The author's abstract of master's work
Content
1. Introduction
2. Work actuality
3. Scientific novelty
4. Work purpose and tasks
5. Practical results
6. The review of existing researches
7. Global overview of research
8. National survey of research
9. Review of research in DonNTU
10. Description of own researches
11. Conclusions
12. Literature
Introduction
In technical analysis, stock market, there are many ways to identify market trends and he forecasting. While professionals and a prefer one type of them, but none of the indicators does not give one hundred percent reliable answer about what will happen to the market in the future. Each indicator has its advantages and disadvantages, which in various situations are crucial.
Indicator – series of data points, which were obtained by applying the formula to price for data. Price data includes any combination of open (open), the highest value (high), the lowest value (low) or closing (close) over time. Some indicators may use only the closing price, while others include in their formula volumes and open interests. [1]
Indicators serve for three broad functions: prevention, validation and prediction.
The indicator may act as a warning to study price action more closely. If the pulse (driving force) is reduced, it may be a signal that can be expected to break through support. Or, if formed a large positive divergence (divergence), it can serve as a warning that necessary to watch for a sharp break through resistance.
Indicators can be used to confirm other technical analysis tools. If there is a sharp change in price on a price chart, corresponding to the intersection of the moving average could serve as a confirmation of this breakthrough. Or, if market breaks through a support line, corresponding to Low on the chart indicator On–Balance–Volume (OBV), that could serve as evidence of weakness. Some investors and traders use indicators to predict the future direction of change in prices.
Work actuality
To maintenance profitable trades the trader must be understood for what markets (more or less volatile), for some situations (separation of the ascending or descending trend for a certain period of consolidation, integration points of correction) is correct or that indicator. You need to know how to combine technical indicators in trading system. Need to examine methods of forming a single trading signal based on the signals of several indicators, and to choose the best according to some criterion. [2,3]
It is also important to understand that over time the market situation is changing. Those decisions, which are most profitable today may be ineffective tomorrow. Therefore, the most important area of research in the field of technical analysis is to develop algorithms that allow you to use the same trading system with the same set of technical indicators and their identical parameters in different market situations. [4]
Relevance of the topic of master's theses are due absence by effective computer trading systems, which have a system of monitoring indicators. Such a system "includes" indicators, which are have "good" specifications and "off" indicators with "bad" characteristics.
Scientific novelty
There are currently more than 3000 technical indicators, which are not effective at certain times. Therefore, in this paper will be developed computerized trading system containing control system market indicators, which will form for generate a "reliable" trading signals.
Work purpose and tasks
The purpose of this study is to develop a computer trading system (CTS), which maximizes the income and other economic indicators while reducing investment risks. A special feature of this CTS is that there exists a subsystem, which manages the market indicators. To achieve this goal it is necessary to perform certain tasks to find the parameters used for the development of CTS, namely:
- to develop and optimize the simplest CTS;;
- to develop and optimize its management subsystem trading position;;
- to make a comparison of computer trading systems commonly used criteria:
- income at the end of the reporting period;
- the standard deviation of return;
- the number of transactions during the reporting period;
- Profit–Factor (PF);;
- TWR.
Subject development and research: algorithms technical market indicators:
Object development and research: the task of creating an algorithm operational control.
Practical results
The practical results of this work will be to develop algorithms for real–time monitoring of technical market indicators. This algorithm can be applied to reduce risks and increase profitability in order to optimize policy traders.
The review of existing researches
Global overview of research
Very close to the subject of my master's thesis design by John Bollinger. He took a market indicators, namely the MACD, MA, RSI, and modified the algorithm to optimize for trading signals. [5]
National survey of research
In connection with the recent emergence of the stock market in Ukraine, which implies a lack of experience, so this topic has not received adequate coverage in studies of local scientists, with the exception of assistant professor, Cand.Tech.Sci Smirnov A., Gizatulin A., Revega D. and Guryanova T.
Review of research in DonNTU
In DonNTU conducted research assistant professor of PMI, Cand.Tech.Sci Smirnov A. together with Revega D. and Gizatulin A. The results work of Smirnov A. and Revega D. described in the article "Methods of selecting indicators of technical analysis of markets for their subsequent integration".
Description of own researches
The aim of the study is to develop and study algorithms for real–time monitoring of technical market indicators. The initial data are taken real stock chart. Therefore had to take an array of closing prices of daily bars for 5 years of the currency pair EUR/USD, since the beginning of 2002 to the end of 2007
For research the trading system, built on a technical indicator used MACD.
The MACD is one of the easiest and most reliable indicators. MACD uses moving averages, which are lagging indicators, and includes some characteristics of trend–following. These lagging indicators are transformed into a pulse oscillator by subtracting the longer moving average from the shorter moving average. The resulting graph forms a line that oscillates above and below zero, without any restrictions from the top and bottom. MACD is a central oscillator so you need to apply the basic rules for the central oscillator. [6,7]
The formula for the "standard" MACD represents the difference between the 26–day and 12–day exponential moving averages. Using shorter moving averages will provide a faster, more sensitive indicator, and using longer moving averages will provide a slower indicator, less prone to rapid reversal. To investigate the trading system will be used by the traditional 12/26 MACD.
Signals inputs:
1. If the MACD line is above zero and crosses the signal line in the upper side, then buy tomorrow at the opening.
2. If today's MACD crosses the zero line in the lower side, then buy tomorrow at the opening.
Signals input
1. We leave a long position, if the MACD line crosses the signal line in the bottom.
2. We leave a long position, if the MACD line crosses the signal line in the upper side. [8]
The figures show the graph of stock exchange prices and values of the oscillator for the first two months of trading
Figure 1 – Bars currency pair EUR\USD for January – February 2002.
Figure 2 – Graph indicator MACD
The next stage of the algorithm is to assess the economic efficiency of each individual indicator. When selecting the most efficient trading system used indirect indicators of its quality, this is usually the average profitability and average risk. The best is a system with a maximum income or minimum risk. However, the effective operation of the trading system in the past does not work effectively in the future. Thus, the error of this approach is not excluded. It is therefore necessary for each indicator to hold a series of trades throughout the analyzed period. Short positions are taken in line with the indicators, and long in accordance with a specially optimized for the length of the counter trend, the length is 6 trading days. In parallel to the work of each indicator trades are constructed income curves and assessed the economic efficiency of each system, based on only one of the oscillator.
Figure 3 – The total income (size animation: 121 KB (124,858 bytes), resolution animation: 483 x 291 pixels, number of repetitions: 7, number of staff: 6, implemented in MS Gif-animator)
Thus, the most profitable was CTS with N = 6. Her average income at the end of the reporting period (Dav) amounted to 69,452 den. units.
Figure 4 – The average income at the end of the reporting period
To this computerized trading system was implemented score based on the following indicators:
1. The ratio of average profit margins of the average size of losses in the base period:
where:
P – total profits for the base period;
L – total loss in the base period;
Np – the number of profitable transactions in the base period;
Nu – the number of unprofitable trades in the base period;
No = Np + Nu – total number of transactions in the base period.
2. The ratio of the number of profitable trades to their total number in the base period:
Application of MTS has practical meaning when К2 >= 0,6..0,7. It tends to the value of this index at the level of К2 >= 0,9..0,95.
3. Profit-Factor (РF) – product of the coefficients К1 и К2:
The economic meaning of РF – this is the mathematic expectation ratio of profit to of the average size of loss for the period (К2 determines the frequency of profitable trades over this period).
4. Coefficient Sharp:
where:
Cs – Sharpe ratio (r=10%) (in the expression for Cs is set equal to I = 10%);
Dav – average income at the end of the reporting period;
I – value of bank interest.
5. SD – standard deviation of the financial results of the MTS (in percent):
where:
n – the number of trades during the reporting period;
Пi – current income CTS;
Пavi – average income at the end of the reporting period. [9]
6. TWR:
where:
Final capital = Initial capital + Dav. [10]
Table 1 shows the comparative characteristics of the functioning of trading systems based on the MACD.
Table 1 – Characteristics of the system based on MACD
| System characteristic
| 1
| 2
| 3
| 4
| 5
|
| Number of trades
| 16
| 11
| 17
| 15
| 8
|
| Number of profitable trades
| 9
| 7
| 8
| 4
| 3
|
| Number of loss trades
| 7
| 4
| 9
| 11
| 5
|
| K1
| 0.922
| 1.11
| 1.188
| 1.035
| 0.954
|
| K2
| 0.563
| 0.636
| 0.47
| 0.267
| 0.357
|
| Profit-Factor
| 0.519
| 0.707
| 0.559
| 0.276
| 0.358
|
| TWR
| 1.0712
| 1.0887
| 1.1209
| 1.1033
| 1.0793
|
| Mean profit
| 7828.856
| 7327.65
| 14717.927
| 10957.2
| 6966
|
| Average risk
| 1078.995
| 1245.063
| 1313.587
| 706.752
| 1308.393
|
| Sharpe ratio (r=10%)
| –2.012
| –2.146
| 3.59
| 1.3543
| –2.319
|
As you can see, these ratios give quite contradictory assessment: some talk about high efficiency, while others not give enough efficiency. This is explained by the fact that they are used in various fields and have different reliability.
Conclusions
In the transition to understanding the profitability curve as a function of quantity traded, and thus the possibility of using mathematical methods of analysis functions is an important step in assessing the efficiency of trading systems.
The proposed algorithm estimates the current quality of the trading system (represented in the final work) can be applied in practice to reduce the investment risk.
In some cases, risk reduction may be accompanied by increase the profitability of trades by reducing the number of losing trades and the amount of maximum loss.
However, this algorithm can only identify the periods of inefficient operation of trading systems and disable them. Traders can not afford to simply withdraw from the bidding and wait for better times in the game. Therefore, the next stage of research should be to develop an algorithm which enables switch used trading systems, depending on the current assessment of the quality of their work.
Literature
1. Алексис С. Б. Технический анализ от «А» до «Я» /Пер. с англ. – М.: Диаграмма, 1999 – 274 с.
2. Швангер Д. Технический анализ. Полный курс: Пер. с англ. – М.: Альпина, 2001 – 1263 с.
3. Tushar S. Chande «Beyond Technical Analysis: How to Develop & Implement a Winning Trading System» – USA, 1997
4. Колби Р. Энциклопедия технических индикаторов рынка. Пер. с англ. – 2–е изд. – М.: Альпина Бизнес Букс, 2004 – 647с.
5. Аппель Дж. Технический анализ. Эффективные инструменты для активного инвестора, Питер, 2007 – 504с.
6. Шнайдер М. Анализ валютных систем/ Дилинговый центр "Pro Finance Group Inc." – http://www.pfgfx.ru
7. Таран В. А. Играть на бирже просто?! – М.: Питер, 2008. – 540 с.
8. Рашид Умаров. Математика в трейдинге/ MQL4 – http://articles.mql4.com
9. Смирнов А. В. Методические указания и задания по курсу «Технический анализ рынков» – Донецк: ДонНТУ, 2003г.
10. Люк Ван Хоф. Оценка эффективности торговой системы/ FX Guild – http://www.fxguild.info
* — При написании данного автореферата магистерская диссертация еще не завершена. Окончательное ее завершение состоится в декабре 2010 г. Текст и материалы диссертации могут быть получены у автора или его руководителя после этой даты.
* — The work is in development for the current day and is planned to be finished by December 1st, 2010. The paper and relevant materials can be obtained either form author or its supervisor after that date.
|
RUS 
UKR 