On the convergence of genetic learning in a double auction market

We study the learning behavior of a population of buyers and a population of sellers whose members are repeatedly randomly matched to engage in a sealed bid double auction. The agents are assumed to be boundedly rational and choose their strategies by imitating successful behavior and adding innovations triggered by random errors or communication with other agents. This process is modelled by a two-population genetic algorithm. A general characterization of the equilibria in mixed population distributions is given and it is shown analytically that only one price equilibria are attractive for the GA dynamics. Simulation results confirm these findings and imply that in cases with random initialization with high probability the gain of trade is equally split between buyers and sellers.     

Common-value auction versus posted-price selling: an agent-based model approach

Previous theoretical work has compared a private-value auction and posted-price market, and an affiliated-value auction and a posted-price market to determine the selling method preferred by sellers. Much less, however, is known about the seller’s preferred selling method when the buyers have a common value of the item. Our objective is to determine if a first-price auction or a posted-price market provides a seller with the larger expected revenue when buyers have a common value of the item being sold. An agent-based posted-price market and an agent-based first-price common-value auction with a reserve price are developed to compare these selling methods. Holding the buyers’ uncertainty about the value of the item constant, the seller prefers the posted-price market when the seller has no uncertainty about the item’s value. When the seller has an equal level of uncertainty as the buyers, the seller’s expected revenue for each market is similar. As the seller’s uncertainty increases beyond the level of the buyers’ uncertainty, the auction with a reserve price eventually becomes the preferred choice.     

Adaptive learning,endogenous uncertainty,and asymmetric dynamics

I present a simple model where forecasting confidence affects aggregate demand. It is shown that this model has similar stability properties, under statistical and evolutionary learning, as a model without a confidence affect. From this setup, I introduce “Expectational Business Cycles” where output fluctuates due to learning, heterogeneous forecasting models and random changes in the efficient forecasting model. Agents use one of two forecasting models to forecast future variables while heterogeneity is dictated via an evolutionary process. Increased uncertainty, due to a shock to the structure of the economy, may result in a sudden decrease in output. As agents learn the equilibrium, output slowly increases to its equilibrium value. Expectational business cycles tend to arrive faster, last longer and are more severe as agents possess less information.     

Foreign direct investment: A genetic algorithm approach

This paper proposes a genetic algorithm (GA) approach as an analytical tool with a carefully defined fitness function for variable selection. Discriminant analysis will be used as a parameter evaluation method for the analysis of inward foreign direct investment (FDI) in China. Results indicate that the proposed GA method is more efficient in classifying “successful or unsuccessful” inward FDI by providing higher accuracy rates while using fewer variables than previous efforts. An implication of this result is that, given a scarcity of resources and the need to promote FDI, the proposed GA may provide a more efficient way to concentrate on those fewer variables found to be important determinants of “successful” FDI inflow.     

Pricing the cloud: a QoS-based auction approach

ABSTRACT

Cloud computing is an evolutionary technology that offers on-demand resources and elastic services through the Internet. Most providers adopt fixed-price mechanisms (e.g. pay-as-you-go). However, a few providers have recently employed auction-like approaches to price cloud services. Meanwhile, cloud consumers pay more attention to Quality of Service (QoS) such as availability, which measures how well a service is performed. This paper proposes a novel auction approach that can efficiently allocate resources according to customers’ QoS preferences. The QoS-based pricing can generate more revenue than a fixed-price strategy. This research lies at the intersection of cloud computing, economics, and information systems.     

Adaptive learning and distributional dynamics in an incomplete markets model

Recent research shows that several DSGE models provide a closer fit to the data under adaptive learning. This paper extends this research by introducing adaptive learning in the model of Krusell and Smith (1998) with uninsurable idiosyncratic risks and aggregate uncertainty. A first contribution of this paper establishes that the equilibrium of this framework is stable under least-squares learning. The second contribution consists of showing that bounded rationality enhances the ability of this model to match the distribution of income in the US. Learning increases significantly the Gini coefficients because of the opposite effects on consumption of the capital-rich and of the capital-poor agent. The third contribution is an empirical exercise that shows that learning can account for increases in the income Gini coefficient of up to 25% in a period of 28 years. Overall, these findings suggest that adaptive learning has important distributional repercussions in this class of models.     

A genetic algorithm based approach to transactional and QoS-aware service selection

As there are various risks of failure when Web services are deployed in unreliable environment, the execution of a composite Web service (CWS) requires the transaction mechanism to guarantee its reliable execution. However, the existing service selection approaches consider QoS and transaction separately and have not considered that transactional properties may affect the QoS such as the execution time of a CWS. This work addresses the importance of considering transactional properties and how the transactional properties affect the QoS simultaneously in the process of service selection by a QoS-aware and transactional-aware selection approach. First, a performance evaluation method is proposed to calculate the execution time of a transactional CWS. Then, a genetic algorithm based approach, which takes into account the execution time, price, transactional property and successful execution rate of CWS, is presented to achieve global optimisation service selection. Finally, experimental results show the efficiency and effectiveness of the approach.     

An options-based approach to analyze auction guarantees in the art market

Auction-house guarantees are becoming a common feature in the art market. We analyze these guarantees within the framework of financial options. This approach allows us to derive analytical (closed-form) expressions to value these positions, considering both, the case in which the painting is sold, and the case in which the painting goes unsold (“bought in”). In addition, we present several risk metrics that are useful to describe from an intuitive viewpoint the exposure of the auction house, and that of a third party (in case the auction house decides to layoff, fully or partially, the risk associated with offering such guarantees). We demonstrate that the expressions we derive satisfy the put-call parity relationship, and we further validate these formulas with a Monte Carlo simulation applied to a realistic example. We also show that the risk associated with such guarantees is lower than what is commonly believed by market practitioners, and we expose the dangers of relying on the Black-Scholes model to value such guarantees. Finally, having explicit expressions to assess the risk involved in these guarantees helps to bring more transparency to a notoriously opaque segment of the art market.     

Credit risk and asymmetric information: A simplified approach

We present a simple model for risky, corporate debt. Debtholders and equityholders have incomplete information about the financial state of the debt issuing company. Information is incomplete because it is delayed for all agents, and it is asymmetrically distributed between debtholders and equityholders. We solve for the equityholders' optimal default policy and for the credit spreads required by debtholders. Delayed information accelerates the equityholders' optimal decision to default. Interestingly, this effect is small, implying only a small impact on credit spreads. Asymmetric information, however, has a major impact on credit spreads. Our model predicts high credit spreads for short-term debt, as observed empirically in credit markets.     

Efficient learning via simulation: A marginalized resample-move approach

In state–space models, parameter learning is practically difficult and is still an open issue. This paper proposes an efficient simulation-based parameter learning method. First, the approach breaks up the interdependence of the hidden states and the static parameters by marginalizing out the states using a particle filter. Second, it applies a Bayesian resample-move approach to this marginalized system. The methodology is generic and needs little design effort. Different from batch estimation methods, it provides posterior quantities necessary for full sequential inference and recursive model monitoring. The algorithm is implemented both on simulated data in a linear Gaussian model for illustration and comparison and on real data in a Lévy jump stochastic volatility model and a structural credit risk model.     

Adaptive learning forecasting,with applications in forecasting agricultural prices

We introduce a new forecasting methodology, referred to as adaptive learning forecasting, that allows for both forecast averaging and forecast error learning. We analyze its theoretical properties and demonstrate that it provides a priori MSE improvements under certain conditions. The learning rate based on past forecast errors is shown to be non-linear. This methodology is of wide applicability and can provide MSE improvements even for the simplest benchmark models. We illustrate the method’s application using data on agricultural prices for several agricultural products, as well as on real GDP growth for several of the corresponding countries. The time series of agricultural prices are short and show an irregular cyclicality that can be linked to economic performance and productivity, and we consider a variety of forecasting models, both univariate and bivariate, that are linked to output and productivity. Our results support both the efficacy of the new method and the forecastability of agricultural prices.     

A genetic algorithm for the partial binary constraint satisfaction problem: an application to a frequency assignment problem

We describe a genetic algorithm for the partial constraint satisfaction problem. The typical elements of a genetic algorithm, selection, mutation and cross-over, are filled in with combinatorial ideas. For instance, cross-over of two solutions is performed by taking the one or two domain elements in the solutions of each of the variables as the complete domain of the variable. Then a branch-and-bound method is used for solving this small instance. When tested on a class of frequency assignment problems this genetic algorithm produced the best known solutions for all test problems. This feeds the idea that combinatorial ideas may well be useful in genetic algorithms.     

Binary quantile regression: a Bayesian approach based on the asymmetric Laplace distribution

This paper develops a Bayesian method for quantile regression for dichotomous response data. The frequentist approach to this type of regression has proven problematic in both optimizing the objective function and making inferences on the parameters. By accepting additional distributional assumptions on the error terms, the Bayesian method proposed sets the problem in a parametric framework in which these problems are avoided. To test the applicability of the method, we ran two Monte Carlo experiments and applied it to Horowitz's (1993) often studied work‐trip mode choice dataset. Compared to previous estimates for the latter dataset, the method proposed leads to a different economic interpretation. Copyright © 2010 John Wiley & Sons, Ltd.     

Comovements and asymmetric tail dependence in state housing prices in the USA: A nonparametric approach

We reexamine the methods used in estimating comovements among US regional home prices and find that there are insufficient moments to ensure a normal limit necessary for employing the quasi‐maximum likelihood estimator. Hence we propose applying the self‐weighted quasi‐maximum exponential likelihood estimator and a bootstrap method to test and account for the asymmetry of comovements as well as different magnitudes across state pairs. Our results reveal interstate asymmetric tail dependence based on observed house price indices rather than residuals from fitting autoregressive–generalized autoregressive conditional heteroskedasticity (AR‐GARCH) models.     

Text-based crude oil price forecasting: A deep learning approach

This study proposes a new, novel crude oil price forecasting method based on online media text mining, with the aim of capturing the more immediate market antecedents of price fluctuations. Specifically, this is an early attempt to apply deep learning techniques to crude oil forecasting, and to extract hidden patterns within online news media using a convolutional neural network (CNN). While the news-text sentiment features and the features extracted by the CNN model reveal significant relationships with the price change, they need to be grouped according to their topics in the price forecasting in order to obtain a greater forecasting accuracy. This study further proposes a feature grouping method based on the Latent Dirichlet Allocation (LDA) topic model for distinguishing effects from various online news topics. Optimized input variable combination is constructed using lag order selection and feature selection methods. Our empirical results suggest that the proposed topic-sentiment synthesis forecasting models perform better than the older benchmark models. In addition, text features and financial features are shown to be complementary in producing more accurate crude oil price forecasts.     

Buying cooperation in an asymmetric environmental differential game

We consider a two-player asymmetric differential game of pollution control. One player is non-vulnerable to pollution, or unwilling to consider damages when choosing her production policy in a non-cooperative game. We characterize the feedback-Nash equilibrium and the cooperative solution. We establish conditions under which the vulnerable player can buy the cooperation of the non-vulnerable player to control her emissions. We further use the Nash bargaining solution to allocate the total cooperative dividend between the two players and propose a time-consistent decomposition overtime of the total payoff.     

Blocking efficiency in an economy with asymmetric information

In this paper, we study a pure exchange atomless economy with asymmetric information and having an ordered Banach space with an interior point in its positive cone as the commodity space. An extension of the main theorem in Vind (1972) to the private core without free disposal is established. As a particular case of this result, a solution to a problem mentioned in Pesce (2010) is derived.     

Forecasting bank failures and stress testing: A machine learning approach

This paper presents a forecasting model of bank failures based on machine-learning. The proposed methodology defines a linear decision boundary that separates the solvent banks from those that failed. This setup generates a novel alternative stress-testing tool. Our sample of 1443 U.S. banks includes all 481 banks that failed during the period 2007–2013. The set of explanatory variables is selected using a two-step feature selection procedure. The selected variables were then fed to a support vector machines forecasting model, through a training–testing learning process. The model exhibits a 99.22% overall forecasting accuracy and outperforms the well-established Ohlson’s score.     

Popularity of reinforcement-based and belief-based learning models: An evolutionary approach

In an evolutionary model, players from a given population meet randomly in pairs each instant to play a coordination game. At each instant, the learning model used is determined via some replicator dynamics that respects payoff fitness. We allow for two such models: a belief-based best-response model that uses a costly predictor, and a costless reinforcement-based one. This generates dynamics over the choice of learning models and the consequent choices of endogenous variables. We report conditions under which the long run outcomes are efficient (or inefficient) and they support the exclusive use of either of the models (or their co-existence).