FRACTAL FORMATION AND TREND TRADING STRATEGY IN FUTURES MARKET Saulius Masteika1,2, Aleksandras V.Rutkauskas1, Audrius Lopata2,3 1
Department of Finance Engineering, Faculty of Business Management, VGTU, Vilnius, Lithuania, EU, E-mail (corresponding author):
[email protected]
2
Department of Informatics, Faculty of Humanities, Vilnius university, Lithuania,EU
3
Information Systems Department, Kaunas University of Technology, Lithuania, EU
Abstract The paper presents the details of trend trading algorithm in futures market. A contribution of this paper lies in a modified chart pattern related to a fractal formation, nonlinearity and chaos theory, broadly discussed by Benoit B.Mandelbrot and Bill M. Williams. As typical fractal pattern often is being applied in conjunction with other forms of technical analysis, like moving averages, Elliott Waves analysis or MACD indicators the proposed pattern is presented as a basic indicator itself. The strategy can be applied as up-trend market forecasting tool. The efficiency of the proposed strategy was tested with the most active North American futures contracts using 10 years historical daily data. Experimental results showed better returns if compared to overall market average- CRB index.
1
Introduction
Last decade expansion of high frequency trading, quantitative analysis and automated trading shows an increasing belief that algorithmic techniques can be helpful for decision making in financial markets [2, 4, 5]. Because of increase in market volatility and growing demand for hedging businesses trading volumes from stock and currency markets move to derivative markets [1]. The need for trading algorithms and quantitative analysis in futures markets are in great demand.
2
The basic concept of the proposed trading strategy
Despite the shortage of hard evidences about the profitability of trading algorithms there are some research papers claiming that the application of momentum trend trading and chart patterns can give some useful information to market participants [3, 9, 10, 11]. The proposed trading strategy is up-trend following system based on continuous bar chart formation.
The proposed continuous chart formation is related to fractal pattern and chaos theory, broadly discussed by Benoit B. Mandelbrot and Ph.D. Bill M.Williams [6, 11]. Decision to search for a new modified fractal was made because of inaccuracies when building back-testing trading strategies based on typical fractals. Inaccuracies appear because of a big variety of possible formations, lack of signal quality measurement and clear rules for closing positions. Typical fractal formations are most often presented in conjunction with other forms of technical analysis, like moving averages, Elliott Waves analysis or MACD indicator [11]. In general, typical fractals can be as a good decision support tool, but not the basic indicator itself. Decision to search for a more accurate, algorithmic, testable and still conformable with trend trading and chaos theory chart pattern was made.
2.1 Chart Pattern for Trading Strategy The proposed short term chart pattern is composed of only 3 consecutive chart bars instead of five. The proposed pattern uses a modified truncated fractal to generate a buy and sell signals. A fractal occurs when there is a pattern with the highest high in the middle and one lower high on each side, as it can be seen in some examples in Fig.1.
Fig. 1 Examples of the proposed chart pattern
The trading strategy opens a long position when the current price (i) tops the previous bar’s (i-1) highest price. If current price bar (i) opens higher than (i-1) and forms a gap, a buy signal is identified as an opening price. A trailing stop order is set as the lowest level of the bar (i-1) price minus a tick size. The trailing stop price is adjusted and moved to the next bar lowest price each time the new bar is formed. If there is a sudden crash in the market and the next bar opens with a gap down, sell price is set as an open price of this following bar.
2.2 Signal strength of the trading pattern In order to avoid inaccuracies related to taking only a particular contract or the most suitable time period for analysis the quality of signals and ranking techniques of contracts must be considered [7]. The proposed trading strategy is looking for futures contracts, which during two day’s period increased in a price the most, and the lowest price of the last bar is higher than a month ago. The biggest increase in
a price of a particular contract during the period between i-4 and i-2 means the best quality and the highest rank. If the chart pattern is formed with several contracts at a time, the one with the highest rank is chosen for a trade.
3. Experimental setup The efficiency of the strategy was tested with historical data from futures exchanges. The daily time series were collected from GLOBEX, NYMEX, ECBOT, CFE, ICE and ICE-NYBOT exchanges. Test period from 2002 till 2012. Most active futures were taken from these sectors: Energies, Metals, Grains, Financials, Indices, Currencies, Softs and Meats. Tick sizes, margin rates and commissions chosen according to Interactive Brokers LLC requirements and pricing structure. Risk ratio was taken 2% per trade of total capital. Backward adjusted data series were used for analysis. Backward adjusted data uses the actual prices of the most recent contract with a backward correction of price discontinuities for successive earlier active delivery months [8]. Considering that trading costs consist of commissions and also possible trade execution slippage, the size of a slippage was 3 ticks on every trade.
4. Experimental results The strategy was back tested applying MatLab software of technical computing. The experimental results of the strategy are presented in the following figure.
a)
b)
Fig.2 Total daily returns using trading strategy
Fig. 2 (a) shows the dynamics of total returns of the strategy. Total returns had outperformed the benchmark (Commodity Research Bureau- CRB Index) change and was nearly 250% at the end of testing period. Fig.2. (a) also shows that despite huge fluctuations in CRB Index (during 2008 market crash) the trading strategy generated rather stable results. These results confirm the capability of the fractal formation application in different market conditions. Fig.2(b) shows a comparison of daily returns, where the biggest daily drawbacks are only up to 4%, while some profitable days generate over 10% returns.
5 Conclusions In the paper experimental research of trading strategy based on truncated fractal formation and applied in futures market was presented. The research was carried out on daily historical data of the most active futures contracts in US markets. The results have given significantly better returns if compared to CRB index. The research has also shown rather stable results considering global economic fluctuations over the testing period. The strategy can be attractive for hedge funds or futures market participants who intend trading short term strategies or implementing risk management techniques, especially when volatility in markets increases.
Acknowledgment This research as Fellowship is being funded by the European Union Structural Funds project ”Postdoctoral Fellowship Implementation in Lithuania” within the framework of the Measure for Enhancing Mobility of Scholars and Other Researchers and the Promotion of Student Research (VP1-3.1-ŠMM-01) of the Program of Human Resources Development Action Plan.
References [1] Acworth Will (2011) “Record Volume 2010 (Annual Volume Survey)”, Futures Industry, March 2011: 12-29. [2] Avellaned M. (2002). Quantitative Analysis in Financial Markets. Collected Papers of the New York University Mathematical Finance Seminar, Vol.3. World Scientific Pub Co Inc, p.400 [3] Friesen, G.C., Weller P.A., Dunham L.M. (2009), Price trends and patterns in technical analysis: a theoretical and empirical examination. Journal of Banking&Finance, Vol. 33(6): 10891100. [4] Hautch N. (2008). Capturing common components in high-frequency time series: A multivariate stochastic multiplicative error model. Journal of Economic Dynamics and Control, Vol.32(12): 3978-4015. [5] Izumi K., Toriumi F., Matsui H. (2009). Evaluation of automated- trading strategies using an artificial market. Neurocomputing 72(16-18): 3469-3476. [6] Mandelbrot B.B., Gomory R.E., Cootner P.H., Fama E.F., Morris W.S., Taylor H.M. (2010). Fractals and Scaling in Finance: Discontinuity, Concentration, Risk. Springer, p.561. [7] Masteika S. (2010) “Short term trading strategy based on chart pattern recognition and trend trading in Nasdaq Biotechnology stock market”, Business Information Systems Workshops, Springer, May 2010, pp.51-57 [8] Masteika, S., Rutkauskas A.V., Alexander Janes A. (2012). Continuous futures data series for back testing and technical analysis. Conference proceedings, 3rd International Conference on Financial Theory and Engineering. IACSIT Press, vol.29, Feb 2012, pp. 265-269. [9] Miffre J., Rallis G. (2007). Momentum strategies in commodity futures markets. Journal of Banking&Finance, Vol.31(6): 1863-1886. [10] Szakmary A.C., Shen Q., Sharma S.C. (2010). Trend-following trading strategies in commodity futures: A re-examination. Journal of Banking & Finance, Vol.34(2): 409-426. [11] Williams B.M., Williams J.G. (2004), Trading chaos: maximize profits with proven technical techniques, 2nd edition, John Wiley & Sons, Inc., p.228.
