Emergence of anti-coordination through reinforcement learning in generalized minority games

In this paper we propose adaptive strategies to solve coordination failures in a prototype generalized minority game model with a multi-agent, multi-choice environment. We illustrate the model with an application to large scale distributed processing systems with a large number of agents and servers. In our set up, agents are assigned responsibility to complete tasks that require unit time. They request servers to process these tasks. Servers can process only one task at a time. Agents have to choose servers independently and simultaneously, and have access to the outcomes of their own past requests only. Coordination failure occurs if more than one agent simultaneously requests the same server to process tasks at the same time, while other servers remain idle. Since agents are independent, this leads to multiple coordination failures. In this paper, we propose strategies based on reinforcement learning that minimize such coordination failures. We also prove a null result that a large category of probabilistic strategies which attempts to combine information about other agents’ strategies, asymptotically converge to uniformly random choices over the servers.

     

Arbitrage and price revelation with asymmetric information and incomplete markets

This paper deals with the issue of arbitrage with differential information and incomplete financial markets, with a focus on information that no-arbitrage asset prices can reveal. Time and uncertainty are represented by two periods and a finite set S of states of nature, one of which will prevail at the second period. Agents may operate limited financial transfers across periods and states via finitely many nominal assets. Each agent i has a private information about which state will prevail at the second period; this information is represented by a subset S_i of S. Agents receive no wrong information in the sense that the “true state” belongs to the “pooled information” set ∩_iS_i, hence assumed to be non-empty.Our analysis is two-fold. We first extend the classical symmetric information analysis to the asymmetric setting, via a concept of no-arbitrage price. Second, we study how such no-arbitrage prices convey information to agents in a decentralized way. The main difference between the symmetric and the asymmetric settings stems from the fact that a classical no-arbitrage asset price (common to every agent) always exists in the first case, but no longer in the asymmetric one, thus allowing arbitrage opportunities. This is the main reason why agents may need to refine their information up to an information structure which precludes arbitrage.     

The stock–bond comovements and cross-market trading

We propose an asset pricing model with heterogeneous agents allocating capital to the stock and bond markets to optimize their portfolios, utilizing the dynamic interaction between the two markets. While some agents focus on the stock market and have more expertise in it, the others specialize in the bond market. Based on their comparative advantages in a particular market, heterogeneous agents constantly revise their investment portfolios by taking into account the time-varying stock–bond return comovements and the changing market conditions. Agents׳ collective investment behavior shapes the stock–bond interlinkage, which feedbacks on their subsequent capital allocations. Using monthly US stock and bond data from January 1990 to June 2014, we estimate the vector autoregression model with threshold and Markov switching mechanisms. We find evidence in support of flight-to-quality and show that it is mainly driven by the technical traders who actively sell stocks and buy bonds during periods of high market uncertainty.     

Heterogeneity in social values and capital accumulation in a changing world

In a society characterized by a multitude of heterogeneous agents and a large number of possibly immaterial goods, each one having distinct social and personal values, we study the impact of these relative values on intergenerational capital accumulation, as a function of economic and social parameters such as capital mobility, productivity and personal and social values discrepancies. Each agent is modelled by a one-period production function and a two-period intertemporal utility. Agents live, produce and consume over one period, but optimize over two periods, so providing a remaining stock of goods for the next generation. This creates a dynamics in capital accumulation depending on social and individual values. A threshold appears in capital stock accumulation that depends on personal and social values’ volatilities, and below which the initial stock will be depleted. Whereas volatility in social values increases the threshold, impairing capital accumulation, adverse shocks in goods’ values may reverse the dynamics of the accumulation process. Finally, capital mobility specifically favors forerunners, but capital accumulation in one or several sectors may shift social values in their direction, at the expense of other sectors.

     

An evaluation of positive psychological state of educational administrators in a tolerance context

Positive psychology leads the problems with regard to subject well-being, hopefullness, motivation, hope and academic success of individuals from their birth to their death. The purpose of this study is to assess the obtained opinions regarding positive psychological state of education directors. Being a qualitative case study, the integrative single statement design was used in the study. The study group consisted of 10 administrators and 25 instructors working in state schools in Northern Cyprus during 2016–2017 academic year in this study. An interview form which was consisted of 5 questions was used in collecting data. The qualitative data analysis program NVIVO 11.0 was used during interviews for classifying and coding data. The opinions of school administrators regarding their positive psychological states was assessed and determined.

     

From constrained optimization to constrained dynamics: extending analogies between economics and mechanics

Economic equilibrium models have been inspired by analogies to stationary states in classical mechanics. To extend these mathematical analogies from constrained optimization to constrained dynamics, we formalize economic (constraint) forces and economic power in analogy to physical (constraint) forces and the reciprocal value of mass. Agents employ forces to change economic variables according to their desire and their power to assert their interest. These ex-ante forces are completed by constraint forces from unanticipated system constraints to yield the ex-post dynamics. The differential-algebraic equation framework seeks to overcome some restrictions inherent to the optimization approach and to provide an out-of-equilibrium foundation for general equilibrium models. We transform a static Edgeworth box exchange model into a dynamic model with procedural rationality (gradient climbing) and slow price adaptation, and discuss advantages, caveats, and possible extensions of the modeling framework.

     

Stochastic processes of temporary equilibria

It is the aim of the paper to study within the framework of an ‘overlapping generation model’ the evolution of temporary equilibria. At date t, there are ‘newborn’ agents and ‘old’ agents who were born in previous periods; the old agents hold cash balances (fiat money) that they carried over from the previous period. At the beginning of period t, all agents receive a random endowment of consumption goods. Then the agents exchange these endowments and money on spot markets at date t (trading in future markets is not considered). Once a temporary equilibrium is reached, the economy move to the next date. Agents who were born at date t then become old and meet agents born at period t+1.It is shown that the evolution of temporary equilibria in this model leads to analyse the ergodic properties of a certain class of Markov processes with stationary transition probability.     

Artificial stock markets with different maturity levels: simulation of information asymmetry and herd behavior using agent-based and network models

This paper aims mainly at building artificial stock markets with different maturity levels by modeling information asymmetry and herd behavior. The developed artificial markets are multi-assets, order-driven and populated by agents having heterogeneous behaviors and information. Agents are defined by their information and their herd behavior levels. Agents trade multiple risky assets based on their wealth, their behaviors and their available information which spread among multiple behavioral networks. In a novel contribution to artificial stock markets literature, agents’ behaviors modeling is mixed with social network simulation to reproduce different degrees of information asymmetry and herd behavior based on several assortative topologies. Several simulations validated the proposed model since univariate and multivariate stylized facts were reproduced both for mature and immature stock markets. The proposed artificial stock market can be considered as a first step toward decision and simulation tools for optimal management, strategy analysis and predictions evolution of immature stock markets.     

The dual attitudinal dynamics of public opinion: an agent-based reformulation of L. F. Richardson’s war-moods model

This study proposes a new model of public opinion dynamics, focused specifically on attitudes toward war. The model consists of citizen-agents who have dual (public and private) attitudes. While they change the two levels of attitude according to different rules of two-layered interactions with neighboring agents, they also change their own attitudes internally. Our model shows that public opinion has “meta-stable” states, which generate overlapping, or two-valued, stable states. This characteristic can explain how public opinion responds to the outbreak and continuation of war very differently, not just in our model, but also in the opinions of European citizens on the First World War and those of American citizens on U.S. wars after the Second World War.     

Empirical properties of a heterogeneous agent model in large dimensions

We introduce a variant of the Adaptive Beliefs System (ABS) of Brock and Hommes (1998) based on returns instead of prices. Agents form their demands according to the degree to which they are trend-following or contrarian. Empirically, the model requires that agents’ demands be coerced by leverage constraints. Using five samples of US stock returns, we show that the fit to realized returns is essentially driven by the total dispersion of the model’s returns. We also find that the latter are more realistic when forecasts are based on short-term estimates and when trend-followers and contrarians have the same ex-ante importance. We then provide evidence that the model is able to mimic most stylized facts observed on financial markets (tail decay, volatility clustering and autocorrelation patterns) quite closely. Finally, we find that portfolio policies designed according to the model’s predictions outperform the naive 1/N portfolio out-of-sample by 2% per annum.     

Political Connections and Accounting Quality under High Expropriation Risk

Abstract

We examine the impact of political connections and accounting quality among Venezuelan industrial firms, which face one of the highest levels of expropriation risk worldwide. Based on prior literature, we expect a negative relationship between expropriation risk and accounting quality as firms manage earnings to avoid ‘benign’ state intervention. We find that politically connected firms have higher accounting quality than non-connected firms, which is consistent with connected firms’ lower risk of expropriation due to connections with high-level government officials or ruling party members. The relationship between accounting quality and political connections appears to be strongly moderated by institutional features like expropriation risk.     

Ex-post favoring ranks: a fairness notion for the random assignment problem

We introduce two notions of ex-post fairness, namely ex-post favoring ranks (EFR) and robust ex-post favoring ranks, which consider whether objects are received by those agents who have the highest rank for them. We examine their compatibility with standard properties of random assignments and state some impossibility theorems. We also propose and formalize a revised version of the Boston mechanism and prove that it provides an EFR random assignment.

     

Networks and market-based measures of systemic risk: the European banking system in the aftermath of the financial crisis

This article investigates the behaviour of the European banking system during the financial crises that occurred in the last decades. Among the various approaches for measuring systemic risk, we consider network analysis, which describes the linkages among financial institutions and their whole structure. We construct a time-varying network of the European banking system. Banks are linked to form a global interconnected system and they mutually influence one another in terms of risk. We model their reciprocal influence via a weighted and directed network, in which weights are related to risk measures that are based on equity returns. Then, we apply two network indicators to investigate the prominence of a bank in spreading and receiving risk from the others. The results enable us to capture many features of the banking system while identifying the global systemically important banks. Moreover, the results of the analysis over time show how interconnections change over periods that are characterized by various economic scenarios.

     

Social networks,mass media and public opinions

Our opinions and ideas are shaped by what our friends said and what we read or watched on mass media. In this paper, we propose a concise and analyzable model to study the effects of mass media modeled as an applied external field, and social networks on public opinions based on the multi-state voter model, and a tuned parameter can control the relative intensity of the effects of mass media and social networks. We consider a generalized scenario where there exist committed or stubborn agents in the networks whose opinions are not affected by their friends or mass media. We find that the fraction of each opinion will converge to a value which only relates to the fractions and degrees of stubborn agents, and the relative intensity between media and network effects. The final agents with media opinion, except the stubborn agents, also include the increment produced by the internal impact of social networks and that caused by the external impact of media. Interestingly the second increment is composed of two parts, one is from the media effect when there are no interactions between agents and the other is from the influence of media on agent opinions caused by social network structure. That is the interactions among agents within social networks can amplify media influence. Finally we also discuss several extensions to the dynamics model which consider more realistic scenarios.     

CREDIBLY IDENTIFYING SOCIAL EFFECTS: ACCOUNTING FOR NETWORK FORMATION AND MEASUREMENT ERROR

Understanding whether and how connections between agents (networks) such as declared friendships in classrooms, transactions between firms, and extended family connections, influence their socio‐economic outcomes has been a growing area of research within economics. Early methods developed to identify these social effects assumed that networks had formed exogenously, and were perfectly observed, both of which are unlikely to hold in practice. A more recent literature, both within economics and in other disciplines, develops methods that relax these assumptions. This paper reviews that literature. It starts by providing a general econometric framework for linear models of social effects, and illustrates how network endogeneity and missing data on the network complicate identification of social effects. Thereafter, it discusses methods for overcoming the problems caused by endogenous formation of networks. Finally, it outlines the stark consequences of missing data on measures of the network, and regression parameters, before describing potential solutions.     

Policy change and learning in the RBC model

What is the impact of surprise and anticipated policy changes when agents form expectations using adaptive learning rather than rational expectations? We examine this issue using the standard stochastic real business cycle model with lump-sum taxes. Agents combine knowledge about future policy with econometric forecasts of future wages and interest rates. Dynamics under learning can have large impact effects and a gradual hump-shaped response, and tend to be prominently characterized by oscillations not present under rational expectations. These fluctuations reflect periods of excessive optimism or pessimism, followed by subsequent corrections.     

Incorporating Transportation Network Structure in Spatial Econometric Models of Commodity Flows

Abstract

We use a spatial econometric extension of the traditional regression-based gravity model to model commodity flows, focusing on a formal methodology for incorporating information regarding the highway network into the spatial connectivity structure of the spatial autoregressive econometric model. We show that our simple approach to incorporating this information in the model produces improved model fit and higher likelihood function values. Empirical estimates of the relative importance of the different types of origin–destination connectivity between regions indicates that the strongest spatial autoregressive effects arise when both origin and destination regions have neighbouring regions located on the highway network.     

Opinion dynamics and wisdom under conformity

We study a dynamic model of opinion formation in social networks. In our model, boundedly rational agents update opinions by averaging over their neighbors’ expressed opinions, but may misrepresent their own opinion by conforming or counter-conforming with their neighbors. We show that an agent׳s social influence on the long-run group opinion is increasing in network centrality and decreasing in conformity. Concerning efficiency of information aggregation or “wisdom” of the society, it turns out that misrepresentation of opinions need not undermine wisdom, but may even enhance it. Given the network, we provide the optimal distribution of conformity levels in the society and show which agents should be more conforming in order to increase wisdom.     

Sulla vitalità di un mercato finanziario

Questo lavoro discute un modello di mercato finanziario basato sulla teoria della vitalità associata allo studio delle inclusioni differenziali. Il modello permette di esplicitare l'esistenza di una relazione di dipendenza tra le quantità di un titolo detenute dagli agenti ed il suo prezzo di mercato. Basandosi su questa, è possibile ottenere rigorosamente l'esistenza di un sentiero temporale per il prezzo senza bisogno di introdurre ipotesiad hoc.

---

Summary This work discusses a model of a financial market based on viability theory and differential inclusions. The model makes it explicit the existence of a relationship between the quantities of a security held by agents and its market price. Based on this, we show how to establish rigorously the existence of a time path for the price which in the literature is usually assumed without much justification.

---

---

Ringraziamo E. Castagnoli per alcuni utili commenti. Le conversazioni con J.-P. Aubin, F. Moriconi e F. Ortu sull'etimologia e l'interpretazione dei concetti di vitalità ci sono state di grande beneficio.