Portfolio of Automated Trading Systems: Complexity and Learning Set Size Issues

In this paper, we consider using profit/loss histories of multiple automated trading systems (ATSs) as N input variables in portfolio management. By means of multivariate statistical analysis and simulation studies, we analyze the influences of sample size (L) and input dimensionality on the accuracy of determining the portfolio weights. We find that degradation in portfolio performance due to inexact estimation of N means and N(N - 1)/2 correlations is proportional to N/L; however, estimation of N variances does not worsen the result. To reduce unhelpful sample size/dimensionality effects, we perform a clustering of N time series and split them into a small number of blocks. Each block is composed of mutually correlated ATSs. It generates an expert trading agent based on a nontrainable 1/N portfolio rule. To increase the diversity of the expert agents, we use training sets of different lengths for clustering. In the output of the portfolio management system, the regularized mean-variance framework-based fusion agent is developed in each walk-forward step of an out-of-sample portfolio validation experiment. Experiments with the real financial data (2003-2012) confirm the effectiveness of the suggested approach.

[1]  Lei Xu,et al.  New Sharpe-ratio-related methods for portfolio selection , 2000, Proceedings of the IEEE/IAFE/INFORMS 2000 Conference on Computational Intelligence for Financial Engineering (CIFEr) (Cat. No.00TH8520).

[2]  Aistis Raudys,et al.  High frequency trading portfolio optimisation: Integration of financial and human factors , 2011, 2011 11th International Conference on Intelligent Systems Design and Applications.

[3]  Anja Vogler,et al.  An Introduction to Multivariate Statistical Analysis , 2004 .

[4]  Stefanos Zafeiriou,et al.  Regularized Kernel Discriminant Analysis With a Robust Kernel for Face Recognition and Verification , 2012, IEEE Transactions on Neural Networks and Learning Systems.

[5]  S. Raudys,et al.  Results in statistical discriminant analysis: a review of the former Soviet union literature , 2004 .

[6]  Nikolaus Hautsch,et al.  A Blocking and Regularization Approach to High Dimensional Realized Covariance Estimation , 2010 .

[7]  Aistis Raudys,et al.  Three decision making levels in portfolio management , 2012, 2012 IEEE Conference on Computational Intelligence for Financial Engineering & Economics (CIFEr).

[8]  Raymond Kan,et al.  Optimal Portfolio Choice with Parameter Uncertainty , 2007, Journal of Financial and Quantitative Analysis.

[9]  Šarūnas Raudys,et al.  Statistical and Neural Classifiers: An Integrated Approach to Design , 2012 .

[10]  S RzepczynskiMark.  Neural Networks in Finance: Gaining Predictive Edge in the Markets (a review) , 2007 .

[11]  Alberto Ferreira de Souza,et al.  Prediction-based portfolio optimization model using neural networks , 2009, Neurocomputing.

[12]  David Disatnik.  Portfolio Optimization Using a Block Structure for the Covariance Matrix , 2010 .

[13]  William N. Goetzmann,et al.  Active Portfolio Management , 1999 .

[14]  Sarunas Raudys,et al.  First-Order Tree-Type Dependence between Variables and Classification Performance , 2001, IEEE Trans. Pattern Anal. Mach. Intell..

[15]  Axel Gandy,et al.  THE EFFECT OF ESTIMATION IN HIGH‐DIMENSIONAL PORTFOLIOS , 2013 .

[16]  D. Berwick.  Disseminating innovations in health care. , 2003, JAMA.

[17]  Sarunas Raudys,et al.  Evolution and generalization of a single neurone: : II. Complexity of statistical classifiers and sample size considerations , 1998, Neural Networks.

[18]  E. Fama.  EFFICIENT CAPITAL MARKETS: A REVIEW OF THEORY AND EMPIRICAL WORK* , 1970 .

[19]  P. Fishburn.  Mean-Risk Analysis with Risk Associated with Below-Target Returns , 1977 .

[20]  Wing-Keung Wong,et al.  ENHANCEMENT OF THE APPLICABILITY OF MARKOWITZ'S PORTFOLIO OPTIMIZATION BY UTILIZING RANDOM MATRIX THEORY , 2009 .

[21]  S. Benninga,et al.  Shrinking the covariance matrix - simpler is better , 2007 .

[22]  C. Anderson‐Cook,et al.  An Introduction to Multivariate Statistical Analysis (3rd ed.) (Book) , 2004 .

[23]  Victor DeMiguel,et al.  Optimal Versus Naive Diversification: How Inefficient is the 1/N Portfolio Strategy? , 2009 .

[24]  Sarunas Raudys,et al.  Multi-agent System Approach to React to Sudden Environmental Changes , 2007, MLDM.

[25]  Clifford S. Ang.  High-Frequency Trading and Price Discovery , 2015 .

[26]  Simon Benninga,et al.  Shrinking the Covariance Matrix , 2007 .

[27]  R. Lourie,et al.  The Statistical Mechanics of Financial Markets , 2002 .

[28]  André Stuhlsatz,et al.  Feature Extraction With Deep Neural Networks by a Generalized Discriminant Analysis , 2012, IEEE Transactions on Neural Networks and Learning Systems.

[29]  A. Stuart,et al.  Portfolio Selection: Efficient Diversification of Investments , 1959 .

[30]  A. Ranaldo,et al.  Forecasting Realized (Co)Variances with a Block Structure Wishart Autoregressive Model , 2008 .

[31]  Edward R. Dougherty,et al.  Analytic Study of Performance of Error Estimators for Linear Discriminant Analysis , 2011, IEEE Transactions on Signal Processing.

[32]  Maciej A. Nowak,et al.  Random matrix filtering in portfolio optimization , 2005 .

[33]  M. Jeffery,et al.  Best Practices in IT Portfolio Management , 2004 .

[34]  Aistis Raudys,et al.  On the Efficient-Market Hypothesis and stock exchange game model , 2010, Expert Syst. Appl..

[35]  E. Fama.  Random Walks in Stock Market Prices , 1965 .

[36]  Sarunas Raudys,et al.  Prediction of Commodity Prices in Rapidly Changing Environments , 2005, ICAPR.

[37]  Lei Xu,et al.  An extended ASLD trading system to enhance portfolio management , 2003, IEEE Trans. Neural Networks.

[38]  B. Malkiel.  The Efficient Market Hypothesis and Its Critics , 2003 .