Search

In this paper, we consider the approach of applying state-of-the-art machine learning algorithms to simulate some financial markets. In this case, we choose the cryptocurrency market based on the assumption that such markets more active today. As a rule, they have more volatility, attracting riskier traders. Considering classic trading strategies, we also introduce an agent with a self-learning strategy. To model the behavior of such agent, we use deep reinforcement learning algorithms, namely Deep Deterministic policy gradient. Next, we develop an agent-based model with following strategies. With this model, we will be able to evaluate the main market statistics, named stylized-facts. Finally, we conduct a comparative analysis of results for constructed model with outcomes of previously proposed models, as well as with the characteristics of real market. As a result, we conclude that our model with a self-learning agent gives a better approximation to the real market than a model with classical agents. In particular, unlike the model with classical agents, the model with a self-learning agent turns out to be not so heavy-tailed. Thus, we demonstrate that for a complete understanding of market processes simulation models should take into account self-learning agents that have a significant presence at modern stock markets. [ABSTRACT FROM AUTHOR]

Using a previously validated agent-based model with fundamentalists and chartists, we investigate the usefulness and impact of direct market intervention. The policy maker diagnoses bubbles by forming an expectation of the future returns, then invests in burgeoning bubbles to develop a sufficient inventory of the risky asset in order to be able to sell adequate amounts of the overpriced asset later countercyclically to fight market exuberance. Preventing bubbles and crashes, this market intervention improves all analysed market return metrics, volatility, skewness, kurtosis and VaR, without affecting long-term growth. This increases the Sharpe ratios of noise traders and of fundamentalists by approximately 28% and 45% respectively. The results are robust even for substantially miscalibrated long-term expected returns. [ABSTRACT FROM AUTHOR]

The paper reexamines an agent-based model of opinion formation under bounded confidence with heterogeneous agents. The paper is novel in that it extends the standard model of opinion dynamics with the assumption that interacting agents share the desire to exchange opinion. In particular, the interaction between agents in the paper is modeled via a dynamic preferential-matching process wherein agents reveal their preferences to meet according to three features: coherence, opinion difference, and agents' positive sentiments towards others. Only preferred matches meet and exchange opinion. Through an extensive series of simulation treatments, it follows that the presence of sentiments, on one hand, hardens the matching process between agents, which leads to less communication. But, on the other hand, it increases the diversity in preferred matches between agents and thereby leads to a better-integrated social network structure, which reflects in a reduction of the opinion variance between agents. Moreover, at combinations of (a) high tolerance, (b) low sensitivity of agents to opinion volatility, and (c) low levels of confidence, agents are occasionally drawn away from the consensus, forming small groups that hold extreme opinions. [ABSTRACT FROM AUTHOR]

Contemporary debates about scientific institutions and practice feature many proposed reforms. Most of these require increased efforts from scientists. But how do scientists' incentives for effort interact? How can scientific institutions encourage scientists to invest effort in research? We explore these questions using a game-theoretic model of publication markets. We employ a base game between authors and reviewers, before assessing some of its tendencies by means of analysis and simulations. We compare how the effort expenditures of these groups interact in our model under a variety of settings, such as double-blind and open review systems. We make a number of findings, including that open review can increase the effort of authors in a range of circumstances and that these effects can manifest in a policy-relevant period of time. However, we find that open review's impact on authors' efforts is sensitive to the strength of several other influences. [ABSTRACT FROM AUTHOR]

The complexity science, with the help of, agent based modeling has recently claimed that the free market economy is producing wealth inequality—a totally opposite perception, which is in practice for more than 10 decades. It is capable of investigating the distribution of money in the market economy and finding solution for unfair gap between the wealth and, thus, overcoming poverty. This paper is an attempt to investigate this claim. This work is inspired from Gooding's work and it, therefore, reproduces his toy trader model, which claims to offer a fair trading environment. This model is extended towards a more human-oriented economy by introducing two possible biases: the variation of market visibility between the traders (by introducing the social networks of traders) and monopoly. The basic aim was to find out the impact of market visibility and monopoly on wealth distribution and poverty. It was learnt through simulations performed, in NetLogo, that the accessibility of traders, when taken as an important factor of a free market economy, positively influences wealth disparity, wealth gap and controls poverty. However, more human-oriented economy, in fact, widens the rich-poor divide and increases poverty. These results suggest that the authorities must control the market and everyone involved in trading should be given equal opportunities. This would also help in controlling the possible monopoly of the traders. The intrinsic physics of the free market would, otherwise, will always result in an unequal distribution of wealth. [ABSTRACT FROM AUTHOR]

We replicate the core model of the well-tested Keynes + Schumpeter agent-based model family, which features an endogenous innovation process in the evolutionary tradition based on invention and imitation. We introduce heterogeneous labor in the form of three different types of workers, representing different skill levels. In addition to a number of other stylized facts, which are reproduced by any Keynes + Schumpeter model, our version also generates wage inequality and labor market polarization due to skill-biased technological change. We introduce various labor market institutions and policies to our artificial economy in order to test, whether and how they affect inequality and polarization. Those policies, which alter relative wages induce an evolution of the technological development towards a lower demand for the relatively expensive type of worker. Policies and institutions that only aim at increasing the relative wages of low- and medium skilled workers therefore prove to be unable to combat inequality in the long run on their own. In order to be effective, those policies must be combined with educative measures that allow the workers to adapt to the changes in labor demand. Our findings have important implications on the design of real-world policies against inequality and polarization, since they shed light on potential unintended consequences of some of these policies. [ABSTRACT FROM AUTHOR]

Population growth and greater human well-being imply increased use of scarce resources, which makes innovative proposals for fair redistribution policies necessary. The utilisation of common-pool resources (CPRs) offers an economising solution to this problem; however, addressing the complex issue of harmonising the participants' interests is challenging. The use of CPRs has been theoretically discussed for decades and is addressed by theory of public goods (PGs) provision. Surprisingly, experimental tests of this theory have observed systematic deviations from the theoretical conclusions. Economists have applied agent-based methods to resolve this puzzle. We follow this approach and propose an agent-based model that innovatively employs voting on the punishment of free/cheap riders. The voting participants simultaneously determine the strictness of the punishment meted out to free/cheap riders and the cost burden of this punishment system. The proposed endogenous shared punishment model works within the threshold public goods game framework. Simulation results show that the presented method is more efficient than models with full or no punishment. Furthermore, the involvement of activists increases the likelihood of PGs provision. Introducing voting into PGs provision increases its efficiency and thus extends the set of available solutions to increase the efficiency of PGs provision. [ABSTRACT FROM AUTHOR]

We develop an agent-based computational model of epidemics with varying amounts of heterogeneity in regional vaccination coverage. We examine the severity of the resulting epidemics and the externalities created between high and low vaccination regions. The resulting size of the epidemics depends upon the heterogeneity of coverage, the interaction structure in the population, and how close the population is to herd immunity. Holding overall vaccination levels constant, the mean epidemic size increases and a larger level of vaccinations is needed to reach herd immunity when there is greater variability in the regional vaccination rates as long as the vaccination level is sufficiently high. This relationship amplifies when the contacts of individuals in the model have a greater share of inter-regional linkages. However, when inter-regional linkages and vaccination levels are sufficiently low, the mean epidemic size can decrease when there is greater variability in the regional vaccination rates. Further, we find that the magnitude of externalities between regions with different levels of vaccine coverage depends crucially on how close the population is to herd immunity. If population level vaccination coverage is sufficiently high (low) shifting connections from low coverage regions to high coverage regions will decrease (increase) mean epidemic size. [ABSTRACT FROM AUTHOR]

Big data technology enables us to access tremendous amounts of information; however, individuals cannot process all available information due to the bounded attention. The impact of this tension upon information seeking and processing behaviors and the resultant decision-making quality is still unclear. By agent-based simulation, we explicitly model the endogenous information choice in a sequential decision-making process, where individuals choose independently how much information and what type of information (shallow information such as the popularity a product, or deep information such as the expected utility of a product) is to be used. It is found that when the information is costly, only a small part of the individuals use deep information and only limited pieces of it, and other individuals simply follow the majority choice. The decrease in the cost of of information cost due to big data can encourage individuals to make use of more information, resulting in a better overall decision quality. However, if the big data only reduces the cost of shallow information but not that of the deep information, the decision quality is diminished because more individuals are induced to adopt the herding strategy. [ABSTRACT FROM AUTHOR]

We propose a simple model to simulate an interaction between banks and a financial market. In our model, banks are exposed to two sources of risks: market risk from their investments in assets external to the banking system and credit risk from lending in the interbank market. By and large, both risks increase during severe financial turmoil. In this scenario, the paper shows the conditions under which individual and the systemic defaults tend to coincide. This paper attempts to conduct a numerical simulation of banking ecosystems by using the actual values of financial items extracted from 89 Japanese banks' balance sheets. From this numerical simulation, we confirm two points: (1) when financial market prices decrease due to crashes in a trend-followers-dominant market, banks lose their net worth coincidentally. Thus, the capital adequacy ratio decreases synchronously, and any bank may not provide other banks with money through the interbank markets. (2) In a contrarians-dominant or contrarians-predominant market, we observed mean-reverting fluctuations in market prices. Bankruptcies happen asynchronously, and market prices eventually decrease. However, other banks may provide the bank suffering from a shortage of assets with money through the interbank markets. We further compare the characteristics of the banking system in four types of market modes. [ABSTRACT FROM AUTHOR]

Leverage trading, which consists of short selling and buying on margin, has been introduced into stock markets in many countries, including China. Ever since, there have been heated debates on how leverage trading influences financial markets. In this paper, an agent-based artificial market model is developed to simulate market behaviors and to analyze the influence of the leverage ratio on liquidity, volatility and price-discovery efficiency. In our artificial market, heterogeneous agents submit limit orders based on the fundamentalist or chartist strategy, and their effective supplies and demands can be increased by short selling or margin trading. Numerical analyses are performed in both one-sided and two-sided markets. We find that in one-sided markets, leverage trading can increase market liquidity and volatility, and decrease price-discovery efficiency. However, in the two-sided market, the increase of liquidity is much smaller, the volatility is decreased, and the price-discovery efficiency is improved. Generally, this model provides some meaningful results, which are supported by many other studies, and these findings underscore the necessity of building up a two-sided market when introducing leverage trading into stock markets. [ABSTRACT FROM AUTHOR]

A new Keynesian model built on an agent-based approach is considered and employed to investigate China's monetary policy and macroeconomic fluctuations. The assumption of perfect rationality used in standard dynamic stochastic general equilibrium (DSGE) models is abandoned. The expectation's heterogeneity, caused by agents behaving according to individual rules through adaptive learning, is one of the agent-based model (ABM) characteristics inserted into the DSGE model. Differential evolution (DE) algorithm is employed to estimate the parameters of an agent-based new Keynesian (ABNK) model, which combined the ABM and the new Keynesian DSGE models. The primary contribution of this study is that the degree of rationality in the economy has been estimated using a model with heterogeneous bounded rationality and adaptive learning. In addition, the determinacy properties of ABNK models with different degrees of heterogeneity are analyzed, which shows that the models that are determinate under the assumptions of rationality may become indeterminate in the presence of heterogeneous expectations. [ABSTRACT FROM AUTHOR]

Recently, new regional innovation policy paradigms emerged that transcend a long-lived dispute on whether either regional specialization, diversification, or rather related variety is most conducive to regional innovativeness. This includes ‘smart specialization’ in which regions are deliberately specialized and connected following technological relatedness, ‘gatekeepers’ in which there are pipelines between dense regional networks, and a ‘regional mix’ of knowledge that allows sustained ‘branching’. We develop and use an agent-based model to study the conjecture that, to stimulate (supra)regional innovativeness, social planners need to consider both, in conjunction, the mix of technological knowledge in regions and the (inter)regional innovation network topology. We use this agent-based model to evaluate the performance of and study the internal mechanisms of these new policy paradigms in a variety of scenarios. To increase the external validity of our findings, we calibrate the knowledge graph searched by the agents in our model to the OECD patent database. We confirm that access to related variety is important, yet that, on top of that, access to incidentally related knowledge is crucial to prevent high-level lock-in and ensure long term technological progress. Moreover, we find that networks with regional gatekeepers are particularly innovative, because these gatekeepers form ‘knowledge hubs’, create short paths to potential partners, and enlarge the total pool of knowledge. In case agents have few relationships, we find exceptionally high performance for the gatekeeper network in combination with regional diversification. The smart specialization network is a solid second option, although it lacks access to incidentally related knowledge and thus will ultimately fall behind whenever agents have relatively few relationships. The study elaborates on specific scenarios to reveal intricacies in the relationship between knowledge distribution, network topology, and the structure of interrelationships between knowledge fields. [ABSTRACT FROM AUTHOR]

We take the model of Alfarano et al. (J Econ Dyn Control 32:101-136, 2008) as a prototype agent-based model that allows reproducing the main stylized facts of financial returns. The model does so by combining fundamental news driven by Brownian motion with a minimalistic mechanism for generating boundedly rational sentiment dynamics. Since we can approximate the herding component among an ensemble of agents in the aggregate by a Langevin equation, we can either simulate the model in full at the micro level, or via an approximate aggregate law of motion. In the simplest version of our model, only three parameters need to be estimated. We explore the performance of a simulated method of moments (SMM) approach for the estimation of this model. As it turns out, sensible parameter estimates can only be obtained if one first provides a rough “mapping” of the objective function via an extensive grid search. Due to the high correlations of the estimated parameters, uninformed choices will often lead to a convergence to any one of a large number of local minima. We also find that the efficiency of SMM is relatively insensitive to the size of the simulated sample over a relatively large range of sample sizes and the SMM estimates converge to their GMM counterparts only for large sample sizes. We believe that this feature is due to the limited range of moments available in univariate asset pricing models, and that the sensitivity of the present model to the specification of the SMM estimator could carry over to many related agent-based models of financial markets as well as to similar diffusion processes in mathematical finance. [ABSTRACT FROM AUTHOR]

This work presents an artificial order-driven market able to reproduce mature and immature stock markets properties in the case of a single traded asset. This agent-based artificial market is designed to simulate characteristics of immature stock markets (high risk and low efficiency) by reproducing their stylized facts related mainly to information asymmetry and herd behavior. These two properties are modeled by combining social network and multi-agent simulations. The constructed scale-free social network, linking the modeled investors, gives rise to both informed and uninformed agents communities. Different assortative topologies are proposed and linked to different degrees of information asymmetry and market maturities. Several simulation experiments show that the modeled information asymmetry and herd behavior succeed in reproducing artificially some important stylized facts characterizing differences between immature and mature stock markets. [ABSTRACT FROM AUTHOR]

One explanation for large stock market fluctuations is its tendency to herd behavior. We put forward an agent-based model where instabilities are the result of liquidity imbalances amplified by local interactions through imitation, and calibrate the model to match some key statistics of actual daily returns. We show that an 'aggregate market-maker' type of liquidity injection is not successful in stabilizing prices due to the complex nature of the stock market. To offset liquidity shortages, we propose the use of locally triggered contrarian rules, and show that these mechanisms are effective in preventing extreme returns in our artificial stock market. [ABSTRACT FROM AUTHOR]

In this paper we introduce an agent-based model of a discretionary consumption sector in which demand is transformed by social emulation among consumers, thereby making producers adapt to demand. Our theoretical approach considers bounded rationality of agents (consumers and producers), heterogeneity of both agents and product characteristics, and the co-evolution of consumer desires, mainly, by social emulation. The proposed dynamics can reproduce some stylized facts that are well known in literature, such as the S-shaped adoption rate curve that many industries develop over their life cycle. Our model also obtains a novel result with relevant theoretical implications: the strictness of requirements, a factor rarely studied in consumer theory, has an important effect on some aggregate variables that are usually explained by the supply side, such as the number of producers or the industrial concentration index. In particular, the minimum number of producers (and then, the maximum Herfindahl index) is obtained for an intermediate degree of consumer requirements, a fact that is empirically validated for the wine market in Spain. [ABSTRACT FROM AUTHOR]

In this paper, we present a stochastic agent-based model and extend an artificial financial market by considering a stochastic process of market fundamentals. The model predicts that groups of noise traders are busy communicating when market uncertainty is high. In particular, we examine the effects of social interactions on price movements, based on parameter estimation of the group behavior. As traders' reactions to new information act much like an endogenous shock on return volatility, however, we cannot easily find an exact solution for the model with social interactions. Thus, simulation-based inference is used for the model validation; we investigate whether our artificial economy can match the empirical moments observed in five major foreign exchange data sets (as closely as possible). The results indicate that the return volatility under scrutiny can be robustly decomposed into news (45-55 %) and social interaction effects (45-55 %). [ABSTRACT FROM AUTHOR]

Auction sellers can use a reserve price to require a minimum bid before items are sold. Theoretical and experimental research has tested the influence of a reserve price in an independent private values auction, but little focus has been given to the influence of a reserve price in a first-price common-value auction. We establish an agent-based first-price common-value auction to determine the impact of the reserve price with two buyers and with three buyers. An agent-based approach to this problem is both a unique contribution to the literature and appropriate since finding analytical solutions with common-value auctions is difficult. The agent-based model approach also allows us to consider buyers that have non-symmetric bid functions. Furthermore, we introduce a combination of numerical integration techniques with a new particle swarm learning algorithm. The buyers in the model choose their expected-net-revenue-maximizing bid price, and sellers choose their expected-revenue-maximizing reserve price. In the two-buyer and three-buyer auction, a reserve price increases the equilibrium winning bid price, decreases the probability that the item is sold, and increases the seller's expected revenue. A reserve price in a two-buyer auction increases the winning bid price more than including an additional buyer in the auction with no reserve price. However, due to only receiving a salvage value when the item does not sell in the auction, the seller's expected revenue is higher in the three-buyer first-price common-value auction without a reserve price than in the two-buyer auction with a reserve price. [ABSTRACT FROM AUTHOR]

We develop an agent-based model of labor market regulation to study the consequences of employment protection legislations for labor market performance. Unlike most of the existing studies of labor market regulation we endogenize the institutional setting. Workers cast their vote on labor market regulation depending on the past payoffs that they accrued when one of two competing parties with different labor market policy platforms was in power. We identify important interactions with exogenous shocks. In more turbulent economic times, employment protection systems can affect labor market performance for some periods even after the shock has subsided. [ABSTRACT FROM AUTHOR]

In this report we present ASPEN, a new agent-based microeconomic simulation model of the U.S. economy being developed at Sandia National Laboratories (SNL). The model is notable because it allows a large number of individual economic agents to be modeled at a high level of detail and with a great degree of freedom. Some of the features of ASPEN are (a) a sophisticated message-passing system which allows individual pairs of agents to communicate with one another, (b) the use of genetic algorithms to simulate certain agents' learning, and (c) a detailed financial sector which includes a banking system and a bond market. Results from runs of the model are also presented. [ABSTRACT FROM AUTHOR]