Can I get some technical trading strategies
The performance of technical trading strategies. A literature review
Table of Contents
LIST OF FIGURES
LIST OF TABLES
LIST OF ABBREVIATIONS
2 TECHNICAL ANALYSIS
2.2 Efficient market hypothesis
2.3 Technical trading strategies
2.3.2 Chart patterns
2.3.3 Moving Average
3 TEST METHODS AND KEY FIGURES
3.1 Hypothesis test
3.2 Bootstrapping method
3.5 Sharpe ratio
4 PROFITABILITY OF TECHNICAL TRADING STRATEGIES
4.1 Profitability of trading strategies in stock markets
4.2 Currency market
6 LITERATURE LIST
List of figures
Figure 1: Random walk around the rational price level 5
Figure 2: Trend reversal formation - shoulder-head-shoulder formation 10
Figure 3: Trend Continuation Formation Flag 11
Figure 4: Trading strategies with the RSI oscillator 13
Figure 5: Hypothesis test - Rejection areas 17
Figure 6: Hypothesis test with t and p values 20
List of tables
Table 1: Evaluation of the 2580 technical trading strategies in the S&P 500 26
Table 2: Evaluation of the 2580 technical trading strategies in the S&P 500 futures market 27
Table 3: Evaluation of the 2580 technical trading strategies in the S&P 500 futures market on 30 minutes of data 28
Table 4: Portfolio returns 31
Table 5: Portfolio properties based on the three-factor model 33
Table 6: Trade results to developed countries 37
Table 7: Trade results with developing countries 38
List of abbreviations
Figure not included in this excerpt
Trading in technical trading signals on the stock exchanges can look back on a long history. As early as the 18th century, Japanese rice traders developed technical trading strategies to forecast prices. The candle chart is used for this technical analysis. The advantages of these charts, as opposed to ordinary line charts, are that they contain more information. You can read the high, low, opening and closing prices for each period. These advantages have been unknown to the West for a long time. This is why the first introductory literature enjoyed broad support among analysts and course speculators (cf. Nison, 1991, p. 1).
By applying technical trading strategies, one tries to generate price trends or optimal entry and exit points on the basis of various indicators, oscillators and chart patterns.
Technical trading strategies are widespread among large and small investors. According to the empirical study by Cheol-Ho Park and Scott H. Irwin in 2007, professional analysts use at least one technical trading system to support their buying and selling decisions (cf. Park / Irwin, 2007, p. 787).
This bachelor thesis deals with the profitability of such trading strategies. The main interest is in determining whether technical trading strategies are profitable. For this purpose, the various results of the empirical studies are shown.
At the beginning, the most commonly used technical trading systems are presented. This includes trading with momentum, chart pattern, moving average and filter strategies. Chapter 2 presents both the basic principle of technical analysis (TA) and the efficient market hypothesis (EMH). For more than 40 years, this hypothesis has represented the thesis that our financial markets work efficiently and that even in the weak form it is not possible to use historical prices to make forecasts about future price developments.
In chapter 3 the test methods and key figures necessary for the elaboration of the empirical studies are presented. The first insights into the profitability of technical trading strategies were made through hypothesis testing. This test was intended to determine whether the results were significant or whether they could have come about purely by chance.
Most recent studies, however, indicate that the usual hypothesis test can lead to incorrect results (cf. Hsu et al., 2016, p. 193). Because the use of several technical trading strategies on the same samples can lead to the significantly best trading strategies occurring randomly. This error is called “data snooping bias” in the literature (cf. White, 2000, p. 1097). The latest studies circumvent this error by using the bootstrapping method. A more detailed explanation can be found in the corresponding chapter.
The autocorrelation is then presented. The occurrence of autocorrelation in the price time series is an essential assumption for the momentum strategy. The empirical studies provide very different results on this topic. When looking at different periods, the autocorrelations in the respective markets differ from one another.
The main aim of the investigations is to answer the question of whether it is possible to systematically achieve excess returns with technical trading. Above all, the excess returns should be higher than the returns of the “buy and hold” strategy of the underlying security or market (underlying). In addition, the risk accepted in technical trading should be less than or equal to the risk of the respective underlying. In order to be able to establish a comparison of the strategies, most studies use the Sharpe ratio as a measure of performance. A detailed explanation is given in chapter 3.5.
Chapter 4 presents the results of the investigations. The studies show significant differences in the various markets as well as times. Therefore, this chapter is divided into stock and currency markets.
2 Technical Analysis
This chapter looks at the trading techniques used in the studies. In TA, a time series analysis of the historical courses is carried out in order to be able to make a prognosis about future course developments. This is mainly done through indicators, oscillators and chart patterns. Carry out technical trading strategy, automated by predefined rules, buy and sell a security in the corresponding market. The goal when using technical trading strategies is to identify a trend early on and to trade it profitably.
Before introducing the technical trading strategies, there is a brief introduction to the efficient market hypothesis, as this thesis serves as the counter-argument to TA. A comprehensive statement on this hypothesis is dispensed with, as it would unnecessarily expand the scope of this bachelor thesis.
2.2 Efficient market hypothesis
The efficient market hypothesis has been a widely used and accepted hypothesis about our financial markets since 1970. The EMH assumes collective investor behavior. Every individual tries to maximize his profits through rational action by regularly updating and evaluating all available information. Further assumptions are that there are no transaction costs and all information available free of charge for each of the market participants. This information includes historical prices, public and inside information. Under these conditions, all information is shown in the price and the current rate represents the most efficient rate (cf. Fama, 1970, p. 387).
Efficient markets are divided into weak, semi-strong and strong forms in the literature.
-In a weakly efficient market, all historical information is already priced into the market.
-In a semi-efficient market, both historical and publicly available information are priced into the market.
-In a highly efficient market, all information known to anyone (e.g. insider information) is already priced into the market (cf. Jensen, 1978, pp. 3-4).
The logical conclusion from this is that any attempt to obtain and evaluate information is unnecessary in order to make price forecasts about the future. The procurement and evaluation of information can result in high costs. Assuming our financial markets are efficient, costs are not offset by returns. The fact that financial institutions spend large amounts of resources on obtaining and evaluating information is all the more contradictory (cf. Aronson, 2007, p. 342 ff.).
Arbitrageurs play an important role in the efficient market. These take advantage of price differences (prices that deviate from the rational level) in order to achieve profits without risk. This can occur above all if some investors incorrectly evaluate information and thereby drive the price above or below the rational price level. The course adapts to the rational level by means of the arbitrageurs. The price movements around the rational price occur randomly (cf. Aronson, 2007, p. 334 ff.).
The random price development is also known as a random walk. A graphic illustration is given in Figure 1. The random walk implies that the course changes are independent of one another. The non-linear dependency can be checked by an autocorrelation test. In principle, various results can be found in the scientific studies on autocorrelations. A deeper insight can be found in the corresponding chapter.
Figure not included in this excerpt
Figure 1: Random walk around the rational price level (source: Aronson, 2007, p. 336).
In the financial markets one can very often observe price developments that deviate from Figure 1. Especially when new information is made public, the phenomena of over- or underreaction should be observed regularly. The developer of the EMH, Eugene F. Fama, argues in his scientific study published in 1998 that course anomalies (e.g. over- and underreaction) appear equally frequently in the courses over the long term. This means that in the event of an overreaction (underreaction) in the courses, there will be an underreaction (overreaction) in the long term, so that it again reaches the rational level. (See Fama, 1998, p. 284).
A study from 2007 sees a contradiction in this line of argument. If there is an overreaction or underreaction and the expectation that the price will return to its mean in the long term, a consideration or analysis of the historical prices leads to prices becoming predictable (cf. Aronson, 2007, p. 339). The EMH thus excludes systematic profit opportunities through technical trading.
Insights into the profitability of technical trading strategies, e.g. momentum strategies in the small-cap market, served as a point of attack for EMH several times. These were often able to provide highly significant results before 1990. Eugen Fama and Kenneth French (1993) then modified the “Capital Asset Pricing Model” (CAPM). This should better describe the expected return through the risk taken. With CAPM, risk-adjusted returns are calculated with a risk factor, namely the beta. This maps the systematic risk. In their studies, Eugen Fama and Kenneth French added two risk factors to the CAPM. It was found that stocks with a small market capitalization and stocks with a high book / market value ratio achieved better results than the overall market (cf. Fama / French, 1993, p. 4).
Market capitalization is determined as a risk factor, since smaller companies have more declines in profits in phases of economic depression and profit less than large companies in phases of economic upswing. The addition of the second risk factor is justified by the fact that companies with a higher book value / market value ratio generate less profit in relation to the assets invested and are therefore not profitable enough (Fama / French, 1993, p. 7 ff.).
Securities with high book / market value ratios are generally referred to as “value stocks”. David R. Aronson mentions in his scientific work that to date no empirical study has been able to provide evidence that “value stocks” -based investments would perform worse during an economic recession (cf. Aronson, 2007, p. 355).
The studies presented in this bachelor thesis often take a position on EMH. Some confirm the EMH and others criticize it. Most of the results from the research indicate that our financial markets have become more efficient over time. More on this can be found in Chapter 4.
2.3 Technical trading strategies
Technical trading strategies are strategies that generate trading signals under predefined conditions. In the empirical studies, entry and exit strategies are assigned to technical trading strategies. These run automatically over a historical time series. The trading signals obtained from this provide data on profitability for the respective technical trading systems.
The counterpart to TA is fundamental analysis. The fundamental analysts evaluate the number of companies in order to make sales or purchase decisions. In principle, no trading signals are generated via the fundamental analysis in the investigations. Some key figures, such as the ratio of book / market value of equity or profit / price, are converted into indicators in momentum strategies in order to generate tradable signals. Therefore, these are also assigned to the technical trade.
The most common methods of generating technical trading signals are presented in the following chapters. Indicators, oscillators and chart patterns use historical price and time information to generate values that can be perceived and traded as trading signals. Most technical trading strategies are mainly those that follow the current course and therefore usually turn later in the direction of the course. This is mainly due to the fact that in technical trading, historical prices are generally used to make trading decisions.
One way to project price fluctuations into the future is to use a trading system that uses the GARCH model. In this bachelor thesis we will mainly only present trailing trading systems, as simple trading systems were used in most of the empirical studies.
Using simple trading systems has both advantages and disadvantages. The advantage is that these are known to many market participants. If these already deliver significant results in terms of profitability, it can be concluded that technical trading is generally profitable. Because the familiarity of a technical trading system can turn the profitability of a trading system for the worse. Profitable technical trading systems tend to become unprofitable trading systems as they become more well known. This is also called the disintegration of a trading strategy (cf. Timmermann / Granger, 2004, p. 22).
The following example should clarify this:
An investor owns a profitable trading system that systematically generates profits. This system e.g. generates buy signals in falling markets. The investor buys securities at certain prices based on the buy signals.
Many investors now have the same trading system. As a result, they also receive the buy signal at the same price level. The high demand cannot be met at this price level, as the majority of market participants want to buy and only a few want to sell. There is an excess of demand and unfulfilled purchase orders.
The price of the security continues to rise. The investors, for whom only a partial execution or no execution at all took place, will place the purchase order above the generated purchase price at the next buy signal in order to be able to receive the purchase execution this time. Therefore, with a new buy signal and a successful buy execution, a price increase occurs before the original price level generated by the technical trading system is reached. At the same time, the rest of the investors who use this trading system miss the opportunity to enter, as the price has already turned upwards in advance.
Now they also decide to place the buy order at an even higher price with the next buy signal in order to be able to trade the trading signal. As a result, the entry price rises far above the price level generated by the trading system. The profitable trading system falls apart and becomes unprofitable.
This example is valid for market participants who buy in falling markets. Nevertheless, the logic remains the same for market participants who use other strategies.
From this knowledge, the question now arises whether studies of the profitability of technical trading systems by using simple technical trading strategies can make a statement about the total population.
2.3.2 Chart patterns
When trading chart patterns, price formations are visualized and identified. The expectation in chart pattern trading is that market participants will behave in the same way in future market situations as they have already done in the past. Based on this expectation, buying and selling decisions are made in order to generate profits.The chart patterns are basically assigned the observed formation as a name. The best-known are shoulder-head-shoulder, triple tips and bottoms as well as flag and wedge formations. In order to get a better understanding of such trading systems, two of the formations mentioned above are presented below.
Below is the shoulder-head-shoulder formation:
Figure not included in this excerpt
Figure 2: Trend reversal pattern - shoulders-head-shoulders.
The formation in Figure 2 forms the "shoulder-head-shoulder" formation (SCS). The price development from the point “left shoulder” to the point “head” forms a higher high. From this point the movement goes to the neckline, which in this case is a support line. The point “right shoulder” draws a weak ascent, which lies below the point “left shoulder”. The movement then returns down to the neckline. This is broken down below. This is followed by a return to the neckline, which in this case presents itself as resistance. The movement then forms a downtrend. This chart pattern is generally viewed as a reversal formation (cf. Murphy, 2013, p. 115).
The pattern below in Figure 3 is a trend continuation pattern and is known as the "flag". You can find this formation quite often in the course of the course. One assumes this formation that a downward trend is intact, which then consolidates upwards in the form of a flag. Breaking the lower trend line (red lines) signals the end of consolidation and the continuation of the trend.
The signal is, however, only valid if the last high in the “flag” does not exceed the higher-level high of the downward movement (cf. Murphy, 2013, p. 151 f.).
Figure not included in this excerpt
Figure 3: Trend continuation formation flag.
Chart formations are often and easily recognizable in the course of the course. Trading such chart formations can quickly lead to high transaction costs. It is usually advisable to incorporate filters into trading strategies in order to reduce false signals. Several such chart pattern trading systems are represented in some studies.
- Do you think India is really sovereign
- What is fictional prose
- How does letrozole work for infertility
- How was Alexander killed
- Which river divides India into two parts
- How can I calculate my car loan
- Playing an instrument relieves stress
- Will Apple ever replace the lightning cable?
- Are you lonely why
- Social media is ruining society
- Nothing can be a logical impossibility
- Which web apps use Knockout js
- What is Profit Sharing
- What is the RAM of OnePlus 6
- Are associated with free will and entropy
- Have people been busy for the past few years
- What is the postcode for Paris
- Why shouldn't we get rid of drug patents?
- Are you frustrated with SIRI on iPhone?
- What are the tastiest flavor combinations
- Why did CSC adopt Xchanging
- Who would win Neferpitou against Isaac Netero?
- How open-minded are scientists today
- How can I make an agglutinative conlang