The application of Dynamic Time Wrapping Algorithm and correlation coefficients in Time Sires Clustering for Index Tracking

Subject Areas : Journal of Investment Knowledge

farid tondnevis ¹ , hasan hakimian ²

1 - Master of Financial Engineering, University of Tehran, Faculty of Management
2 - Master of Financial Engineering, University of Tehran, Faculty of Management

Received: 2018-08-07 Accepted : 2018-10-08 Published : 2020-11-21

Keywords: Index tracking, Tracking Error, Correlation Coefficients, DTW,

Abstract :

Index tracking -one of the most popular methods of passive portfolio management- tries to construct a portfolio with same return as a financial market index. This research investigates the application of a binary programming model in time series clustering for index tracking. Various similarity measurements –Pearson, Kendall, Spearman, Erdem and Dynamic Time Wrapping based distance- have been applied in this research. Out of Sample test on Market Ratio and Tracking error of portfolios based on 50 more active companies index of Tehran Stock Exchange in second, third and fourth season of 1396 and first season of 1397 shows that all portfolios successfully replicate the performance of index and the tracking error of Pearson correlation based portfolio in lower than Others. Paired comparison test on Tracking Error of portfolios shows that tracking Error of Pearson Correlation based portfolio is significantly (99% confidence level) lower than other portfolios.

References:

• امیری،م؛ کرمی، ش؛ ناصرپور، ع "ردیابی شاخص بورس اوراق بهادار با در نظر گرفتن محدودیت زیان گریزی با استفاده از رویکرد جدید بیگ بنگ بیگ کرانچ," دانش سرمایه گذاری,صفحات 83 تا 105، 1395.


• راعی، ر؛ پویان‌فر، ا؛ مدیریت سرمایه‌گذاری پیشرفته، انتشارات سمت، 1384.


• سبزواری، ج؛ سجادی، م؛ سجادی، م "حل مسئله ردیابی شاخص با استفاده از الگوریتم فراابتکاری ترکیبی کرم شب تاب," مدیریت مهندسی و رایانش نرم, صفحات 125 تا 143، 1394.


• فلاح پور، س؛ تندنویس، ف؛ هاشمی، ا, "بهینه‌سازی پرتفوی ردیاب شاخص با استفاده از مدل تک شاخصی پایدار برمبنای شاخص 50 شرکت فعالتر بورس اوراق بهادار تهران," مهندسی مالی و مدیریت اوراق بهادار، صفحات 115 تا 134، 1394.


• فلاح‌پور، س؛ تندنویس، ف, "کاربرد رویکرد بهینه‌سازی استوار در تشکیل پرتفوی سهام مبتنی بر شاخص با درنظر گرفتن ردیاب پارامتر‌ها," تحقیقات مالی, دوره 17، شماره 2 صفحه 325-340, 1394.


• Rudd, "Optimal selection of passive portfolios," Financial Management, p. 9(1):57–66., 1980.


• Chen and H. Roy, "Robust portfolio selection for index tracking," Computers & Operations Research, p. 39 829–837, 2012.


• Dos and S. Cincotti, "Clustering of Financial time series with application to index and enhanced-index tracking portfolio, DIBE-CINEF,," Universitµa di Genova,, Via Opera Pia 11a, 16145 Genova, Italia, 2005.


• Cornuejols and Tutuncu, “Optimization Methods in Finance, Carnegie Mellon University, 2005.


• Berndt and J. Clifford, "Using dynamic time warping to find patterns in time series," in the AAAI Workshop on Knowledge Discovery in Databases, 1994.


• Douglas g. Bonett, Thomas a. wright, “sample size requirements for estimating Pearson, Kendall and Spearman correlations” journal of psychometrik, vol. 65, no. 1, 23-28, 2000


• Erdogan, D. Goldfarb and G. Iyengar, "Robust portfolio management. CORC Technical Report TR," Columbia University, 2004.


• Konno and A. Wijayanayake, "Minimal cost index tracking under nonlinear transaction costs and minimal transaction unit constraints.," International Journal of Theoretical and Applied Finance, p. 4(6):939–958., 2001.


• Mrkowitz, "Portfolio selection," The journal of finance, pp. 77-91, 1952.


• Markowitz, "Portfolio selection: efficient diversification of investments," Yale university press, 1959.


• Rohweder, "Implementing stock selection ideas: does tracking error optimization do any good?," Journal of Portfolio Management, p. 24(3):49–59., 1998.


• Beasley, N. Meade and T. Chang, "An evolutionary heuristic for the index tracking problem," European Journal of Operational Research, p. 148(3): 621–643., 2003.


• Roll and D. Ronald, "Optimal benchmark tracking with small portfolios," Journal of Portfolio Management, pp. 28(2):9-33, 2002.


• Shen, "A Novel Similarity Measure Model for Multivariate Time Series Based on LMNN and DTW," Springer Science+Business, 2016.


• Gilli and E. Kellezi, "The threshold accepting heuristic for index tracking.," in Financial engineering, e-commerce, and supply chain. Kluwer In: Pardalos P, Tsitsiringos VK, editors., 2002.


• Meade and G. Salkin, "Developing and maintaining an equity index fund," Journal of the Operational Research Society, p. 41(7):599–607., 1990.


• Meade and G. Salkin, "Index funds-construction and performance measurement.," J. Opl Res. Soc., pp. 40, 871-879., 1989.


• Erdem, "A new correlation coefficient for bivariate time-series data," Physica A, pp. 274-284,, 2014.


• Clarke, C. Krase and M. Statman, "Tracking error, regret and tactical asset allocation," The Journal of Portfolio Management, pp. 20, 16–24., 1994.


• Roll, "A mean/variance analysis of tracking error," Journal of Portfolio Management, p. 18:13–22, 1992.


• Rudolf, Markus and Jurgen, "A linear model for tracking error minimization," Journal of Banking & Finance, p. 23(1): 85–103., 1999.


• TSAY, Analysis of financial time series, John Wiley & Sons,, 2010.


• Focardi and F. Fabozzi, "A methodology for index tracking based on timeseries clustering," Quantitative Finance, p. 4(4):417–25., 2004.


• Philippon and M. Richardson, "Measuring Systemic Risk," November 1, 2009.


• Meihua, X. Chengxian, X. Fengmin and X. Hong, "A mixed 0–1 LP for index tracking problem with CVaR risk constraints," Springer Science+Business Media, LLC, 2011.


• Sharpe, "A simplified model for portfolio analysis," Management science,, pp. vol. 9, pp. 277-293,, 1963.