FROM DISCRETE-TIME MODELS TO CONTINUOUS-TIME, ASYNCHRONOUS MODELING OF FINANCIAL MARKETS

Most agent-based simulation models of financial markets are discrete-time in nature. In this paper, we investigate to what degree such models are extensible to continuous-time, asynchronous modeling of financial markets. We study the behavior of a learning market maker in a market with information asymmetry, and investigate the difference caused in the market dynamics between the discrete-time simulation and continuous-time, asynchronous simulation. We show that the characteristics of the market prices are different in the two cases, and observe that additional information is being revealed in the continuous-time, asynchronous models, which can be acted upon by the agents in such models. Because most financial markets are continuous and asynchronous in nature, our results indicate that explicit consideration of this fundamental characteristic of financial markets cannot be ignored in their agent-based modeling.     

Management decision support using long-term market simulation

In recent years Agent-based Computational Economics (ACE) has become an increasingly important method in market simulation. After liberalization of many former governmental owned or controlled industries the used operations research models are not longer sufficient to simulate market behavior due to individual action and increasing competition. Agent-based simulation appears to be an alternative approach considering also individual behavior and competition. Some short-term simulation approaches have shown promising results for the simulation in the domain of electricity markets. Picking up the desire for a long-term oriented simulation, this paper presents a basic agent-based model considering the investment decision within long-term planning of electricity markets. Additionally, regulatory agents are introduced as a third side in the market simulation to represent governmental decisions. This results in the definition of three types of agents representing electricity generating companies, consumers and governmental instances.     

A multi-agent approach for planning activities in decentralized electricity markets

Despite the rigour and ability of game theory to cope with oligopolistic electric markets, it fails to model many existing behaviours in the real-world circumstances. The traditional models such as statistical extrapolation or econometrics are not capable to anticipate the changes in the pattern of the market prices due to the future structural changes. Furthermore, in such free and open markets, there is a more intense need for each participant to benefit from a certain level of autonomy, while keeping some capabilities to interact, communicate, collaborate and negotiate with other participants in an efficient way. As a result, there is a need for a novel framework of modelling that could include game theoretical assumptions as well as other more complex assumptions. Agent technologies in general and agent-based simulation in particular offer this possibility. This paper proposes, in a decision-making perspective, a new multi-agent architecture specifically designed to support planning activities in decentralized electricity markets, with a certain level of flexibility. In this model, synthetic agents are created allowing flexible representations of the multi-functional market players and possible mergers and coalitions in the electricity market.     

Modeling spot price dependence in Australian electricity markets with applications to risk management

We examine the dependence structure of electricity spot prices across regional markets in Australia. One of the major objectives in establishing a national electricity market was to provide a nationally integrated and efficient electricity market, limiting market power of generators in the separate regional markets. Our analysis is based on a GARCH approach to model the marginal price series in the considered regions in combination with copulae to capture the dependence structure between the marginals. We apply different copula models including Archimedean, elliptical and copula mixture models. We find a positive dependence structure between the prices for all considered markets, while the strongest dependence is exhibited between markets that are connected via interconnector transmission lines. Regarding the nature of dependence, the Student-t copula provides a good fit to the data, while the overall best results are obtained using copula mixture models due to their ability to also capture asymmetric dependence in the tails of the distribution. Interestingly, our results also suggest that for the four major markets, NSW, QLD, SA and VIC, the degree of dependence has decreased starting from the year 2008 towards the end of the sample period in 2010. Examining the Value-at-Risk of stylized portfolios constructed from electricity spot contracts in different markets, we find that the Student-t and mixture copula models outperform the Gaussian copula in a backtesting study. Our results are important for risk management and hedging decisions of market participants, in particular for those operating in several regional markets simultaneously.     

Agent-Based Social Simulation in Markets

We show that certain desired behavioural properties of agent-based models can be deterministically induced by an appropriate mathematical structure. We also point out problems related to the handling of parameters, and of the modelling of time, in agent-based models. Our purpose is to illustrate some problems of agent-based social simulations in markets, as a first step towards the more ambitious goal of providing a methodology for such simulations.     

A Simulation Framework for Heterogeneous Agents

We introduce a generic simulation framework suitable for agent-based simulations featuring the support of heterogeneous agents, hierarchical scheduling, and flexible specification of design parameters. One key aspect of this framework is the design specification: we use a format based on the Extendible Markup Language (XML) that is simple-structured yet still enables the design of flexible models. Another issue in agent-based simulations, especially when ready-made components are used, is the heterogeneity arising from both the agents' implementations and the underlying platforms. To tackle such obstacles, we introduce a wrapper technique for mapping the functionality of agents living in an interpreter-based environment to a standardized JAVA interface, thus facilitating the task for any control mechanism (like a simulation manager) because it has to handle only one set of commands for all agents involved. Again, this mapping is made by an XML-based definition format. We demonstrate the technique by applying it to a simple sample simulation of two mass marketing firms operating in an artificial consumer environment.     

Agent-based computational modeling of the stock price-volume relation

From the perspective of the agent-based model of stock markets, this paper examines the possible explanations for the presence of the causal relation between stock returns and trading volume. Using the agent-based approach, we find that the explanation for the presence of the stock price-volume relation may be more fundamental. Conventional devices such as information asymmetry, reaction asymmetry, noise traders or tax motives are not explicitly required. In fact, our simulation results show that the stock price-volume relation may be regarded as a generic property of a financial market, when it is correctly represented as an evolving decentralized system of autonomous interacting agents. One striking feature of agent-based models is the rich profile of agents' behavior. This paper makes use of the advantage and investigates the micro-macro relations within the market. In particular, we trace the evolution of agents' beliefs and examine their consistency with the observed aggregate market behavior. We argue that a full understanding of the price-volume relation cannot be accomplished unless the feedback relation between individual behavior at the bottom and aggregate phenomena at the top is well understood.     

Design classes for hybrid simulations involving agent-based and system dynamics models

Hybrid simulation involves the use of multiple simulation paradigms, and is becoming an increasingly common approach to modelling modern, complex systems. Despite growing interest in its use, little guidance exists for modellers regarding the nature and variety of hybrid simulation models. Here, we concentrate on one particular hybrid – that involving agent-based and system dynamics models. Based on an up-to-date review of the literature, we propose three basic types of hybrid agent-based system dynamics simulations, referred to here as interfaced, integrated and sequential hybrid designs. We speculate that the classification presented may also be useful for other classes of hybrid simulations.     

基于Agent的复杂系统分布仿真建模方法的研究

基于Agent的分布仿真是研究大型复杂系统的一种有效的、重要的方法。为了减小复杂系统仿真的复杂度,增加仿真模型的重用和可维护性,需要研究基于Agent分布仿真的建模方法。首先对复杂系统及其特性进行了分析,对基于Agent的仿真进行了全面的论述,然后对基于Agent的复杂系统仿真中的复杂系统建模分析、Agent建模分析以及Agent的分布进行了分析,给出了基于Agent的复杂系统分布仿真的建模步骤,最后给出了在此建模思想指导下的金融证券市场的建模过程。     

Care HPS: A high performance simulation tool for parallel and distributed agent-based modeling

Parallel and distributed simulation is a powerful tool for developing complex agent-based simulation. Complex simulations require parallel and distributed high performance computing solutions. It is necessary because their sequential solutions are not able to give answers in a feasible total execution time. Therefore, for the advance of computing science, it is important that High Performance Computing (HPC) techniques and solutions be proposed and studied. In literature, we can find some agent-based modeling and simulation tools that use HPC. However, none of these tools are designed to enable the HPC expert to be able to propose new techniques and solutions without great effort. In this paper, we introduce Care High Performance Simulation (HPS), which is a scientific instrument that enables researchers to: (1) develop techniques and solutions of high performance distributed simulations for agent-based models; and, (2) study, design and implement complex agent-based models that require HPC solutions. Care HPS was designed to easily and quickly develop new agent-based models. It was also designed to extend and implement new solutions for the main issues of parallel and distributed solutions such as: synchronization, communication, load and computing balancing, and partitioning algorithms. We conducted some experiments with the aim of showing the completeness and functionality of Care HPS. As a result, we show that Care HPS can be used as a scientific instrument for the advance of the agent-based parallel and distributed simulations field.     

基于Agent仿真模型的校核与验证研究

首先对国内外基于Agent仿真模型的校核与验证方面的发展情况进行了述评;然后,提出了一个完整的基于Agent仿真模型的校核与验证框架,该方案中包括表面验证、参数灵敏度分析、模型校准与运行时验证;最后以环境经济政策仿真模型中的校核与验证为例对该框架进行了简单说明。     

基于Agent计算金融的计算机仿真研究综述

传统的金融学研究方法为"由上及下",此类方法倚重于总体把握,而忽视个体行为,特别是缺乏个体以及个体与环境之间的互动.基于Agent的计算金融是一种新的研究金融市场行为的工具.首先阐述了金融市场本身的复杂性和传统金融学存在的一些困惑.在这基础上提出了基于Agent的计算金融方法.其次,概要介绍了目前基于Agent的人工金融市场的主要设计方法,并且提出了在设计过程中要注意的一些问题.最后,分析了这种方法相对于传统金融学研究方法的优点和存在的问题,并进一步提出了该领域的未来研究方向.     

An extensible simulation environment and movement metrics for testing walking behavior in agent-based models

Human movement is a significant ingredient of many social, environmental, and technical systems, yet the importance of movement is often discounted in considering systems’ complexity. Movement is commonly abstracted in agent-based modeling (which is perhaps the methodological vehicle for modeling complex systems), despite the influence of movement upon information exchange and adaptation in a system. In particular, agent-based models of urban pedestrians often treat movement in proxy form at the expense of faithfully treating movement behavior with realistic agency. There exists little consensus about which method is appropriate for representing movement in agent-based schemes. In this paper, we examine popularly-used methods to drive movement in agent-based models, first by introducing a methodology that can flexibly handle many representations of movement at many different scales and second, introducing a suite of tools to benchmark agent movement between models and against real-world trajectory data. We find that most popular movement schemes do a relatively poor job of representing movement, but that some schemes may well be “good enough” for some applications. We also discuss potential avenues for improving the representation of movement in agent-based frameworks.     

Reverse Engineering Financial Markets with Majority and Minority Games Using Genetic Algorithms

Using virtual stock markets with artificial interacting software investors, aka agent-based models, we present a method to reverse engineer real-world financial time series. We model financial markets as made of a large number of interacting boundedly rational agents. By optimizing the similarity between the actual data and that generated by the reconstructed virtual stock market, we obtain parameters and strategies, which reveal some of the inner workings of the target stock market. We validate our approach by out-of-sample predictions of directional moves of the Nasdaq Composite Index.     

Automating agent-based modeling: Data-driven generation and application of innovation diffusion models

Simulation modeling is useful to understand the mechanisms of the diffusion of innovations, which can be used for forecasting the future of innovations. This study aims to make the identification of such mechanisms less costly in time and labor. We present an approach that automates the generation of diffusion models by: (1) preprocessing of empirical data on the diffusion of a specific innovation, taken out by the user; (2) testing variations of agent-based models for their capability of explaining the data; (3) assessing interventions for their potential to influence the spreading of the innovation. We present a working software implementation of this procedure and apply it to the diffusion of water-saving showerheads. The presented procedure successfully generated simulation models that explained diffusion data. This progresses agent-based modeling methodologically by enabling detailed modeling at relative simplicity for users. This widens the circle of persons that can use simulation to shape innovation.     

A model of friendship networks based on social network analysis

It is important for teachers to understand the characteristics of forming and changing of friendship networks to manage their classes effectively. Many studies have analyzed friendship networks using agent-based models. Most of the models used were designed for capturing basic phenomena in a micro-viewpoint. The purpose of our study is to simulate actual friendship networks. In this paper, we present an agent-based friendship-network model based on our analysis of friendship networks among university students. We focused on Hiramatsu’s model based on a cost-benefit analysis and modified the model with the aim of closing to the actual friendship. With our modified model, the benefit function was based on their cluster structure. The results with the modified model better matched the actual friendship networks. The clique size in the modified model approached that in the actual friendship networks, and the values of several network properties approached those of the actual friendship networks.     

Application of the G-JF discrete-time thermostat for fast and accurate molecular simulations

A new Langevin–Verlet thermostat that preserves the fluctuation–dissipation relationship for discrete time steps is applied to molecular modeling and tested against several popular suites (AMBER, GROMACS, LAMMPS) using a small molecule as an example that can be easily simulated by all three packages. Contrary to existing methods, the new thermostat exhibits no detectable changes in the sampling statistics as the time step is varied in the entire numerical stability range. The simple form of the method, which we express in the three common forms (Velocity-Explicit, Störmer–Verlet, and Leap-Frog), allows for easy implementation within existing molecular simulation packages to achieve faster and more accurate results with no cost in either computing time or programming complexity.     

Agent-based material handling and inventory planning in warehouse

Agent technology has become popular in the last few years as an effective approach to develop software and hardware systems. Our research applies this technology to warehouse systems. Since warehouse system is crucial to the success of supply-chain systems, objective of our research is to implement a framework of warehouse system based on the agent technology. This paper proposes an agent-based model for warehouse system, which is called AWAS (Agent-based model for WArehouse System). AWAS is composed of three subsystems, or agent-based communication system (ACS), agent-based material handling system (AMATH), and agent-based inventory planning and control system (AIPCON). These subsystems are designed to cooperate together to facilitate just-in-time exchange of orders and materials. Under these subsystems, seven kinds of basic agents are defined, including customer, supplier, order, inventory, product, supplier-order, and automatic-guided vehicle (AGV) agents. We will describe our approach in designing and implementing warehouse systems simulation with agents. Using a prototype system which was developed based on the model, the paper presents some results of simulation to show the validity of the model.     

Behavioral Calibration and Analysis of a Large-Scale Travel Microsimulation

This article reports on the calibration and analysis of a fully disaggregate (agent-based) transport simulation for the metropolitan area of Zurich. The agent-based simulation goes beyond traditional transport models in that it equilibrates not only route choice but all-day travel behavior, including departure time choice and mode choice. Previous work has shown that the application of a novel calibration technique that adjusts all choice dimensions at once from traffic counts yields cross-validation results that are competitive with any state-of-the-art four-step model. While the previous study aims at a methodological illustration of the calibration method, this work focuses on the real-world scenario, and it elaborates on the usefulness of the obtained results for further demand analysis purposes.