A real time clustering and SVM based price-volatility prediction for optimal trading strategy Subhabrata Choudhury a,1 , Subhajyoti Ghosh b,2 , Arnab Bhattacharya c , Kiran Jude Fernandes d,3,4 , Manoj Kumar Tiwari e,n,5 a Department of Metallurgical & Materials Engineering, Indian Institute of Technology Kharagpur, Kharagpur 721302, India b Department of Ocean Engineering and Naval Architecture, Indian Institute of Technology Kharagpur, Kharagpur 721302, India c University of Pittsburgh, Pittsburgh, PA 15213, United States d Department of Management, Durham University Business School, Mill Hill Lane, Durham University, Durham DH1 3LB, United Kingdom e Department of Industrial Engineering and Management Indian Institute of Technology, Kharagpur 721302, India article info Article history: Received 17 October 2012 Received in revised form 22 July 2013 Accepted 10 October 2013 Communicated by Bijaya Ketan Panigrahi Available online 29 October 2013 Keywords: Stock market Clustering Self-Organizing Maps Trading strategy Support vector machine abstract Financial return on investments and movement of market indicators are fraught with uncertainties and a highly volatile environment that exists in the global market. Equity markets are heavily affected by market unpredictability and maintaining a healthy diversified portfolio with minimum risk is undoubt- edly crucial for any investment made in such assets. Effective price and volatility prediction can highly influence the course of the investment strategy with regard to such a portfolio of equity instruments. In this paper a novel SOM based hybrid clustering technique is integrated with support vector regression for portfolio selection and accurate price and volatility predictions which becomes the basis for the particular trading strategy adopted for the portfolio. The research considers the top 102 stocks of the NSE stock market (India) to identify set of best portfolios that an investor can maintain for risk reduction and high profitability. Short term stock trading strategy and performance indicators are developed to assess the validity of the predictions with regard to actual scenarios. & 2013 Elsevier B.V. All rights reserved. 1. Introduction The global financial markets are again fraught with uncertain- ties at every level of investment and over every possible invest- ment vehicle. Recent developments like down grading of US credit rating by Standards and Poor's (S and P) from the embellished AAA to a prudent AA þ and ongoing Euro credit crunch involving massive government debts have forced several countries into tailspin and the contagion have heavily affected many economies all around the world, taking the investors by surprise and proving even their worst case predictions wrong. The implication of such astounding events could be seen in the massive price surge in the global gold markets whereas a complete opposite scenario evolved in the US equity, stock and commodities market which was complemented by a weakened dollar and an even frailer euro. An overwhelming majority of investors and investment institu- tions tend to formulate their strategies based on extrapolating simple recent trends and calculate the portfolio return-risk trade- off to formulate an optimal one. The fallacy lies in the predictions and decisions based only upon price movements of the indices or individual stock in the market and the technical analysis for the various strategies for a range of investment vehicles. The concept of risk or volatility takes a very important meaning in this context. Determining the standard deviation or variance of a particular asset class or its derivative becomes absolutely crucial in giving a holistic view of the market uncertainties existing. The structural exogenous changes in the market are extremely hard to predict, hence the analysis used in this piece of research focuses on giving an investor a robust tool that can accurately gauge the mood of the market and the asset under consideration which becomes crucial in formulating a trading strategy that matches the risk averseness or risk empathy of individual investors or conglomerates. Over reaction is extremely dangerous at arriving at investment deci- sions which becomes the cornerstone of formulating analytical or heuristic solutions towards strategy formulation that can hedge against such paranoia upon any disruptive influence. As no model is fool proof, the performance should be gauged by real time Contents lists available at ScienceDirect journal homepage: www.elsevier.com/locate/neucom Neurocomputing 0925-2312/$ - see front matter & 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.neucom.2013.10.002 n Corresponding author. Tel.: þ91 3222 283 746. E-mail addresses: subhabrata.iitkgp@gmail.com (S. Choudhury). g.subhajyoti90@gmail.com (S. Ghosh), cfcarnabiitkgp@gmail.com (A. Bhattacharya), mkt09@hotmail.com, mkt009@gmail.com (M.K. Tiwari). 1 Tel.: þ91 974 993 5575. 2 Tel.: þ91 99 337 954 32. 3 Tel.: þ44 191 33 45512. 4 URL: http://www.durham.ac.uk/k.j.fernandes. 5 Web: http://sites.google.com/site/mktiwari09. Neurocomputing 131 (2014) 419–426