Abstracts of the Accepted Papers

of the Research Area: CIEF

NM&NC Vol. 2, No 3 ( Nov 2006 )

Title:

RANKING STOCKS USING THE FUZZY MULTIPLE CRITERIA DECISION
MAKING APPROACH

Author: CHUNG-TSEN TSAO
Affliation: Department of Finance, National Pingtung Institute of Commerce, Taiwan

Abstract:

This work applies the Fuzzy Multiple Criteria Decision Making (FMCDM) approach to assist in
making stock investment decisions. The proposed approach applies to quantitative data and can
accommodate qualitative information, which is normally difficult to be integrated by traditional
finance and accounting methods. A major-sub-criteria hierarchy is established to reduce the
possibility of over-weighing some dependent criteria existing in a single-level structure. The
ratings of stocks versus qualitative sub-criteria and the weights of major and sub-criteria are assessed in linguistic terms represented by fuzzy numbers. Each sub-criterion is in a benefit, cost, or balanced nature. New standardization methods for cost-nature and balanced-nature criteria are
presented. The algorithms of membership functions of the final aggregation are derived from the
roots of cubic equations of multiplications of triple fuzzy numbers. Since these algorithms are
clearly developed, the investor can easily calculate the fuzzy aggregation values. The defuzzified
final aggregation judges the performance of alternative stocks. Moreover, the ratio of the market
price to performance (PP) is suggested to filter the over and under-pricing of alternative stocks. A
set of buying and selling rules are recommended based on the performance and PP ratio. Finally,
an empirical example of evaluating a set of TSE-listed stocks tests the proposed approach.

Keywords: stock investment decisions; FMCDM

NM&NC Vol. 2, No 3 ( Nov 2006 )

Title:

Boosting-based framework for portfolio strategy discovery and optimization

Author: Valeriy V. Gavrishchaka

Affliation: Head of Quantitative Research, Alexandra Investment Management, 767 3-d Avenue, New York

Abstract:

Increasing availability of the multi-scale market data exposes that leads to the acceptable performance of the portfolio strategy. In this work, a boosting-based framework for a direct trading strategy and portfolio optimization is introduced. Due to inherent adaptive control of the parameter space dimensionality, this technique can work with very large pools of base strategies and financial instruments that are usually prohibitive for other portfolio optimization frameworks. Unlike existing approaches, this framework can be effectively used for the coupled optimization of the portfolio capital/asset allocation and dynamic trading strategies. Generated portfolios
of trading strategies not only exhibit stable and robust performance but also remain interpretable. Encouraging preliminary results based on real market data are presented and discussed.

Keywords: Boosting; ensemble learning; portfolio optimization; trading strategies.

NM&NC Vol. 2, No 3 ( Nov 2006 )

Title:

PREDICTION OF BIRD FLU A(H5N1) OUTBREAKS IN TAIWAN BY ONLINE

AUCTION: EXPERIMENTAL RESULTS

Author: SUN-CHONG WANG
Affliation: Institute of Systems Biology and Bioinformatics, National Central University,Jhongli City

Abstract:

The ability of accurate epidemic prediction facilitates early preparation for the disease and minimizes losses due to any strikes. We devised a platform on the Web for users to exchange their information/opinions on possible avian flu outbreaks in Taiwan. The likelihood of the first human infection from bird flu in Taiwan in, say, December 2005 is securitized in the form of a futures contract. Incentives are introduced via a tournament: users trade the futures in the market on our Web server in order to win the awards at the end of the tournament. We ran such a tournament during the period between December 2005 and February 2006. The results of the futures’ prices correctly predicted no outbreaks of bird flu among the residents in Taiwan during the 3-month period, suggesting that the design of the futures exchange on the Web be a potentially useful
tool for event forecasting. Another crucial aspect of the experiment is that, associated with the price convergence, the transaction volume also quickly converges to zero, which is closely related to the famous no-trade theorem in theoretical economics.

Keywords: avian influenza; futures exchange; market; no-trade theorem

NM&NC Vol. 2, No 3 ( Nov 2006 )

Title:

What causes persistence of stock return volatility? One possible explanation with an artificial stock

market

Author: Ryuichi Yamamoto

Affliation: Department of International Trade, National Chengchi University, Taiwan

Abstract:

This paper explores a possible cause of persistence in stock return volatility. Artificial stock markets are examined with different learning mechanisms, i.e., imitative and experiential learning. The simulation result shows that an economy with imitative learning gives rise to persistence of return volatility while an experiential learning economy does not. We find that volatility becomes persistent as investors learn through imitating the prediction methods of others. Imitation is crucial to producing the persistence in stock return volatility.

Keywords: Asset Pricing; Learning; Evolution; Volatility Clustering

NM&NC Vol. 2, No 3 ( Nov 2006 )

Title:

Forecasting High-Frequency Financial Data Volatility via Nonparametric
Algorithms: Evidence from Taiwan’s Financial Marketsecision-Making Model for Stock

Markets

Author: Wo-Chiang Lee
Affliation: Department of Finance and Banking, Aletheia University, Taiwan

Abstract:

This paper uses two computational intelligence algorithms, namely, artificial neural networks
(ANN) and genetic programming (GP), for forecasting the volatility of highfrequency
TAIEX financial data with four different horizons and compares the out-sample
forecasting performance with the GARCH(1,1), EGRACH(1,1) and GJR-GARCH(1,1)
models. Based on intraday integrated volatility, the mean squared error (MSE), mean
absolute error (MAE), mean absolute percentage error (MAPE), Theil’s U and the VaR
backtest are used as performance indexes. Our empirical results reveal that the GP and
ANN perform reasonably well in forecasting out-sample volatility compared to other
parametric volatility forecasting models for most of the performance indexes. Our results
also suggest that nonparametric computational intelligence algorithms are powerful for
modeling the volatility of high-frequency intraday financial data.

Keywords: Integrated volatility; genetic programming; artificial neural networks

NM&NC Vol. 2, No 3 ( Nov 2006 )

Title:

Graphs, Networks and ACE

Author: Shu-Heng Chen

Affliation: Department of Economics, National Chengchi University, Taiwan

Abstract:

Following the standard probabilistic approach, we shall explicitly show the relation between a microscopic and a macroscopic view of an economy in the context of a discrete choice model. Two crucial issues to do with the graphical applications to the network economy are addressed. The first one concerns the representation richness of the graph,whereas the second one concerns the formation and evolution of the graph when it is applied to social networks. This study can be a starting point to see the relevance of agent-based computational economics (ACE) to the network economy, in particular after bringing in the interaction mechanism associated with a network topology.

Keywords: Network Topology; Graph; Agent-Based Computational Economics.

NM&NC Vol. 2, No 3 ( Nov 2006 )

Title:

Predicting uncertain outcomes using information markets: Trader behavior and information

aggregation

Author: Chao-Hsien Chu
Affliation: School of Information Systems, Singapore Management University,80 Stamford Road, Singapore

Abstract:

Forecasting seems to be a ubiquitous endeavor in human societies. In this paper, information
markets are introduced as a promising mechanism for predicting uncertain
outcomes. Information markets are markets that are specially designed for aggregating
information and making predictions on future events. A generic model of information
markets is proposed. We derive some fundamental properties on when information
markets can converge to the direct communications equilibrium, which aggregates all
information across traders and is the best possible prediction for the event under consideration.
Information markets, if properly designed, have substantial potential to facilitate
organizations in making better informed decisions.

Keywords: Information market; Prediction; Trader behavior; Information aggregation

NM&NC Vol. 1, # 3 ( Nov 2005 )

Title:

Co-Evolving Business Models: A Case Study with the Internet Service Provider (ISP)

Industry

Author: Ian Fenty,¹ Eric Bonabeau,² Juergen Branke³

Affliation:

¹Icosystem Corporation, 10 Fawcett Street, Cambridge, MA 02138, USA

²Icosystem Corporation, 10 Fawcett Street, Cambridge, MA 02138, USA

³Institute AIFB, University of Karlsruhe, 76128 Karlsruhe, Germany

Abstract:

In this paper, co-evolution is used to examine the long-term evolution of business models in an industry. Two types of co-evolution are used: synchronous, whereby the entire population of business models is replaced with a new population at each generation, and asynchronous, whereby only one individual is replaced.

Keywords: Agent-based modeling; co-evolution

NM&NC Vol. 1, # 2 ( July 2005 )

Title:

Decision-Making Model for Stock Markets Based on Particle Swarm Optimization

Algorithm

Author: Jovita Nenortaite and Rimvydas Simutis
Affliation: Kaunas Faculty of Humanities, Vilnius University, Muitines 8

Abstract:

The objective of this paper is to introduce the decision-making model for stock markets. The proposed model is based on the study of historic data and the application of Artificial Neural Networks ( ANN ) and Particles Swarm Optimization ( PSO ) algorithm. In the proposed decision-making model the ANN are recommendations concerning the purchase of the stocks. Subsequently, the application of PSO algorithm is made. The core idea of this algorithm application is to select the "global best" ANN for future investment decisions and to adapt the weights of other ANN towards the weights of the best network. The experimental investigation results presented in this paper show the potentiality of PSO algorithm applications for the decision-making in the stock markets.

Keywords: Stock markets; genetic algorithms; swarm intelligence

NM&NC Vol. 1, # 2 ( July 2005 )

Title:

Avalanche Dynamics of the Financial Market

Author: Pei-Ling Zhou,¹ Chun-Xia Yang,^2,* Tao Zhou,³Min Xu,⁴Jun Liu and⁵ Bing-Hong Wang

Affliation:

¹Department of Electronic Science and Technology, University of Science and Technology of China

Hefei Anhui, 230026, People's Republic of China

⁴ Department of Modern Physics, University of Science and Technology of China

Hefei Anhui, 230026, People's Republic of China

Abstract:

A parsimonious percolation model for stock market is proposed, of which the avalanche dynamics agree with the real-life one as well. We have also investigated how the interaction parameter p affects the price dynamics. Simulation results about the formation of the bullish/bearish market and corresponding avalanche taking place in the market indicate that the magnified "herd behavior" resulting from the evolution of p may be the origin of the observed avalanche phenomena.

Keywords: Complex system; financial market; percolation; nonlinear dynamics; avalanche.

NM&NC Vol. 1, # 2 ( July 2005 )

Title:

Applying the Genetic-Based Neural Networks to Volatility Trading

Author: Shinn-Wen Wang
Affliation: Department of Business Administration, College of Management, National Changhua University

of Education

Abstract:

The Black-Scholes options pricing model is widely applied in various options contracts, including contract design, trading, assets evaluation, and enterprise value estimation, etc. Unfortunately, this theoretical model limited by the influences of many unexpected real world phenomena due to six unreasonable assumptions. If we were to soundly take these phenomena into account, the opportunity to gain an excess return would be created. This research therefore combines both the remarkable effects caused by the implied volatility smile ( or skew ) and the tick-jump discrepancy between the underlying and derivative prices to establish a two-phase options arbitrage model using a genetic-based neural network ( GNN ). Using evidence from the warrant market in Taiwan, it is shown that the GNN model with arbitrage operations is superior in terms of performance to the original Black-Scholes-based arbitrage model. The GNN model is found to be suitable for application to various options markets as the valuation factors are modified. This paper helps to integrate the theoretical model with important practical considerations.

Keywords: Black-Scholes; volatility skews; warrant; arbitrage;

NM&NC Vol. 1, # 2 ( July 2005 )

Title:

A Hybrid Genetic Algorithm for the Maximum Likelihood Estimation of Models with

Multiple Equilibria: A First Report

Author: Victor Aguirregabiria,¹ Pedro Mira²

Affliation:

¹ Department of Economics, Boston University, USA

² CEMFI, Casado del Alisal, 5, Madrid, Madrid 28004, Spain

Abstract:

This paper presents a hybrid genetic algorithm to obtain maximum likelihood estimates of parameters in structural econometric models with multiple equilibria. The algorithm combines a pseudo maximum likelihood ( PML ) procedure with a genetic algorithm ( GA ). The GA searches globally over the large space of possible combinations of multiple equilibria in the trial value of the structrual parameters.

Keywords: Genetic algorithms; maximum likelihood estimation; multiple equilibria

NM&NC Vol. 1, # 1 ( March 2005 )

Title:

Extended Daily Exchange Rates Forecast Using Wavelet Temporal

Resolutions

Author: Prof. Mak Kaboudan

Affliation: Business Statistics,School of Business,University of Redlands, California, USA

Abstract:

Applying genetic programming and artificial neural networks to raw as well as wavelet-transformed exchange rate data showed that genetic programming may have good extended forecasting abilities. Although it is well known that most predictions of exchange rates using many alternative techniques could not deliver better forecasts than the random walk model, in this paper employing natural computational strategies to forecast three different exchange rates produced two extended forecasts (that go beyond one-step-ahead) that are better than naïve random walk predictions. Sixteen-step-ahead forecasts obtained using genetic programming outperformed the one- and sixteen-step-ahead random walk US dollar/Taiwan dollar exchange rate predictions. Further, sixteen-step-ahead forecasts of the wavelet-transformed US dollar/Japanese Yen exchange rate also using genetic programming outperformed the sixteen-step-ahead random walk predictions of the exchange rate. However, random walk predictions of the US dollar/British pound exchange rate outperformed all forecasts obtained using genetic programming. Random walk predictions of the same three exchange rates employing raw and wavelet-transformed data also outperformed all forecasts obtained using artificial neural networks.

Keywords: Genetic programming; artificial neural networks; Haar wavelets.

NM&NC Vol. 1, # 1 ( March 2005 )

Title:

The apprentice wizard: monetary policy, complexity and learning

Author: Domenico Delli Gatti,¹ Edoardo Gaffeo,^2,* Mauro Gallegati,³Antonio Palestrini⁴

Affliation:

¹Institute of Quantitative Methods and Economic Theory, Catholic University of Milan, Italy

²Department of Economics and CEEL, University of Trento, Italy

³ Department of Economics, Polytechnic University of Marche, Ancona, Italy

⁴ Department of Law in Society and History, University of Teramo, Italy

Abstract:

This paper aims at reassessing some central issues of monetary policy by offering a model in which a central bank tries to stabilize fluctuations in aggregate output and inflation in an adaptive complex economy. We resort to evolutionary algorithms to model the central bank behaviour under discretion, and confront the efficiency of discretion with the choice of full commitment to a fixed rule.

JEL classification: E52, E58

Keywords: Monetary policy; Taylor rule; Learning.