These results corroborate the latter ones: it is only for the most informed player that it is rewarding to stay fu…
This review deals with several microscopic (``agent-based'') models of financial markets which have been studied by economists and physicists over the last decade: Kim-Markowitz, Levy-Levy-Solomon, Cont-Bouchaud, Solomon-Weisbuch, Lux-Marchesi, Donangelo-Sneppen and Solomon-Levy-Huang. After an overview of simulation approaches in financial economics, we first give a summary of the Donangelo-Sneppen model of monetary exchange and compare it with related models in economics literature. Our selective review then outlines the main ingredients of some influential early models of multi-agent dynami…
We propose an artificial market model based on deterministic agents. The agents modify their ask/bid price depending on past price changes. The temporal development of market price fluctuations is calculated numerically. A probability density function of market price changes has power law tails. Autocorrelation coefficient of the changes has an anti-correlation, and autocorrelation coefficient of squared changes (volatility correlation function) has a long time correlation. A probability density function of intervals between successive trading follows a geometric distribution. GARCH type stoch…
Real world markets display power-law features in variables such as price fluctuations in stocks. To further understand market behavior, we have conducted a series of market experiments on our web-based prediction market platform which allows us to reconstruct transaction networks among traders. From these networks, we are able to record the degree of a trader, the size of a community of traders, the transaction time interval among traders and other variables that are of interest. The distributions of all these variables show power-law behavior. On the other hand, agent-based models have been p…
We use generating functional analysis to study minority-game type market models with generalized strategy valuation updates that control the psychology of agents' actions. The agents' choice between trend following and contrarian trading, and their vigor in each, depends on the overall state of the market. Even in `fake history' models, the theory now involves an effective overall bid process (coupled to the effective agent process) which can exhibit profound remanence effects and new phase transitions. For some models the bid process can be solved directly, others require Maxwell-construction…
This paper reports the effort of using agent-based mix-game model to predict financial time series. It introduces simple generic algorithm into the prediction methodology, and gives an example of its application to forecasting Shanghai Index. The results show that this prediction methodology is effective and agent-based mix-game model is a potential good model to predict time series of financial markets
We extend to the multi-asset case the framework of a discrete time model of a single asset financial market developed in Ghoulmie et al (2005). In particular, we focus on adaptive agents with threshold behavior allocating their resources among two assets. We explore numerically the effect of this diversification as an additional source of complexity in the financial market and we discuss its destabilizing role. We also point out the relevance of these studies for financial decision making.
A network effect is introduced taking into account competition, cooperation and mixed-type interaction amongst agents along a generalized Verhulst-Lotka-Volterra model. It is also argued that the presence of a market capacity enforces an indubious limit on the agent's size growth. The state stability of triadic agents, i.e., the most basic network plaquette, is investigated analytically for possible scenarios, through a fixed point analysis. It is discovered that: (i) \market" demand is only satisfied for full competition when one agent monopolizes the market; (ii) growth of agent size is enco…
We review the recent approaches to modelling financial markets based on multi-agent systems. After a brief summary of the basic stylised facts observed in real-market time-series we discuss some simple agent-based systems which are currently used to model financial markets. One of the most prominent examples is here the Minority Game (MG), which we address in some more detail. After a brief discussion of its basic setup and general phenomenology we summarise the main findings of the statistical mechanics analysis and discuss the emergence of stylised facts in extensions of the MG near their ph…
Generalization of the minority game to more than one market is considered. At each time step every agent chooses one of its strategies and acts on the market related to this strategy. If the payoff function allows for strong fluctuation of utility then market occupancies become inhomogeneous with preference given to this market where the fluctuation occured first. There exists a critical size of agent population above which agents on bigger market behave collectively. In this regime there always exists a history of decisions for which all agents on a bigger market react identically.
Manipulation in Prediction Markets: An Agent-based Modeling Experiment
Institute for New Economic Thinking, University of Oxford（牛津大学新经济思维研究所）；Mathematical Institute, University of Oxford（牛津大学数学研究所）；Institute for Globally Distributed Open Research and Education (IGDORE)（全球分布式开放研究与教育学院）；Economics Institute, Pontifical Catholic University of Chile（智利天主教大学经济研究所）
We study social behaviour of agents on capital markets when these are perturbed by small perturbations. We use the mean field method. Social behaviour of agents on capital markets is described: volatility of the market, aversion constant and equilibrium states are discussed. Relaxation behaviour of agents on the capital market is studied. Equation of motion for the agent average number is of the relaxation type. Development of the group of agents in the states corresponding to minimum of the aim function is either linear either exponentially damped. There exist characteristic volatility consta…
We propose a simple statistical-physics-inspired model for the effect of intrinsic fluctuations on supply and demand in markets. The model consists of agents that trade in two types of goods of which the total number is separately conserved. The relative preference of an individual agent for the two types of goods is determined by a utility that is identical for all agents. Market supply and demand curves are computed and compared for various motivated choices of the distribution of goods over the agents. In particular, we compare the "mean-field" case, in which all agents have the same number…
We introduce an auto-regressive model which captures the growing nature of realistic markets. In our model agents do not trade with other agents, they interact indirectly only through a market. Change of their wealth depends, linearly on how much they invest, and stochastically on how much they gain from the noisy market. The average wealth of the market could be fixed or growing. We show that in a market where investment capacity of agents differ, average wealth of agents generically follow the Pareto-law. In few cases, the individual distribution of wealth of every agent could also be obtain…
Urban housing markets, along with markets of other assets, universally exhibit periods of strong price increases followed by sharp corrections. The mechanisms generating such non-linearities are not yet well understood. We develop an agent-based model populated by a large number of heterogeneous households. The agents' behavior is compatible with economic rationality, with the trend-following behavior found to be essential in replicating market dynamics. The model is calibrated using several large and distributed datasets of the Greater Sydney region (demographic, economic and financial) acros…
Phenomena which involves collective choice of many agents who are interacting with each other and choosing one of several alternatives, based on the limited information available to them, frequently show switching between two distinct phases characterized by a bimodal and an unimodal distribution respectively. Examples include financial markets, movie popularity and electoral behavior. Here we present a model for this biphasic behavior and argue that it arises from interactions in a local neighborhood and adaptation & learning based on information about the effectiveness of past choices.
Housing markets are inherently spatial, yet many existing models fail to capture this spatial dimension. Here we introduce a new graph-based approach for incorporating a spatial component in a large-scale urban housing agent-based model (ABM). The model explicitly captures several social and economic factors that influence the agents' decision-making behaviour (such as fear of missing out, their trend following aptitude, and the strength of their submarket outreach), and interprets these factors in spatial terms. The proposed model is calibrated and validated with the housing market data for t…
We introduce a stochastic heterogeneous interacting-agent model for the short-time non-equilibrium evolution of excess demand and price in a stylized asset market. We consider a combination of social interaction within peer groups and individually heterogeneous fundamentalist trading decisions which take into account the market price and the perceived fundamental value of the asset. The resulting excess demand is coupled to the market price. Rigorous analysis reveals that this feedback may lead to price oscillations, a single bounce, or monotonic price behaviour. The model is a rare example of…
