igi global proof

54 International Journal of Interdisciplinary Telecommunications and Networking, 4(4), 54-63, October-December 2012

Empirical Case Study of Binary Options Trading:

An Interdisciplinary Application of Telecommunications Methodology to Financial Economics Gaetano Giunta, Signal Processing for Telecommunications and Economics (SP4TE) Lab, University of ROMA TRE, Rome, Italy Francesco Benedetto, Signal Processing for Telecommunications and Economics (SP4TE) Lab, University of ROMA TRE, Rome, Italy

ABSTRACT

IGI GLOBAL PROOF

This paper presents an interdisciplinary application of information & communication methodology to financial economics. The empirical case study reported in this contribution consists of a preliminary example of binary options stock trading. The authors have investigated the performance of a simple algorithm which includes one buy/sell order per week. They have analyzed real sets of historical stock quotes, evidencing the asymmetry of achievable economic returns. In fact, the devised algorithm has denoted a (simulated) overall trading gain in the 87% of cases. A discussion, correlating such trend to the typical behavior of occasional traders, is finally reported Keywords:

Binary Options Stock Trading, Emerging Telecommunications Applications, Financial Data Extrapolation, Financial Engineering, Network Platforms for Internet Banking, Statistical Analysis

INTRODUCTION AND MOTIVATIONS In the last years, there has been an explosive growth in the research area relating economics and mathematical modeling (Gradojevic & Gencay, 2011; Bekiros, 2011), especially in the fields of business and banking researches and applications (Nair et al., 2010; Liu & Xiao,

2009; Taskaya & Ahmad, 2003; Nuti et al., 2011). However, it is more important to dynamically follow the non-stationary processes’ fluctuations to provide optimal information for automatic trading, than statistically modeling data sequences (Ehlers, 2001; Zhang & Kedmey, 2011). In fact, stock traders usually try to profit from short-term price volatility with trades lasting anywhere, from several seconds to several weeks (Wikipedia, 2012a).

DOI: 10.4018/jitn.2012100104 Copyright © 2012, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited.

International Journal of Interdisciplinary Telecommunications and Networking, 4(4), 54-63, October-December 2012 55

The increasing development of on-line trading and Internet banking have boosted the growth of proprietary methods (Slamka et al., in press). In this sense and according to (Tsakalozos et al., 2011), we do not know what we do not know. Among the proprietary algorithms based on heuristic concepts, one popular trading strategy is binary option trading (Raw, 2008). In finance, a binary option is a type of option where the payoff is either some fixed amount of some asset or nothing at all (Wikipedia, 2012b). We have searched through the literature and we have found that a Google search with the queries ´“Telecommunications” + “binary option”´ returns zero results (Google, 2012). In most specific database of IEEE Xplore, there are two papers found with the query “binary options” (IEEE, 2012): namely Wang et al. (1998) and Yuan and Xiao (2011). But the paper (Wang et al., 1998) deals with binary options meaning the classical binary hypothesis testing. Hence, Yuan and Xiao (2011) is the only paper about binary options from the economic viewpoint, in the IEEE database. In particular, the authors in Yuan and Xiao (2011) present a new numerical method for pricing binary options, showing with numerical examples that the proposed algorithm is conditional stable and convergent. The binary options trading is becoming more and more popular because it provides easy indications to operate in a dynamic manner, and suited to occasional operators that usually trade by their remote Internet on-line platforms. This paper aims to highlight the huge possibility in exploiting telecommunication methodologies in synergy with stock trading, showing an empirical case study. We propose to exploit the hidden market trends of stock prices (Drakakis, 2009) for application to stock trading. In particular, our case study consists of a preliminary example of trading stocks with a simple algorithm for binary options, which includes one buy / sell order for week of a fixed amount of cash (or exchange equivalent) to limit the maximum risk of the investment to that fixed amount. The advantage of using a simple trading

model reflects in the fact that the performance evaluation is straightforwardly based on the return in terms of cash and stock portfolios at the end of the trading session. The remainder of this work is organized as follows. First we detail the proposed data processing technique for financial stock trading, and then we discuss a case study reported along with the results’ discussion. Our discussion and conclusions are finally depicted in the last section.

DATA PROCESSING ALGORITHM FOR FINANCIAL STOCK TRADING In this section, we explain the data processing methodology we have applied on financial signals. First of all, we have analyzed financial stocks using different data processing techniques, typical of telecommunications signals, to extrapolate the hidden periodicities. In particular, we have exploited cumbersome methods, such as the multiple signal classification (MUSIC) (Kay & Demeure, 1984), Pisarenko, and Prony algorithms (Kay & Marple, 1981), as well as simpler techniques such as statistical correlation / covariance analysis, and Fourier transform-based methods (periodogram) (Oppenheim & Schafer, 1975). Each of the previous-mentioned techniques has been applied to different ensembles of financial stocks and we have observed the same trend: there is a positive correlation with a distance of 1 day and a negative correlation with a distance of 1 week from the maximum. Other correlations exist but they are random and characterized by more non-stationarity. In a recent development, an interesting technique was proposed and applied on financial signals to capture the sign of the increment of the signal instead of the exact future value (Tsakalozos et al., 2011). The authors in Tsakalozos et al. (2011) use the Empirical Mode Decomposition (EMD), which offers a Fourier series-like expansion of any

IGI GLOBAL PROOF

Copyright © 2012, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited.


(discrete) signal. Here, we move further using the phase of the predicted value, in order to buy/sell financial stocks, accordingly. The authors in Tsakalozos et al. (2011) deal with short-term (hourly) periodicities, while here we are exploiting long-term (weekly) periodicities. It has also to be noted that the work in Tsakalozos et al. (2011) is well suited for regular traders that hourly observe the stock values. Conversely, our method is well suited for an occasional on-line trading (the observations of the stocks are once a day, the decisions are once a week). In agreement with the authors in Tsakalozos et al. (2011) that suggest exploiting a Fourier series-like expansion of the financial data, here we have decided to focus on simpler extrapolation approaches (based on statistical autocovariance and discrete Fourier transform). In particular, given a random series x(n) of N data, its auto-covariance is defined as follows (Oppenheim & Schafer, 1975): Covx (i ) =

2 1 N −1 ⋅ ∑ x (n ) ⋅x * (n − i ) − µ (1) N n =0

IGI GLOBAL PROOF

where x*(n) means complex conjugate, i = 0, ±1, ±2, …, and the mean μ is expressed as: µ=

As an example, the two-week autocovariances (Figure 1a and Figure 1b) and periodograms (Figure 1c and Figure 1d), averaged over fifty weeks, are here shown for two financial stocks. The negative correlation for lags of about one week (five working days) is evident in several of the analyzed financial stocks. In the following of this manuscript, we assume that one stationary periodicity of the financial stocks is characterized by a two-week period. This assumption is supposed to be valid with the unique purpose of providing a simple tool for the binary options trading and for easily evaluating its performance. The optimization of the extrapolation algorithm is out of the scope of this work, which aims instead to highlight the synergy between signal processing for telecommunications and financial trading, discussing a preliminary educational case study. The rationale of our algorithm is as follows. Given a stock price sequence of two weeks, our method suggests to buy/sell a fixed amount of the stock on a predicted day in the incoming week. To this aim, two weeks of five working days are stored with one sample per day, corresponding to 10 samples. Hence, we evaluate the 10-points DFT and extract the frequency coefficient K = 1, corresponding to the sinusoidal period of two weeks (the maximum and minimum of the sinusoid are spaced one week apart). Then, we evaluate the phase of the Fourier coefficient K = 1, since it determines the days in which a maximum or a minimum has occurred in the last two weeks. The extracted phase information represents the extrapolated information we use in the next week, to forecast a probable maximum or minimum of the stock prices, assuming the (likely) stationarity of the series under investigation. Now, we extrapolate in the next week the sinusoid (with a two-week period) based on the sinusoid extracted from the previous two weeks and with the same phase. The pictorial Figure 2a and Figure 2b illustrate the extrapolated sinusoid of the next (third) week, for the case in which a maximum (action: sell) or a minimum (action: buy) occurs, respectively.

1 N −1 ⋅ ∑ x (n ) N n =0

(2)

The sequence of N complex numbers x(0), …, x(N-1) is transformed into another sequence of N complex numbers X(0), …, X(N-1) according to the discrete Fourier transform (DFT): N −1

X (K ) = ∑ x (n ) ⋅ e

−i ⋅2 π ⋅

K ⋅n N

(3)

n =0

where K is the frequency index. Finally, the periodogram is obtained by means of the following (Oppenheim & Schafer, 1975): 2

PX (K ) = X (K )

(4)



Figure 1. Two-week autocovariances (Figure 1a and Figure 1b) and periodograms (Figure 1c and Figure 1d), averaged over fifty weeks, for two financial stocks

IGI GLOBAL PROOF Figure 2. Extrapolated sinusoid of the next (third) week for the case in which: (a) A maximum, (b) Or a minimum will occur

Therefore, in the following (third) week, one half period of a sinusoid can be identified. This extrapolated sinusoid has always one maximum or one minimum, according to the

early computed phase. This information will now be used by the binary options trader, by selling or buying on the exact day of the maximum or minimum, respectively.



CASE STUDY REPORT In this section, the sensitivity and the efficiency of the proposed approach for financial trading are evaluated. In particular, to analyze the performance of our method, we have used

real sets of historical stock prices, taken from Finance.Yahoo (2012). We have decided to use different sets of financial stocks of multinational manufacturers, i.e., the top 5 most stable (Table 1) stocks according to Investopedia (2012), namely: Basic Materials, Transportation, Con-

Table 1. Investment return report for the analyzed stocks

IGI GLOBAL PROOF



sumer Goods, Technology, and Healthcare. This choice is motivated by the fact that these sets should be less affected by price fluctuations and financial storms. In this way, we give guess to be able to assume the near stationarity of the analyzed data sets. The names of the companies belonging to these five areas, whose stocks are here analyzed, are listed in alphabetic order as follows: 1. Hot stocks in basic material (Investorguide, 2012): Aegean Marine Petroleum Network, Crown Cork & Seal Company, Goodyear Tire & Rubber Company, Greif Bros., L.B. Foster Company, Meadwestvaco, Shiner International, Tredegar; 2. Top-rated stocks in the transportation sector (MSN, 2012): Allegiant Travel, CSX Corp, Grupo Aeroportuario ADR, Heartland Express, J.B. Hunt Transport Services Inc, Kirby Corp, Norfolk Southern Corp, Ryanair Holdings ADR, TAL International Group Inc, Union Pacific Corp;

or, alternatively, we have simulated a buying decision trading for an exchange equivalent of always 1000 cash units. The buying or selling operations are performed, for the sake of simplicity, at the opening quotes of the day in which the extrapolated maximum (action: sell) or minimum (action: buy) occurs. Obviously, the initial portfolio must allow the next buying or selling actions to the investor (Table 2). For each analyzed stock, Table 1 shows the results of the number of performed selling and buying actions, of the cash and stock profit, and of the overall return of the investment (with sign) in cash units. In particular, the overall return of the investment1 has been evaluated adding both the residual cash units and quotes equivalent and subtracting the initial investment. The stocks are anonymized to meet the Finance.Yahoo requirements on disclosure and data responsibility2. It can be easily seen that the proposed method reaches an overall gain in the 87% of the analyzed stocks, i.e., in 33 cases out of 38. The overall statistics3 of the number of selling actions, of the cash and stock profit, and of the overall return are reported in Table 2. Moreover, Figure 3 visually depicts these statistics by means of histograms and fitting curves. In particular, Figure 3a refers to the histogram of the number of selling actions, Figure 3b and Figure 3c show the statistics of the cash and stock profit, respectively. Finally, Figure 3d represents the statistical behavior of the overall return. It has to be noted that the histograms related to the stock profit, as well as the one referring to the overall return are really skewed, justifying the overall gain reached in 33 cases out of 38 (87%).

IGI GLOBAL PROOF

3. Popular consumer goods companies (Moderninvesting, 2012): Coca-Cola, Johnson & Johnson, PepsiCo, Procter & Gamble, Under Armour; 4. Technology area, best selling mobile phones (Wikipedia, 2012c): Apple, HTC, Huawei, LG, Motorola, Nokia, Samsung, Sony Ericsson, RIM, ZTE; 5. More stable healthcare stocks (Seekingalpha, 2012): Exilixis Inc, Momenta Pharmaceuticals Inc, Neurocrine Biosciences Inc, Syneron Medical Ltd, Thoratec Corp.

For each of the previous-mentioned stocks, we have analyzed one year of historical data (stock prices) from April, 1st 2011 to March, 31st 2012, considering the opening stock quotes of five working days per week. We have simulated the management of a portfolio, equally made of cash and stocks. Then, for each week of the year, we have simulated a selling trading decision for a number of stocks equivalent to 1000 cash units

DISCUSSION AND CONCLUSION We have presented a preliminary educational case study exploiting a simple algorithm of data processing for binary options, which includes one buy / sell order per week. The devised method has denoted a (simulated) overall trading



Table 2. Overall statistics of the investment return report

Figure 3. Histograms of: (a) number of sell actions; (b) stock profit ($); (c) cash profit ($); (d) overall return ($)

IGI GLOBAL PROOF



gain in the 87% of cases, analyzing real sets of historical stock quotes. We have not focused on cumbersome and sophisticated data processing methods to extract hidden periodicities in the analyzed stocks. Indeed, we were interested in verifying the presence of a (stable) alternate periodicity of about one week. Obviously, other periodicities (of different periods) exist in the stock prices and may be exploited, but we believe that the weekly periodicity is the more stable one since it is typical of the behavior of occasional customers that use on-line trading platforms. According to our opinion, the (inverse) weekly periodicity related to occasional traders, is accountable to the following causes: 1. Behavioral and sociological reasons: occasional (non regular) customers typically use on-line trading platforms just once a week (not daily), since trading is not their main job; 2. Occasional customers have reduced cash availability: hence, they want to realize profit as soon as possible, typically try to retrieve their money by selling the stocks in the first week after they have performed a buying action; 3. The delay of the banking transactions. In fact, occasional customers have to wait 2-4 days before the banks credit to their accounts the exchange stock equivalent after the selling operation. This introduces an intrinsic delay in the selling/buying chain, justifying the alternate correlation at a distance greater than 3 days;

REFERENCES Bekiros, S. D. (2011). Sign prediction and volatility dynamics with hybrid neurofuzzy approaches. IEEE Transactions on Neural Networks, 22(12-2), 2353-2362. Drakakis, K. (2009). Application of signal processing to the analysis of financial data. IEEE Signal Processing Magazine, 26(5), 160–158. doi:10.1109/ MSP.2009.933377 Ehlers, J. F. (2001). Rocket science for traders: Digital signal processing applications. New York, NY: John Wiley & Sons. Google. (2012). Binary options and telecommunications. Retrieved July 3, 2012, from http://www. google.com/search?q=binary+options&ie=utf8&oe=utf-8&channel=fs#hl=en&gs_ nf=1&pq=binary%20options&cp=1&gs_ id=28&xhr=t&q=%22binary+options%22+%2Bte lecommunications&pf=p&hs=Y8Y&channel=fs&s client=psy-ab&oq=%22binary+options%22+%2Bte lecommunications&fp=1&bav=on.2,or.r_gc.r_pw.r_ qf.&cad=b&sei=VmZgUN_zEs-yhAeXpYHACw Gradojevic, N., & Gencay, R. (2011). Financial applications of nonextensive entropy. IEEE Signal Processing Magazine, 28(5), 116–141. doi:10.1109/ MSP.2011.941843

IGI GLOBAL PROOF

As a final conclusion, we believe that telecommunications methodologies fruitfully apply to financial stock trading, according to the preliminary results analyzed in the presented case study. Further investigations will be useful to define the most appropriate and efficient forecasting models to adopt for stock trading.

Investopedia. (2012). Top 5 most stable stocks. Retrieved July 3, 2012, from http://www.investopedia.com/financial-edge/0611/Top-5-Most-StableStocks.aspx Investorguide. (2012). Basic materials. Retrieved July 3, 2012, from http://www.investorguide.com/ sector/Basic-Materials Kay, S. M., & Demeure, C. (1984). The highresolution spectrum estimator—A subjective entity. Proceedings of the IEEE, 72(12), 1815–1816. doi:10.1109/PROC.1984.13091 Kay, S. M., & Marple, S. L. Jr. (1981). Spectrum analysis—A modern perspective. Proceedings of the IEEE, 69(11), 1380–1419. doi:10.1109/ PROC.1981.12184 Liu, Z., & Xiao, D. (2009). An automated trading system with multi-indicator fusion based on D-S evidence theory in Forex Market. In Proceedings of the 6th International Conference on Fuzzy Systems and Knowledge Discovery (pp. 239-243).



Moderninvesting. (2012). Investing in consumer goods stocks. Retrieved July 3, 2012, from http:// www.moderninvesting.net/investing-in-consumergoods-stocks/ MSN. (2012). Top-rated stocks. Retrieved July 3, 2012, from http://investing.money.msn.com/investments/ industry-top-stocks?Category=Sector&Choice=7 Nair, B. B., Mohandas, V. P., Sakthivel, N. R., Nagendran, S., Nareash, A., & Nishanth, R. …Manoj Kumar, D. (2010). Application of hybrid adaptive filters for stock market prediction. In Proceedings of the International Conference on Communications and Computational Intelligence (pp. 443-447). Nuti, G., Mirghaemi, M., Treleaven, P., & Yingsaeree, C. (2011). Algorithmic trading. IEEE Computer, 44(11), 61–69. doi:10.1109/MC.2011.31 Oppenheim, A. V., & Schafer, R. W. (1975). Digital signal processing. Upper Saddle River, NJ: Prentice Hall. Raw, H. (2008). Binary options: Fixed odds financial bets. St. Albans, UK: Harriman House. Seekingalpha. (2012). 5 overlooked healthcare stocks to consider now. Retrieved July 3, 2012, from http:// seekingalpha.com/article/376181-5-overlookedhealthcare-stocks-to-consider-now

Wikipedia. (2012a). Stock trader. Retrieved July 3, 2012, from http://en.wikipedia.org/wiki/Stock_trader Wikipedia. (2012b). Binary option. Retrieved July 3, 2012, from http://en.wikipedia.org/wiki/ Binary_option Wikipedia. (2012c). List of best-selling mobile phones. Retrieved July 3, 2012, from http:// en.wikipedia.org/wiki/List_of_best-selling_mobile_phones Yahoo. (2012). Finance. Retrieved July 3, 2012, from http://finance.yahoo.com Yuan, G., & Xiao, Q. (2011). A new numerical method for pricing binary options in the CEV process. In Proceedings of the International Conference on E-Business and E-Government (pp. 1-4). Zhang, X.-P., & Kedmey, D. (2011). TechWare: Financial data and analytic resources. IEEE Signal Processing Magazine, 28(5), 138–141. doi:10.1109/ MSP.2011.941840

ENDNOTES

IGI GLOBAL PROOF

Slamka, C., Skiera, B., & Spann, M. (in press). Prediction market performance and market liquidity: A comparison of automated market makers. IEEE Transactions on Engineering Management.

1.

2.

3.

Taskaya, T., & Ahmad, K. (2003). Bimodal visualisation: A financial trading case study. In Proceedings of the International Conference on Information Visualization (pp. 320-326). Tsakalozos, N., Drakakis, K., & Rickard, S. (2011). Signal extrapolation using Empirical Mode Decomposition with financial applications. In Proceedings of the International Conference on Acoustics, Speech and Signal Processing (pp. 5744-5747). Wang, X.-G., Shen, H. C., & Qian, W.-H. (1998). A hypothesis testing method for multisensory data fusion. In Proceedings of the International Conference on Robotics and Automation (pp. 3407-3412).

The evaluated investment return is a gross return, i.e., for trading orders before-tax and operating costs. To obtain the value of the net return after-tax, one should subtract also the surcharge fees related to the use of the on-line trading platform. From (Yahoo, 2012): “All information provided “as is” for informational purposes only, not intended for trading purposes or advice. Neither Yahoo! nor any of independent providers is liable for any informational errors, incompleteness, or delays, or for any actions taken in reliance on information contained herein. By accessing the Yahoo! site, you agree not to redistribute the information found therein.” The percentile values in Tab. 2 are numerically interpolated. The statistics related to the number of buying actions are not reported since they can be easily evaluated from the ones of the selling actions (i.e., number of buying actions = 50 – number of selling actions)



Gaetano Giunta received the Electronic Engineering degree from the University of Pisa, Italy, and the PhD degree in Information and Communication Engineering from the University of Rome La Sapienza, Italy, in 1985 and 1990, respectively. In 1986, he obtained a research grant from the Italian Research Council (CNR) of Pisa, Italy. He was also (since 1989) a Research Fellow of the Signal Processing Laboratory (LTS), Swiss Federal Institute of Technology (EPFL), Lausanne, Switzerland. In 1992, he became an Assistant Professor with the INFO-COM Department, University of Rome La Sapienza. Since 1998, he has taught Digital Signal Processing in the Third University of Rome. From 2001 to 2005, he was with the Third University of Rome as an Associate Professor of Telecommunications. Since 2005, he has been a Full Professor of Telecommunications with the same University. He is currently the director of the Signal Processing for Telecommunications and Economics Laboratory at the Third University of Rome. His research interests include signal processing for mobile communications, video communications and security, spread-spectrum systems, satellite and wireless networks. Prof. Giunta has been a representative member of CNIT (Italian Inter-Universities Consortium for Telecommunications) and the IEEE Societies of Communications, Signal Processing, and Vehicular Technology. He has also served as a reviewer for several IEEE transactions, IET (formerly IEE) proceedings, and EURASIP journals, and a TPC member for several international conferences and symposia in the same fields. Francesco Benedetto was born in Rome, Italy, in 1977. He received the Dr.Eng. Degree in electronic engineering and the PhD degree in telecommunication engineering from the Third University of Rome, Italy, in May 2002 and April 2007, respectively. In 2007, he was a research fellow of the Department of Applied Electronics of the Third University of Rome. Since 2008, he has been an Assistant Professor of Telecommunications at the Third University of Rome (2008-2012, Applied Electronics Dept.; 2013-present, Economics Dept.). In particular, he has published numerous research articles on multimedia communications and video coding, ground-penetrating radar (GPR) signal processing, spread-spectrum code synchronization for 3G communication systems and satellite systems (GPS and GALILEO), correlation estimation, and spectral analysis. His research interests are in the field of digital signal and image processing in telecommunications, code acquisition, and synchronization for the 3G mobile communication systems and multimedia communication. Dr. F. Benedetto is a reviewer for the IEEE Transactions on Communications, the IEEE transactions on vehicular technology, the IEEE Transactions on Geoscience and Remote Sensing, the IEEE Transactions on Wireless Communications, and the IEEE Communications Letters. He is a member of the TPC (Technical Program Committee) of the European International Conference on Signal Processing (EUSIPCO), and he also served as a reviewer for many IEEE International Conferences.

IGI GLOBAL PROOF


igi global proof

igi global proof

Suggest Documents

igi global proof

igi global proof

igi global proof

igi global proof

igi global proof

igi global proof

igi global proof

igi global proof - iPEAR

igi global proof

igi global proof

igi global proof

igi global proof - iPEAR

igi global proof

igi global proof

igi global proof

igi global proof

igi global proof

IGI Global

igi proof

IGI Global

collaborative design - IGI Global

Preface - IGI Global

Preface - IGI Global

Preface - IGI Global