Serge Hayward - Academia.edu (original) (raw)

Papers by Serge Hayward

Research paper thumbnail of Genetically Optimized Artificial Neural Network for Financial Time Series Data Mining

Lecture Notes in Computer Science, 2006

This paper examines stock prices forecasting and trading strategies' development with means o... more This paper examines stock prices forecasting and trading strategies' development with means of computational intelligence (CI), addressing the issue of an artificial neural network (ANN) topology dependency. Simulations reveal optimal network settings. Optimality of discovered ANN topologies' is explained through their links with the ARMA processes, thus presenting identified structures as nonlinear generalizations of such processes. Optimal settings examination demonstrates the weak relationships between statistical and economic criteria. The research demonstrates that fine-tuning ANN settings is an important stage in the computational model set-up for results' improvement and mechanism understanding. Genetic algorithm (GA) is proposed to be used for model discovery, making technical decisions less arbitrary and adding additional explanatory power to the analysis of economic systems with CI. The paper is a step towards the econometric foundation of CI in finance. The choice of evaluation criteria combining statistical and economic qualities is viewed as essential for an adequate analysis of economic systems.

Research paper thumbnail of Genetically Optimised Artificial Neural Network for Financial Time Series Data Mining

Research Papers in Economics, Jul 4, 2006

Research paper thumbnail of Chapter I Financial Modeling and Forecasting 1 Financial Modeling and Forecasting with an Evolutionary Artifical Neural Network

In this chapter, I consider a design framework of a computational experiment in finance. The exam... more In this chapter, I consider a design framework of a computational experiment in finance. The examination of statistics used for economic forecasts evaluation and profitability of investment decisions, based on those forecasts, reveals only weak relationships between them. The “degree of improvement over efficient prediction ” combined with directional accuracy are proposed in an estimation technique, as an alternative to the conventional least squares. Rejecting a claim that the accuracy of the forecast does not depend upon which error-criteria are used, profitability of networks trained with L 6 loss function appeared to be statistically significant and stable. The best economic performances are realized for a 1-year investment horizon with longer training not leading to enhanced accuracy. An improvement in profitability is achieved for models optimized with genetic algorithm. Computational intelligence is advocated for searching optimal relationships among economic agents ’ risk a...

Research paper thumbnail of Building Financial Time Series Predictions with Evolutionary Artificial Neural Network

Price forecasting and trading strategies modeling are examined with major international stock ind... more Price forecasting and trading strategies modeling are examined with major international stock indexes under different time horizons. Results demonstrate that an accurate prediction is equally important as a stable saving rate for long-term survivability. The best economic performances are achieved for a one-year investment horizon with longer training not necessarily leading to improved accuracy. Thin markets’ dominance by a particular traders’ type (e.g. short memory agents) results in a higher likelihood to learn with computational intelligence tools profitable strategies, used by dominant traders. An improvement in profitability is achieved for models optimized with genetic algorithm and fine-tuning of training/validation/testing distribution. Copyright © 2004 IFAC

Research paper thumbnail of Chapter I Financial Modeling and Forecasting with an Evolutionary Artifical Neural Network

In this chapter, I consider a design framework of a computational experiment in finance. The exam... more In this chapter, I consider a design framework of a computational experiment in finance. The examination of statistics used for economic forecasts evaluation and profitability of investment decisions, based on those forecasts, reveals only weak relationships between them. The “degree of improvement over efficient prediction” combined with directional accuracy are proposed in an estimation technique, as an alternative to the conventional least squares. Rejecting a claim that the accuracy of the forecast does not depend upon which error-criteria are used, profitability of networks trained with L 6 loss function appeared to be statistically significant and stable. The best economic performances are realized for a 1-year investment horizon with longer training not leading to enhanced accuracy. An improvement in profitability is achieved for models optimized with genetic algorithm. Computational intelligence is advocated for searching optimal relationships among economic agents’ risk att...

Research paper thumbnail of Heterogeneous Agents Past and Forward Time Horizons in Setting Up a Computational Model

Computing in Economics and Finance, 2004

Price forecasting and trading strategies modelling are examined with major international stock in... more Price forecasting and trading strategies modelling are examined with major international stock indexes under different time horizons. Results demonstrate that an accurate prediction is equally important as a stable saving rate for long-term survivability. The best economic performances are achieved for a one-year investment horizon with longer training not necessarily leading to improved accuracy. Thin markets" dominance by a particular traders" type (e.g. short memory agents) results in a higher likelihood to learn with computational intelligence tools profitable strategies, used by dominant traders. An improvement in profitability is achieved for models optimized with genetic algorithm and fine-tuning of training/validation/testing distribution

Research paper thumbnail of Building Financial Time Series Predictions with Neural Genetic System

Research paper thumbnail of Financial Modeling and Forecasting with an Evolutionary Artificial Neural Network

Computational Economics

In this chapter, I consider a design framework of a computational experiment in finance. The exam... more In this chapter, I consider a design framework of a computational experiment in finance. The examination of statistics used for economic forecasts evaluation and profitability of investment decisions, based on those forecasts, reveals only weak relationships between them. The “degree of improvement over efficient prediction” combined with directional accuracy are proposed in an estimation technique, as an alternative to the conventional least squares. Rejecting a claim that the accuracy of the forecast does not depend upon which error-criteria are used, profitability of networks trained with L6 loss function appeared to be statistically significant and stable. The best economic performances are realized for a 1-year investment horizon with longer training not leading to enhanced accuracy. An improvement in profitability is achieved for models optimized with genetic algorithm. Computational intelligence is advocated for searching optimal relationships among economic agents’ risk atti...

Research paper thumbnail of Predicting Prices of Financial Assets: From Classical Economic Analysis to Agent-Based Computational Economics

Research paper thumbnail of Multiscale Representation of Agents Heterogeneous Beliefs in Analysis of CAC40 Prices with Frequency Decomposition

This paper focuses on the time series' decomposition and economic representation of its constitue... more This paper focuses on the time series' decomposition and economic representation of its constituent parts. Wavelet transforms are used for adaptive analysis of local behaviour of heterogeneous agents. Unlike fully revealing equilibrium of homogeneous beliefs, in the environment with heterogeneous beliefs prices are driven by prevailing expectations of market participants. Thus, forecasting future prices, one must form expectations of others forecasts. Evolution of agents' expectations largely governs the adaptive nature of market prices. Overlapping beliefs of heterogeneous agents prevent the effective examination of expectation formation and price forecasting by traditional methods. In the approach proposed in this paper, a time series is decomposed into a combination of underlying series, representing beliefs of major clusters of agents. The analysis of individual parts improves statistical inference, isolating effectively nonstationary and nonlinearly features. Emergent local behaviour is also more receptive to prediction. The overall forecast (weighted combination of individual forecasts) is found to be determined and evolved depending on specific market conditions. On the statistical level, the data generating mechanism is considered as complex multistructured system, with individual layers corresponding to particular frequencies. Reflecting the time preferences of agents, trading strategies being homogeneous intra-type are heterogeneous inter-type for agents with distinct time preferences. Overall market activity at each moment, providing the dynamic feedback across agents' types, generates market prices. The frequency decomposition of a time series identifies the local and global structures and separates short and long time dynamics. The Genetic Algorithm is applied to determine the optimal decomposition of the signal and representation of heterogeneous traders. The Artificial Neural Network is trained to learn information at the scale level that is hidden in the aggregate. The resulting models seek to enhance the understanding of the underlying data generating mechanisms of financial time series and to develop new approaches for financial forecasting.

Research paper thumbnail of Genetic Algorithm Optimization of an Artificial Neural Network for Financial Applications

Model discovery and performance surface optimization with genetic algorithm demonstrate profitabi... more Model discovery and performance surface optimization with genetic algorithm demonstrate profitability improvement with an inconclusive effect on statistical criteria. The examination of relationships between statistics used for economic forecasts evaluation and profitability of investment decisions reveals that only the ‘degree of improvement over efficient prediction’ shows robust links with profitability. If profits are not observable, this measure is proposed as an evaluation criterion for an economic prediction. Also combined with directional accuracy, it could be used in an estimation technique for economic behavior, as an alternative to conventional least squares.

Research paper thumbnail of Evolutionary Artificial Neural Network Optimisation in Financial Engineering

Proceedings of the Fourth International Conference on Hybrid Intelligent Systems, Dec 5, 2004

Analytical examination of loss functions' families demonstrates that investors' utility m... more Analytical examination of loss functions' families demonstrates that investors' utility maximisation is determined by their risk attitude. In computational settings, stock traders' fitness is assessed in response to a slow-step increase in the value of the risk aversion coefficient. The experiment rejects the claims that the accuracy of the forecast does not depend upon which error-criteria are used and none of them is related to the profitability of the forecast. Profitability of networks trained with L/sub 6/ loss function appeared to be statistically significant and stable, although links between loss functions and accuracy of forecasts were less conclusive.

Research paper thumbnail of Quantitative Forecasting and Modeling Stock Price Fluctuations

Considering the effect of economic agents’ preferences on their actions, relationships between co... more Considering the effect of economic agents’ preferences on their actions, relationships between conventional summary statistics and forecasts’ profit are investigated. Analytical examination demonstrates that investors’ utility maximization is determined by their risk attitude. The computational experiment rejects the claims that the accuracy of the forecast does not depend upon which error-criteria are used. Profitability of networks trained with L6 loss function appeared to be statistically significant and stable.

Research paper thumbnail of Agent-Based Modelling with Wavelets and an Evolutionary Artificial Neural Network: Applications to CAC 40 Forecasting

Analysis of separate scales of a complex signal provides a valuable source of information, consid... more Analysis of separate scales of a complex signal provides a valuable source of information, considering that different financial decisions occur at different scales. Wavelet transform decomposition of a complex time series into separate scales and their economic representation is a focus of this study. An evolutionary / artificial neural network (E/ANN) is used to learn the information at separate scales and combine it into meaningfully weighted structures. Potential applications of the proposed approach are in financial forecasting and trading strategies development based on individual preferences and trading styles.

Research paper thumbnail of Genetically Optimized Artificial Neural Network for Financial Time Series Data Mining

Lecture Notes in Computer Science, Dec 9, 2006

This paper examines stock prices forecasting and trading strategies' development with means o... more This paper examines stock prices forecasting and trading strategies' development with means of computational intelligence (CI), addressing the issue of an artificial neural network (ANN) topology dependency. Simulations reveal optimal network settings. Optimality of discovered ANN topologies' is explained through their links with the ARMA processes, thus presenting identified structures as nonlinear generalizations of such processes. Optimal settings examination demonstrates the weak relationships between statistical and economic criteria. The research demonstrates that fine-tuning ANN settings is an important stage in the computational model set-up for results' improvement and mechanism understanding. Genetic algorithm (GA) is proposed to be used for model discovery, making technical decisions less arbitrary and adding additional explanatory power to the analysis of economic systems with CI. The paper is a step towards the econometric foundation of CI in finance. The choice of evaluation criteria combining statistical and economic qualities is viewed as essential for an adequate analysis of economic systems.

Research paper thumbnail of Genetically Optimised Artificial Neural Network for Financial Time Series Data Mining

Research paper thumbnail of Risk aversion and agents’ survivability in a financial market

Considering the effect of economic agents’ preferences on their actions, the relationships betwee... more Considering the effect of economic agents’ preferences on their actions, the relationships between conventional summary statistics and forecast profits are investigated. An analytical examination of loss function families demonstrates that investors’ utility maximisation is determined by their risk attitudes. In computational settings, stock traders’ fitness is assessed in response to a slow step increase in the value of the risk aversion coefficient. The experiment rejects the claims that the accuracy of the forecast does not depend upon which error-criteria are used and that none of them is related to the profitability of the forecast. The profitability of networks trained with L6 loss function appeared to be statistically significant and stable, although links between the loss functions and the accuracy of forecasts were less conclusive.

Research paper thumbnail of Heterogeneity of Price Discovery Processes in Financial Markets

This paper considers heterogeneity-driven asymmetry in the stock market with the low frequency sh... more This paper considers heterogeneity-driven asymmetry in the stock market with the low frequency shocks penetrating the entire market, whereas the high frequency shocks are short-lived and often have no impact outside of their boundaries. Testing the long memory versus structural brake hypotheses identifies a number of sample periods when structural breaks spuriously induce the long memory effect on a particular frequency, without their presence across all frequencies. Similarly, testing the structural brakes versus phase shifts hypotheses detects periods when the low and high frequencies move into and out of phase with each other, resulting in phase shifts rather than structural breaks, claimed by other studies. Distinguishing long memory, structural breaks and phase shifts enhances the understanding of the series' emergent nonstationary behaviour. A heterogeneous beliefs model with expectations differentiated according to their time dimension is developed. Decomposing a time se...

Research paper thumbnail of Predicting Prices of Financial Assets: From Classical Economics to Intelligent Finance

New Mathematics and Natural Computation, 2011

Determining the circumstances under which it is possible to make any sort of prediction is a fund... more Determining the circumstances under which it is possible to make any sort of prediction is a fundamental question in financial research. The presence of complex and robust statistical characteristics, exhibited by most financial time series, raise doubts on the simple relationship between information and price changes, as implied by the efficient market hypothesis. In this paper, we consider the main competing economic hypotheses and examine different approaches for learning the price behaviour in financial markets. Our analysis reveals the need to approach the problem from a new perspective. In financial markets, traders are not only adapting, but also determine and form the economic mechanism essentially by their actions. In these settings, financial markets are evolutionary structures of competing trading strategies; prices in such markets are driven endogenously by the induced expectations. A combination of economics, computer and cognitive science in cross-disciplinary study of...

Research paper thumbnail of The Role of Heterogeneous Agents’ Past and Forward Time Horizons in Formulating Computational Models

Computational Economics, 2005

Price forecasting and trading strategies modelling are examined with major international stock in... more Price forecasting and trading strategies modelling are examined with major international stock indexes under different time horizons. Results demonstrate that an accurate prediction is equally important as a stable saving rate for long-term survivability. The best economic performances are achieved for a one-year investment horizon with longer training not necessarily leading to improved accuracy. Thin markets' dominance by a particular traders' type (e.g. short memory agents) results in a higher likelihood to learn with computational intelligence tools profitable strategies, used by dominant traders. An improvement in profitability is achieved for models optimized with genetic algorithm and fine-tuning of training/validation/testing distribution.

Research paper thumbnail of Genetically Optimized Artificial Neural Network for Financial Time Series Data Mining

Lecture Notes in Computer Science, 2006

This paper examines stock prices forecasting and trading strategies' development with means o... more This paper examines stock prices forecasting and trading strategies' development with means of computational intelligence (CI), addressing the issue of an artificial neural network (ANN) topology dependency. Simulations reveal optimal network settings. Optimality of discovered ANN topologies' is explained through their links with the ARMA processes, thus presenting identified structures as nonlinear generalizations of such processes. Optimal settings examination demonstrates the weak relationships between statistical and economic criteria. The research demonstrates that fine-tuning ANN settings is an important stage in the computational model set-up for results' improvement and mechanism understanding. Genetic algorithm (GA) is proposed to be used for model discovery, making technical decisions less arbitrary and adding additional explanatory power to the analysis of economic systems with CI. The paper is a step towards the econometric foundation of CI in finance. The choice of evaluation criteria combining statistical and economic qualities is viewed as essential for an adequate analysis of economic systems.

Research paper thumbnail of Genetically Optimised Artificial Neural Network for Financial Time Series Data Mining

Research Papers in Economics, Jul 4, 2006

Research paper thumbnail of Chapter I Financial Modeling and Forecasting 1 Financial Modeling and Forecasting with an Evolutionary Artifical Neural Network

In this chapter, I consider a design framework of a computational experiment in finance. The exam... more In this chapter, I consider a design framework of a computational experiment in finance. The examination of statistics used for economic forecasts evaluation and profitability of investment decisions, based on those forecasts, reveals only weak relationships between them. The “degree of improvement over efficient prediction ” combined with directional accuracy are proposed in an estimation technique, as an alternative to the conventional least squares. Rejecting a claim that the accuracy of the forecast does not depend upon which error-criteria are used, profitability of networks trained with L 6 loss function appeared to be statistically significant and stable. The best economic performances are realized for a 1-year investment horizon with longer training not leading to enhanced accuracy. An improvement in profitability is achieved for models optimized with genetic algorithm. Computational intelligence is advocated for searching optimal relationships among economic agents ’ risk a...

Research paper thumbnail of Building Financial Time Series Predictions with Evolutionary Artificial Neural Network

Price forecasting and trading strategies modeling are examined with major international stock ind... more Price forecasting and trading strategies modeling are examined with major international stock indexes under different time horizons. Results demonstrate that an accurate prediction is equally important as a stable saving rate for long-term survivability. The best economic performances are achieved for a one-year investment horizon with longer training not necessarily leading to improved accuracy. Thin markets’ dominance by a particular traders’ type (e.g. short memory agents) results in a higher likelihood to learn with computational intelligence tools profitable strategies, used by dominant traders. An improvement in profitability is achieved for models optimized with genetic algorithm and fine-tuning of training/validation/testing distribution. Copyright © 2004 IFAC

Research paper thumbnail of Chapter I Financial Modeling and Forecasting with an Evolutionary Artifical Neural Network

In this chapter, I consider a design framework of a computational experiment in finance. The exam... more In this chapter, I consider a design framework of a computational experiment in finance. The examination of statistics used for economic forecasts evaluation and profitability of investment decisions, based on those forecasts, reveals only weak relationships between them. The “degree of improvement over efficient prediction” combined with directional accuracy are proposed in an estimation technique, as an alternative to the conventional least squares. Rejecting a claim that the accuracy of the forecast does not depend upon which error-criteria are used, profitability of networks trained with L 6 loss function appeared to be statistically significant and stable. The best economic performances are realized for a 1-year investment horizon with longer training not leading to enhanced accuracy. An improvement in profitability is achieved for models optimized with genetic algorithm. Computational intelligence is advocated for searching optimal relationships among economic agents’ risk att...

Research paper thumbnail of Heterogeneous Agents Past and Forward Time Horizons in Setting Up a Computational Model

Computing in Economics and Finance, 2004

Price forecasting and trading strategies modelling are examined with major international stock in... more Price forecasting and trading strategies modelling are examined with major international stock indexes under different time horizons. Results demonstrate that an accurate prediction is equally important as a stable saving rate for long-term survivability. The best economic performances are achieved for a one-year investment horizon with longer training not necessarily leading to improved accuracy. Thin markets" dominance by a particular traders" type (e.g. short memory agents) results in a higher likelihood to learn with computational intelligence tools profitable strategies, used by dominant traders. An improvement in profitability is achieved for models optimized with genetic algorithm and fine-tuning of training/validation/testing distribution

Research paper thumbnail of Building Financial Time Series Predictions with Neural Genetic System

Research paper thumbnail of Financial Modeling and Forecasting with an Evolutionary Artificial Neural Network

Computational Economics

In this chapter, I consider a design framework of a computational experiment in finance. The exam... more In this chapter, I consider a design framework of a computational experiment in finance. The examination of statistics used for economic forecasts evaluation and profitability of investment decisions, based on those forecasts, reveals only weak relationships between them. The “degree of improvement over efficient prediction” combined with directional accuracy are proposed in an estimation technique, as an alternative to the conventional least squares. Rejecting a claim that the accuracy of the forecast does not depend upon which error-criteria are used, profitability of networks trained with L6 loss function appeared to be statistically significant and stable. The best economic performances are realized for a 1-year investment horizon with longer training not leading to enhanced accuracy. An improvement in profitability is achieved for models optimized with genetic algorithm. Computational intelligence is advocated for searching optimal relationships among economic agents’ risk atti...

Research paper thumbnail of Predicting Prices of Financial Assets: From Classical Economic Analysis to Agent-Based Computational Economics

Research paper thumbnail of Multiscale Representation of Agents Heterogeneous Beliefs in Analysis of CAC40 Prices with Frequency Decomposition

This paper focuses on the time series' decomposition and economic representation of its constitue... more This paper focuses on the time series' decomposition and economic representation of its constituent parts. Wavelet transforms are used for adaptive analysis of local behaviour of heterogeneous agents. Unlike fully revealing equilibrium of homogeneous beliefs, in the environment with heterogeneous beliefs prices are driven by prevailing expectations of market participants. Thus, forecasting future prices, one must form expectations of others forecasts. Evolution of agents' expectations largely governs the adaptive nature of market prices. Overlapping beliefs of heterogeneous agents prevent the effective examination of expectation formation and price forecasting by traditional methods. In the approach proposed in this paper, a time series is decomposed into a combination of underlying series, representing beliefs of major clusters of agents. The analysis of individual parts improves statistical inference, isolating effectively nonstationary and nonlinearly features. Emergent local behaviour is also more receptive to prediction. The overall forecast (weighted combination of individual forecasts) is found to be determined and evolved depending on specific market conditions. On the statistical level, the data generating mechanism is considered as complex multistructured system, with individual layers corresponding to particular frequencies. Reflecting the time preferences of agents, trading strategies being homogeneous intra-type are heterogeneous inter-type for agents with distinct time preferences. Overall market activity at each moment, providing the dynamic feedback across agents' types, generates market prices. The frequency decomposition of a time series identifies the local and global structures and separates short and long time dynamics. The Genetic Algorithm is applied to determine the optimal decomposition of the signal and representation of heterogeneous traders. The Artificial Neural Network is trained to learn information at the scale level that is hidden in the aggregate. The resulting models seek to enhance the understanding of the underlying data generating mechanisms of financial time series and to develop new approaches for financial forecasting.

Research paper thumbnail of Genetic Algorithm Optimization of an Artificial Neural Network for Financial Applications

Model discovery and performance surface optimization with genetic algorithm demonstrate profitabi... more Model discovery and performance surface optimization with genetic algorithm demonstrate profitability improvement with an inconclusive effect on statistical criteria. The examination of relationships between statistics used for economic forecasts evaluation and profitability of investment decisions reveals that only the ‘degree of improvement over efficient prediction’ shows robust links with profitability. If profits are not observable, this measure is proposed as an evaluation criterion for an economic prediction. Also combined with directional accuracy, it could be used in an estimation technique for economic behavior, as an alternative to conventional least squares.

Research paper thumbnail of Evolutionary Artificial Neural Network Optimisation in Financial Engineering

Proceedings of the Fourth International Conference on Hybrid Intelligent Systems, Dec 5, 2004

Analytical examination of loss functions' families demonstrates that investors' utility m... more Analytical examination of loss functions' families demonstrates that investors' utility maximisation is determined by their risk attitude. In computational settings, stock traders' fitness is assessed in response to a slow-step increase in the value of the risk aversion coefficient. The experiment rejects the claims that the accuracy of the forecast does not depend upon which error-criteria are used and none of them is related to the profitability of the forecast. Profitability of networks trained with L/sub 6/ loss function appeared to be statistically significant and stable, although links between loss functions and accuracy of forecasts were less conclusive.

Research paper thumbnail of Quantitative Forecasting and Modeling Stock Price Fluctuations

Considering the effect of economic agents’ preferences on their actions, relationships between co... more Considering the effect of economic agents’ preferences on their actions, relationships between conventional summary statistics and forecasts’ profit are investigated. Analytical examination demonstrates that investors’ utility maximization is determined by their risk attitude. The computational experiment rejects the claims that the accuracy of the forecast does not depend upon which error-criteria are used. Profitability of networks trained with L6 loss function appeared to be statistically significant and stable.

Research paper thumbnail of Agent-Based Modelling with Wavelets and an Evolutionary Artificial Neural Network: Applications to CAC 40 Forecasting

Analysis of separate scales of a complex signal provides a valuable source of information, consid... more Analysis of separate scales of a complex signal provides a valuable source of information, considering that different financial decisions occur at different scales. Wavelet transform decomposition of a complex time series into separate scales and their economic representation is a focus of this study. An evolutionary / artificial neural network (E/ANN) is used to learn the information at separate scales and combine it into meaningfully weighted structures. Potential applications of the proposed approach are in financial forecasting and trading strategies development based on individual preferences and trading styles.

Research paper thumbnail of Genetically Optimized Artificial Neural Network for Financial Time Series Data Mining

Lecture Notes in Computer Science, Dec 9, 2006

This paper examines stock prices forecasting and trading strategies' development with means o... more This paper examines stock prices forecasting and trading strategies' development with means of computational intelligence (CI), addressing the issue of an artificial neural network (ANN) topology dependency. Simulations reveal optimal network settings. Optimality of discovered ANN topologies' is explained through their links with the ARMA processes, thus presenting identified structures as nonlinear generalizations of such processes. Optimal settings examination demonstrates the weak relationships between statistical and economic criteria. The research demonstrates that fine-tuning ANN settings is an important stage in the computational model set-up for results' improvement and mechanism understanding. Genetic algorithm (GA) is proposed to be used for model discovery, making technical decisions less arbitrary and adding additional explanatory power to the analysis of economic systems with CI. The paper is a step towards the econometric foundation of CI in finance. The choice of evaluation criteria combining statistical and economic qualities is viewed as essential for an adequate analysis of economic systems.

Research paper thumbnail of Genetically Optimised Artificial Neural Network for Financial Time Series Data Mining

Research paper thumbnail of Risk aversion and agents’ survivability in a financial market

Considering the effect of economic agents’ preferences on their actions, the relationships betwee... more Considering the effect of economic agents’ preferences on their actions, the relationships between conventional summary statistics and forecast profits are investigated. An analytical examination of loss function families demonstrates that investors’ utility maximisation is determined by their risk attitudes. In computational settings, stock traders’ fitness is assessed in response to a slow step increase in the value of the risk aversion coefficient. The experiment rejects the claims that the accuracy of the forecast does not depend upon which error-criteria are used and that none of them is related to the profitability of the forecast. The profitability of networks trained with L6 loss function appeared to be statistically significant and stable, although links between the loss functions and the accuracy of forecasts were less conclusive.

Research paper thumbnail of Heterogeneity of Price Discovery Processes in Financial Markets

This paper considers heterogeneity-driven asymmetry in the stock market with the low frequency sh... more This paper considers heterogeneity-driven asymmetry in the stock market with the low frequency shocks penetrating the entire market, whereas the high frequency shocks are short-lived and often have no impact outside of their boundaries. Testing the long memory versus structural brake hypotheses identifies a number of sample periods when structural breaks spuriously induce the long memory effect on a particular frequency, without their presence across all frequencies. Similarly, testing the structural brakes versus phase shifts hypotheses detects periods when the low and high frequencies move into and out of phase with each other, resulting in phase shifts rather than structural breaks, claimed by other studies. Distinguishing long memory, structural breaks and phase shifts enhances the understanding of the series' emergent nonstationary behaviour. A heterogeneous beliefs model with expectations differentiated according to their time dimension is developed. Decomposing a time se...

Research paper thumbnail of Predicting Prices of Financial Assets: From Classical Economics to Intelligent Finance

New Mathematics and Natural Computation, 2011

Determining the circumstances under which it is possible to make any sort of prediction is a fund... more Determining the circumstances under which it is possible to make any sort of prediction is a fundamental question in financial research. The presence of complex and robust statistical characteristics, exhibited by most financial time series, raise doubts on the simple relationship between information and price changes, as implied by the efficient market hypothesis. In this paper, we consider the main competing economic hypotheses and examine different approaches for learning the price behaviour in financial markets. Our analysis reveals the need to approach the problem from a new perspective. In financial markets, traders are not only adapting, but also determine and form the economic mechanism essentially by their actions. In these settings, financial markets are evolutionary structures of competing trading strategies; prices in such markets are driven endogenously by the induced expectations. A combination of economics, computer and cognitive science in cross-disciplinary study of...

Research paper thumbnail of The Role of Heterogeneous Agents’ Past and Forward Time Horizons in Formulating Computational Models

Computational Economics, 2005

Price forecasting and trading strategies modelling are examined with major international stock in... more Price forecasting and trading strategies modelling are examined with major international stock indexes under different time horizons. Results demonstrate that an accurate prediction is equally important as a stable saving rate for long-term survivability. The best economic performances are achieved for a one-year investment horizon with longer training not necessarily leading to improved accuracy. Thin markets' dominance by a particular traders' type (e.g. short memory agents) results in a higher likelihood to learn with computational intelligence tools profitable strategies, used by dominant traders. An improvement in profitability is achieved for models optimized with genetic algorithm and fine-tuning of training/validation/testing distribution.