An endogenous network model of banking systems

In this article, we propose an endogenous network model for banking systems. Starting from the balance sheets of a banking system and its update algorithms, the interbank credit lending was established based on the partner selection mechanism. Through simulation analysis, we find that the endogenous network model displays a structure with multiple liquidity centers, that the in‐degree distribution exhibits a Pareto tail, and that the bank asset distribution is a lognormal distribution with a Pareto tail. © 2016 Wiley Periodicals, Inc. Complexity 21: 512–520, 2016     

Credit spreads, endogenous bankruptcy and liquidity risk

In this paper, we consider a bond valuation model with both credit risk and liquidity risk to show that credit spreads are not negligible for short maturities. We adopt the structural approach to model credit risk, where the default triggering barrier is determined endogenously by maximizing equity value. As for liquidity risk, we assume that bondholders may encounter liquidity shocks during the lifetime of corporate bonds, and have to sell the bond immediately at the price, which is assumed to be a fraction of the price in a perfectly liquid market. Under this framework, we derive explicit expressions for corporate bond, firm value and bankruptcy trigger. Finally, numerical illustrations are presented.     

Classes of network connectivity and dynamics

Many kinds of complex systems exhibit characteristic patterns of temporal correlations that emerge as the result of functional interactions within a structured network. One such complex system is the brain, composed of numerous neuronal units linked by synaptic connections. The activity of these neuronal units gives rise to dynamic states that are characterized by specific patterns of neuronal activation and co‐activation. These patterns, called functional connectivity, are possible neural correlates of perceptual and cognitive processes. Which functional connectivity patterns arise depends on the anatomical structure of the underlying network, which in turn is modified by a broad range of activity‐dependent processes. Given this intricate relationship between structure and function, the question of how patterns of anatomical connectivity constrain or determine dynamical patterns is of considerable theoretical importance. The present study develops computational tools to analyze networks in terms of their structure and dynamics. We identify different classes of network, including networks that are characterized by high complexity. These highly complex networks have distinct structural characteristics such as clustered connectivity and short wiring length similar to those of large‐scale networks of the cerebral cortex. © 2002 Wiley Periodicals, Inc.     

Structure and dynamics of the global financial network

In this paper, we study the evolution of the network topology for the global financial market. We evaluate the level of diversification and participation of developed and emerging economies in cross-border exposures and find that the gross exposure network is dense, the vulnerability matrix is sparse, and the network’s fragility changes over time. Prior to the financial crisis in 2008, the network was relatively fragile, whereas it became more resilient afterwards, showing a reduction in financial institutions’ risk appetite. Our results suggest that financial regulators should track down the network evolution in their systemic risk assessment.     

Modeling hybrid network dynamics under random perturbations

We define stochastic network models based on internal structural dynamics of individual nodes which undergo Markovian switching subject to noise and external shocks. Steady-state stability and reliability criteria are established in closed form and show explicit dependence on system parameters.     

内生网络环境下2-步邻域内的局部策略互动及其仿真

考察内生网络环境下局中人与2-步邻域内的邻居进行的局部协同对策,较为完整地给出了均衡网络的结构特性,以及费用参数和互动半径对于均衡结构的影响. 基于 NetLogo仿真系统,编制了局部互动仿真模拟实验程序. 仿真结果显示,网络生成的动态进程对于网络均衡结果存在很大影响. 结果对于解决社会和经济领域中的互动问题可提供策略性指导.     

内生网络环境下具有异质连接费用的局部策略互动

考察内生网络环境下局中人之间的局部策略互动, 网络中的局中人只与直接邻居进行协同对策. 网络生成的过程中, 建立连接的费用是异质的~(具有两种水平), 与采取有效行动的局中人建立连接时执行高水平费用, 与采取风险占优行动的局中人建立连接时执行低水平费用. 在异质连接费用的情形下, 首次较为完整地给出了均衡网络的结构特性和局中人的行动选择, 并分析了费用参数对均衡结果的影响.     

Filtering for risk assessment of interbank network

Our paper contributes to the recent macroprudential policy addressing the resilience of financial systems in terms of their interconnectedness. We argue that beneath an interbank market, there is a fundamental latent network that affects the liquidity distributions among banks. To investigate the interbank market, we propose a framework that identifies such latent network using a statistical learning procedure. The framework reverse engineers overnight signals observed as banks conduct their reserve management on a daily basis. Our simulation-based results show that possible disruptions in funds supply are highly affected by the interconnectedness of the latent network. Hence, the proposed framework serves as an early warning system for regulators to monitor the overnight market and to detect ex-ante possible disruptions based on the inherent network characteristics.     

Learning Bayesian network classifiers by risk minimization

Bayesian networks (BNs) provide a powerful graphical model for encoding the probabilistic relationships among a set of variables, and hence can naturally be used for classification. However, Bayesian network classifiers (BNCs) learned in the common way using likelihood scores usually tend to achieve only mediocre classification accuracy because these scores are less specific to classification, but rather suit a general inference problem. We propose risk minimization by cross validation (RMCV) using the 0/1 loss function, which is a classification-oriented score for unrestricted BNCs. RMCV is an extension of classification-oriented scores commonly used in learning restricted BNCs and non-BN classifiers. Using small real and synthetic problems, allowing for learning all possible graphs, we empirically demonstrate RMCV superiority to marginal and class-conditional likelihood-based scores with respect to classification accuracy. Experiments using twenty-two real-world datasets show that BNCs learned using an RMCV-based algorithm significantly outperform the naive Bayesian classifier (NBC), tree augmented NBC (TAN), and other BNCs learned using marginal or conditional likelihood scores and are on par with non-BN state of the art classifiers, such as support vector machine, neural network, and classification tree. These experiments also show that an optimized version of RMCV is faster than all unrestricted BNCs and comparable with the neural network with respect to run-time. The main conclusion from our experiments is that unrestricted BNCs, when learned properly, can be a good alternative to restricted BNCs and traditional machine-learning classifiers with respect to both accuracy and efficiency.     

Identification of nonlinear dynamics using a general spatio-temporal network

The so-called spatio-temporal neural network is considered. This is a neural network where the conventional weight multiplication operation is replaced by a linear filtering operation. General learning algorithms are derived for such a network, both in the discrete-time and in the continuous-time domains. The problem of deterministic nonlinear system identification is considered as an application of spatio-temporal neural networks. Nonlinear system identification is one of the challenging problems in the field of dynamic systems, with limited successful results using conventional methods. Neural network approaches have so far been encouraging, but further exploration is needed. The capabilities of the derived algorithms and of the considered architectures to effectively identify deterministic nonlinear systems is demonstrated through examples.     

Complex dynamics in an OLG model of neoclassical growth with endogenous retirement age and public pensions

This study analyses the dynamics of a general equilibrium, overlapping-generations (closed) economy with pay-as-you-go public pensions and tax-financed health investments that affect the retirement time when old. Life of the typical agent is divided between youth (firth period) and old age (second period). The latter period of life is, in turn, divided between work time and retirement time in a proportion contingent on an individual’s state of health. We show that: (i) a unique non-trivial steady state exists, and (ii) when the labour income tax rates used to finance health expenditure or public pensions are included in an intermediate range of values, complex dynamics occur when individuals have perfect foresight. This holds because the increase either in the fraction of time spent working when old or disability pensions reduces savings and capital accumulation. In addition, dynamic phenomena such as multiple bubbling structures related to the bifurcation diagram can be observed. Under some general assumptions, we show that the rise in health care expenditure and/or public pensions initially destabilises the steady-state equilibrium and causes complex dynamics but eventually acts as a stabilising device.     

Small world network models of the dynamics of HIV infection

It has long been recognised that the structure of social networks plays an important role in the dynamics of disease propagation. The spread of HIV results from a complex network of social interactions and other factors related to culture, sexual behaviour, demography, geography and disease characteristics, as well as the availability, accessibility and delivery of healthcare. The small world phenomenon has recently been used for representing social network interactions. It states that, given some random connections, the degrees of separation between any two individuals within a population can be very small. In this paper we present a discrete event simulation model which uses a variant of the small world network model to represent social interactions and the sexual transmission of HIV within a population. We use the model to demonstrate the importance of the choice of topology and initial distribution of infection, and capture the direct and non-linear relationship between the probability of a casual partnership (small world randomness parameter) and the spread of HIV. Finally, we illustrate the use of our model for the evaluation of interventions such as the promotion of safer sex and introduction of a vaccine.     

Modulation of synchronization dynamics in a network of self-sustained systems

This paper addresses the combined modulatory effects of non-nearest neighbor oscillators and local injection on synchronized states dynamics with their corresponding stability boundaries in a network of self-sustained systems. The Whittaker method and Floquet theory are used to predict analytically the stability of these states for identical and non-identical coupling parameters. Charts revealing the modulation of synchronized states and their stability boundaries at the second order of interaction in the cases of identical and non-identical coupling parameters are constructed with and without an external signal locally injected in the network. Numerical simulations validate and complement the results of analytical surveys. The limits of the stability regions are numerically explored when a small amount of Gaussian white noise is also injected in the network.     

Reverse logistics network design and planning utilizing conditional value at risk

Nowadays, due to some social, legal, and economical reasons, dealing with reverse supply chain is an unavoidable issue in many industries. Besides, regarding real-world volatile parameters, lead us to use stochastic optimization techniques. In location–allocation type of problems (such as the presented design and planning one), two-stage stochastic optimization techniques are the most appropriate and popular approaches. Nevertheless, traditional two-stage stochastic programming is risk neutral, which considers the expectation of random variables in its objective function. In this paper, a risk-averse two-stage stochastic programming approach is considered in order to design and planning a reverse supply chain network. We specify the conditional value at risk (CVaR) as a risk evaluator, which is a linear, convex, and mathematically well-behaved type of risk measure. We first consider return amounts and prices of second products as two stochastic parameters. Then, the optimum point is achieved in a two-stage stochastic structure regarding a mean-risk (mean-CVaR) objective function. Appropriate numerical examples are designed, and solved in order to compare the classical versus the proposed approach. We comprehensively discuss about the effectiveness of incorporating a risk measure in a two-stage stochastic model. The results prove the capabilities and acceptability of the developed risk-averse approach and the affects of risk parameters in the model behavior.     

Simulating influenza pandemic dynamics with public risk communication and individual responsive behavior

Individual responsive behavior to an influenza pandemic has significant impacts on the spread dynamics of this epidemic. Current influenza modeling efforts considering responsive behavior either oversimplify the process and may underestimate pandemic impacts, or make other problematic assumptions and are therefore constrained in utility. This study develops an agent-based model for pandemic simulation, and incorporates individual responsive behavior in the model based on public risk communication literature. The resultant model captures the stochastic nature of epidemic spread process, and constructs a realistic picture of individual reaction process and responsive behavior to pandemic situations. The model is then applied to simulate the spread dynamics of 2009 H1N1 influenza in a medium-size community in Arizona. Simulation results illustrate and compare the spread timeline and scale of this pandemic influenza, without and with the presence of pubic risk communication and individual responsive behavior. Sensitivity analysis sheds some lights on the influence of different communication strategies on pandemic impacts. Those findings contribute to effective pandemic planning and containment, particularly at the beginning of an outbreak.     

Cluster synchronization in colored community network with different order node dynamics

In this paper, a colored community network model with different order nodes dynamics is introduced for the first time. The color of nodes or edges are assumed to be identical if they belong to the same community and nonidentical if they belong to different communities. The color of edges between any pair of communities are assumed to be identical if they connect the same pair of communities and nonidentical if they connect different pair of communities. Further, the order of the node dynamics in different communities can be totally different. Firstly, based on the Lyapunov stability theory, adaptive feedback controllers are designed for achieving cluster synchronization. Secondly, periodically intermittent controllers are designed for achieving cluster synchronization and the synchronization conditions are derived by using mathematical induction method and the analysis technique. Finally, several numerical examples are provided to illustrate the effectiveness of the derived theoretical results.     

Examining undergraduate student retention in mathematics using network analysis and relative risk

Higher education faces challenges in retaining students who require a command of numeracy in their chosen field of study. This study applies an innovative combination of relative risk and social network analysis to enrolment data of a single cohort of commencing students from an Australian regional university. Relative risk, often used in epidemiology studies, is used to strategically investigate whether first year mathematics subjects at the university demonstrated a higher risk of attrition when compared to other subjects offered in the first year of study. The network analysis is used to illustrate the connections of those mathematics subjects, identifying service subjects through their multiple connections. The analysis revealed that attrition rates for eight of the nine subjects were within acceptable limits, and this included identified service subjects. The exception highlighted the issue of mathematics competencies in this cohort. This combined analytical technique is proposed as appropriate for use when investigating attrition and retention at faculty and institutional levels, including the determination of levels of intervention and support for any subject.     

The network structure and systemic risk in the global non-life insurance market

This paper contributes to the literature on systemic risk by assessing the systemic importance of insurers in the global non-life insurance market. First, we estimate the bilateral reinsurance claims matrix using the aggregate outstanding reinsurance data from ISIS and theoretically analyze the interconnectedness in the global reinsurance network using network indicators. The robustness of the estimated matrix is fully assured by sensitivity analysis. Second, we theoretically analyze the contagious defaults introducing the Eisenberg–Noe framework. Reinsurers play a dominant role in the reinsurance network and most of them are included in our data sample. The network analysis finds that some reinsurers with large centrality measures are central in the hierarchical structure of the network. The default analysis shows the occurrences of many stand-alone defaults and only one contagious default via the global reinsurance network after the global financial crisis. In addition, one stress test based on a hypothetical severe stress scenario predicts a few occurrences of contagious defaults in the future. It follows from these analyses that systemic risk via the global reinsurance network is relatively restricted in the global non-life insurance market. In conclusion, our methodology would help supervisory authorities develop an assessment approach for interconnectedness in the global reinsurance network and aid the implementation of insurer stress tests for default contagion.