See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/325451690

An Analyze on Effectiveness of Candlestick Reversal Patterns for Vietnamese

Stock Market

Preprint · May 2018

CITATIONS READS

0 52

5 authors, including:

Bui Ngoc Anh Le Van Thanh

FPT University FPT University
10 PUBLICATIONS   4 CITATIONS    6 PUBLICATIONS   3 CITATIONS   

SEE PROFILE SEE PROFILE

Ban Tran Quy Hoa Duong Xuan

FPT University FPT University
3 PUBLICATIONS   0 CITATIONS    3 PUBLICATIONS   2 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

COMBINATION OF DIMENSIONALITY REDUCTION AND USER CLUSTERING FOR COLLABORATIVE-FILTERING View project

Auto - Car View project

All content following this page was uploaded by Ngo Tung Son on 17 July 2018.

The user has requested enhancement of the downloaded file.

 An Analyze on Effectiveness of Candlestick Reversal
Patterns for Vietnamese Stock Market
Ngo Tung Son Le Van Thanh Tran Quy Ban
ICT Department, FPT University, ICT Department, FPT University, ICT Department, FPT University,
Hanoi, Vietnam Hanoi, Vietnam Hanoi, Vietnam
[email protected] [email protected] [email protected]

Duong Xuan Hoa Bui Ngoc Anh

ICT Department, FPT University, ICT Department, FPT University,
Hanoi, Vietnam Hanoi, Vietnam
[email protected] [email protected]

ABSTRACT Similarly, the HNX-Index rose 13.35% to 132.46, UPCoM up

The Japanese candlestick model is a common technical analysis 10.5% [1]. The goal of investors is always to gain maximum profit
used to understand the behavior and predict the trend in the on their investments. There are two approaches they commonly use
financial market. There are many studies have been conducted to [2]: the fundamental analysis (FA) approach which focuses on all
assess the effectiveness of candlestick patterns in different markets. publicly available information and the technical analysis (TA)
This paper describes an empirical research to examine the approach which based on an investigation of historical information.
predictability and profitability of the candlestick reversal patterns Elton et al pointed out that the effectiveness of each approach type
analysis on the Vietnamese stock market over the period from Jan depends on the type of market [3]. Candlestick is one of the very
2, 2013, to May 15, 2018. The purpose is to determine whether famous models in TA. It uses the open, high, low and closing
candlesticks patterns can be used to achieve significant gains or to (denoted by OHLC) values for each time period you want to display
predict future price trends. We apply the use of Pivot Simple [4]. The hollow or colored portion of the candlestick is called the
Moving Average to indicate the reversal signals of the market and body. The shadow represented by long thin lines above and below
compute the profit. Our results show the tested reversal the body. It describes the high/low range of the prices. Figure 1
candlesticks do not demonstrate the ability to predict the market represents high, low, open, and close prices of each displayed time
trend and generated profitability is low on the stock market in period. The candlestick patterns can be classified into two classes:
Vietnam. The main contribution of this paper is a statistical continuation and reversal patterns. Continuation patterns indicate
analysis of the effectiveness of candlestick patterns as predictors of that the current trend will continue, while reversal patterns indicate
the behavior of those stocks. that the current trend will reverse [5]. Reversal patterns are bullish
or bearish, the bearish patterns are inverses of the bullish patterns
CCS Concepts and vice versa.
• Information systems → Data analytics
Upper shadow  High Upper shadow  High
Keywords
Vietnamese Stock Market, Technical Analysis, Candlestick, Trend
Prediction, Profitability. Open Close

1. INTRODUCTION
Body  Body 
In developed countries such as the United States, Europe and Japan,
investment activity has grown exponentially and consisted of large
financial institutions and individual investors with abundant funds.
In recent years in Vietnam, the demand for personal investment has Close Open
increased sharply as financial markets have begun to grow and
develop. The stock market, derivatives ... gradually becomes an
Lower shadow  Low Lower shadow  Low
attractive investment. According to Vietnam State Securities
Commission, the stock market capitalization in Vietnam reached
Figure 1 Formation of Candlestick
over $ 183 billion, accounting for 82% of GDP. The VN Index
ended the first quarter of 2017 reached 1,174.46, up 19.33% 2. RELATED WORKS
compared to the end of 2017 and led the world in terms of growth. In the past, many researchers were conducted to evaluate the
predictive power of candlestick patterns in different stock
SAMPLE: Permission to make digital or hard copies of all or part of this
work for personal or classroom use is granted without fee provided that markets. There are both positive and negative conclusions about
copies are not made or distributed for profit or commercial advantage and the candlestick model. Several researchers believe that
that copies bear this notice and the full citation on the first page. To copy candlestick patterns do not bring value to investors in the stock
otherwise, or republish, to post on servers or to redistribute to lists, market: Marshall et al (2006) [6] examined the profitability of the
requires prior specific permission and/or a fee. candlestick patterns. In their study, they used price data from 35
Conference’10, Month 1–2, 2010, City, State, Country. stocks in the Dow Jones Industrial Average (DJIA) for the period
Copyright 2010 ACM 1-58113-000-0/00/0010 …$15.00. from 1992 to 2001. They found that candlestick patterns had no
DOI: http://dx.doi.org/10.1145/12345.67890 financial or financial value for the investors. Goo et al tested the
performance of candlestick patterns on the Taiwan stock market particular market and the combination with other techniques. In
from 1997 to 2006. They found that the best performing candle many cases, the holding strategies also play very important roles in
pattern in three to four days hold while the bullish pattern works bringing profit to the investors.
best during a 9-10 day period [7]. Marshall et al (2008) [8] again
examined the predictive power and profitability of candlesticks 3. EXPERIMENTAL DESIGN
for a 10-day holding period. The tests were conducted using the This study employs a daily sample from 9 components stocks of
largest 100 stocks on the Tokyo Stock Exchange between 1975 HOSE index for the period from January 2, 2013 to March 15,
and 2002 by the same technique they have done in 2006. Finally, 2018 (downloaded from cophieu68.vn [15]): CTG (Vietinbank),
they concluded that the candlesticks had no predictive power as GAS (Petro Vietnam), HPG (Hoa Phat Group), MSN (Masan
well as any profit on the Japanese stock market. Horton studied Group Corporation), VCB (Bank for Foreign Trade of Vietnam),
the profitability of eight candlestick patterns on 349 stocks, VIC (Vingroup), VNM (VietNam Dairy Products), FPT (FPT
comparing their gains with buy and hold strategies. He concluded Corporation), and HAP (Hapaco Corporation). The 10 tested
that the Stars, Crows, and Dojis candlestick models were not candlesticks patterns are defined in Figure 2. The uptrend and
profitable [9]. Max Jönsson tested the predictive power of the downtrend are defined in the section 3.1 and profit of each
candlestick pattern on the Swedish stock market between 2007 candlestick are calculated based on equation in the section 3.2. We
and 2015 for his bachelor thesis [10]. He pointed out that first evaluate the predictive power of all bullish and bearish
candlestick analysis was not of value for profit to investors on the patterns, and then calculate the return of each candlestick on
Swedish stock market. different holding period.
The above studies supported Morris's view when he said: "market 3.1 Trend Definition
psychology cannot be measured by statistics" [5]. On the other
Many researches have used Moving Average (MA) to identify the
hand, many studies have shown that candlesticks yield certain
trend of finance market. Lu et al used the five-day simple moving
values: Hércules et al evaluated 16 patterns on 10 component
average (SMA) to define trend of tested candlestick patterns [12,
stocks of Bovespa. The data collected over 5 years from 2005 to
13]. In this study, we apply five-day pivot-based MA described by
2009 in Brazil, 9 patterns can be considered relatively regular
Occhoa [16]. The pivot point is derived by taking the average of the
because they occur at frequencies higher than 1% [11]. Lu et al
high, low, and close prices of a period of candle: 𝑝𝑡 = (H𝑡 + L𝑡 +
(2015) use 26 (after pre-processing) stocks of the DJIA index for
C𝑡 ) / 3
the period from January 2, 1992 to December 31, 2012. To
explore the predictive and lucrative potential of eight three-day Where 𝑝𝑡 is pivot point value, H𝑡 L𝑡 and C𝑡 are respectively High
reversal patterns. They show that the type of holding plays a price, Low price and Closing price of day 𝑡 𝑡ℎ . The five-day moving
major role in the effectiveness of candlestick trading strategies. is given by:
4
Especially candlestick trading strategies can create better value 1
for investors on riskier assets [12]. Lu et al (2016) analyze the 1- 𝑀𝐴𝑡 = ∑ 𝑝𝑡−𝑖
5
day candlestick pattern using data from 30 component stocks of 𝑖=0
the DJIA, with some patterns still proving profitable especially Declare 𝑚𝑡,𝑡−𝑖 is a binary variable to mark if 2 consecutive days is
since 1992 [13]. F. Martinsson and I. Liljeqvist [14] tested both bullish. Where:
the Swedish OMXS30 market and the UK FTSE100 market using 1 if 𝑀𝐴𝑡 > 𝑀𝐴𝑡−𝑖
𝑚𝑡,𝑡−𝑖 = {
10 candlestick patterns. They performed an analysis on the effects 0 𝑜𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒
of Relative Strength Index (RSI) and stop loss configurations on The uptrend is identified if the MA is increasing consecutively for
a trading algorithm based on candlestick patterns. Their result at least five of the past six days. However, we will relax this
shows that different RSI and stop loss configurations has a condition by requiring only 4/5 times of accuracy:
substantial impact on the performance of a trading algorithm 4
based on candlestick patterns. This information shows that the ∑ 𝑚𝑡−𝑖,𝑡−𝑖−1 ≥ 4
effectiveness of the candlestick pattern depends very much on the 𝑖=0

Figure 2: 10 tested Candlestick Patterns. Images from www.OnlinetradingConcepts.com

Similarly, the downtrend is identified as: times detected, bearish candlesticks patterns have 44.4% prediction
4 accuracy.
∑ 𝑚𝑡−𝑖−1,𝑡−𝑖 ≥ 4
𝑖=0
We then perform a 2 tailed t-test to evaluate if a candlestick is a
statistically significant trend reverser with:
Null hypothesis: 𝐻0 : 𝑝𝑐 − 𝑝 = 0 49%
56%
Alternative hypothesis: 𝐻1 : 𝑝𝑐 − 𝑝 ≠ 0
Where 𝑝 is the overall probability of a downtrend / uptrend being
reversed. 𝑝𝑐 denotes the probability of trend reversal success on Fail
each pattern 𝑐 𝑡ℎ .
Correct
3.2 Profit Calculation
The profits are calculated with the following assumptions: buy with 51%
the ‘open’ price after the pattern appears, and sell with the ‘close’ 44%
price of the last day of holding period.
C𝑡+𝑛 −𝑂𝑡+1 C𝑡+𝑛 −𝑂𝑡+1
𝑓𝑢𝑝𝑡𝑟𝑒𝑛𝑑 and 𝑓𝑑𝑜𝑤𝑛𝑡𝑟𝑒𝑛𝑑 = −
𝑂𝑡+1 𝑂𝑡+1
Bullish Bearish
Where 𝑡 denotes the day that tested pattern occur. 𝑛 denotes the
number of days in the holding period. C𝑡+𝑛 represents the closing
price at day 𝑡 + 𝑛. 𝑂𝑡+1 is the opening price at day 𝑡 + 1. Nison Figure 3: Rate of reversal over the entire stock market of
suggested that candlestick model is useful for short-term trading Vietnam during selected period for experiment.
[4]. We assess the profit for holding 1 to 10 days after patterns We then evaluated the reversal power of individual patterns. The
appear. The average rates of return are tested by a 1 tailed t-tested statistical test results are presented in Table 1 and Figure 5
for significance after the calculation of the mean rate of return on visualizes the comparison of the rates of success and failure
each pattern during each of the holding periods. predictions of the patterns. With the 5% significant level, none of
Null hypothesis: 𝐻0 : 𝑟𝑖𝑗 = 0 the candlestick patterns is a statistically significant trend reversal
predictor. The Bearish Engulfing is the most common pattern, with
Alternative hypothesis: 𝐻1 : 𝑟𝑖𝑗 > 0 91 times successful predict trend reversal over 205 times occurs, it-
Where 𝑟𝑖𝑗 is represents the mean rate of return for each holding -has a success rate of 44.39% and p-value of 74%, not a statistical
period 𝑖𝑡ℎ and candlestick pattern 𝑗𝑡ℎ . significant predictor. Piercing is the rarest pattern being only
identified 45 times, however it’s the second most accurate patterns
We don’t take into account transaction cost in this study, because it
with success rate of 53.33% and p-value is 68%. The Bullish
will distract the main focus which is finding a profitable trading
Engulfing, identified 177 times with accuracy of 53.67%, is not
strategy. Moreover, transaction costs are not the same across the
only one of the most common patterns, but also the most accurate
world, so that many addition constraints will have to be properly
patterns with success rate of 53.67%. Therefore it has the lowest p-
considered to adopt for the study.
value of 37%, still well above the critical 5% significance level.
4. RESULTS This part of the section shows the result of profitability examination
Our analysis begins with an examination of the accurate reversal of each candlestick patterns with holding periods from one to ten
signals that 10 candlestick patterns have generated overall 9 stocks, days.
for both bullish and bearish patterns. Figure 3 illustrates the rate of The average returns are calculated for all 9 stocks are tested
reversal over the entire stock market of Vietnam during selected whether they are significantly higher than 0 or not. Table 2 shows
period for the experiment. For the 607 bullish candlesticks detected the statistical test result for profitability on the holding periods. We
on downtrend, there are 310 times the trend followed by are can observe that the Bullish Engulfing, Bullish Harami, Piercing
successfully reversed. Figure 4 shows the frequency of the for some particular holding periods are statistically significant,
particular patterns. The average prediction accuracy of bullish especially during the 4, 5 to 10 day holding period. Meanwhile,
candlestick patterns is 51.1%. The similar result happened for other patterns such as Hammer, Inverted Hammer, and Dark Cloud
bearish patterns: with 311 times accurate predictions over 701 Cover... are not significant difference.

Table 1: The statistical test result for predictive power of experimented Candlestick patterns
Patterns Success Reverse Frequency Prediction Rate t-value p-value
Bullish Engulfing 95 177 53,67% 0,90 0,37
Bullish Harami 90 170 52,94% 0,73 0,46
Hammer 46 97 47,42% -0,22 0,82
Inverted Hammer 55 118 46,61% -0,37 0,72
Piercing 24 45 53,33% 0,42 0,68
Bearish Engulfing 91 205 44,39% -0,34 0,74
Bearish Harami 82 194 42,27% -0,71 0,48
Dark Cloud Cover 24 56 42,86% -0,32 0,75
Hanging Man 47 98 47,96% 0,25 0,80
Shooting Star 67 148 45,27% -0,14 0,89
 Table 2: The statistical test result for profitability on the holding periods
Holding dayth
Patterns Statistic Factors 1 2 3 4 5 6 7 8 9 10
Bullish Engulfing Mean 0.001 0.002 0.004 0.006 0.008 0.010 0.008 0.009 0.009 0.010
T-score 0.517 0.801 1.473 1.774 2.504 2.693 2.161 2.214 2.040 2.167
p-value 0.303 0.212 0.071 0.039 0.007 0.004 0.016 0.014 0.021 0.016
Bullish Harami Mean 0.002 0.005 0.005 0.003 0.002 0.007 0.008 0.009 0.012 0.012
T-score 0.766 1.726 1.552 0.732 0.523 1.663 1.783 1.993 2.630 2.551
p-value 0.222 0.043 0.061 0.232 0.301 0.049 0.038 0.024 0.005 0.006
Hammer Mean -0.002 0.001 0.001 0.003 0.004 0.002 0.003 0.002 0.002 0.004
T-score -1.040 0.201 0.184 0.667 1.129 0.526 0.696 0.314 0.456 0.666
p-value 0.849 0.421 0.427 0.253 0.131 0.300 0.244 0.377 0.325 0.253
Inverted Hammer Mean 0.000 0.001 0.002 0.004 0.005 0.004 0.003 0.001 0.001 0.002
T-score -0.011 0.348 0.525 0.983 0.958 0.867 0.612 0.210 0.123 0.336
p-value 0.505 0.364 0.300 0.164 0.170 0.194 0.271 0.417 0.451 0.369
Piercing Mean 0.003 0.004 0.008 0.010 0.016 0.020 0.020 0.021 0.019 0.019
T-score 1.014 1.010 1.389 1.775 2.528 2.835 2.917 2.752 2.523 2.134
p-value 0.158 0.159 0.086 0.041 0.008 0.003 0.003 0.004 0.008 0.019
Bearish Engulfing Mean -0.001 -0.003 -0.003 -0.006 -0.007 -0.007 -0.009 -0.013 -0.014 -0.012
T-score -0.412 -1.373 -1.055 -2.134 -1.978 -2.099 -2.321 -3.127 -3.276 -2.594
p-value 0.660 0.914 0.854 0.983 0.975 0.981 0.989 0.999 0.999 0.995
Bearish Harami Mean -0.001 -0.003 -0.006 -0.010 -0.011 -0.015 -0.017 -0.020 -0.021 -0.024
T-score -0.396 -1.355 -2.006 -2.759 -2.869 -3.566 -3.941 -4.534 -4.388 -4.773
p-value 0.654 0.912 0.977 0.997 0.998 1.000 1.000 1.000 1.000 1.000
Dark Cloud Cover Mean -0.002 -0.006 -0.004 -0.006 -0.005 -0.002 -0.007 -0.012 -0.008 -0.005
T-score -0.794 -1.633 -0.874 -1.326 -0.842 -0.363 -0.881 -1.463 -1.035 -0.553
p-value 0.785 0.946 0.807 0.905 0.798 0.641 0.809 0.925 0.847 0.709
Hanging Man Mean -0.003 -0.003 -0.005 -0.009 -0.013 -0.013 -0.013 -0.016 -0.018 -0.020
T-score -1.095 -1.001 -1.138 -1.797 -2.543 -2.624 -2.367 -2.681 -2.745 -2.965
p-value 0.862 0.840 0.871 0.962 0.994 0.995 0.990 0.996 0.996 0.998
Shooting Star Mean -0.003 -0.006 -0.007 -0.008 -0.013 -0.013 -0.014 -0.012 -0.015 -0.016
T-score -1.280 -2.363 -2.734 -2.733 -3.618 -2.986 -2.989 -2.755 -3.130 -3.141
p-value 0.899 0.990 0.996 0.996 1.000 0.998 0.998 0.997 0.999 0.999

Taking pivot-SMA(5) trend into account, 3/5 bullish patterns prove

that they are profitable for the periods of 6 to 10 days. The Bullish
Shooting Star
Engulfing and Piercing models are even continuity lucrative in their
Hanging Man
4 to 10 day holding periods. Especially, Piercing is the most
Dark Cloud Cover
profitable when giving highest pattern ~ 2% after 6 holding days.
Bearish Harami
Bearish Engulfing This is quite similar to Max [10] when Piercing is also proved to be
Piercing most profitable. The results also illustrate that the holding strategies
Inverted Hammer based on the Bearish Patterns such as Shooting Star, Hanging Man,
Hammer Bearish Harami and Bearish Engulfing do not bring any profit.
Bullish Harami
Bullish Engulfing
5. CONCLUSION
This study is one of the earliest assessments of the candlestick
0 50 100 150 200 250 model performance on the Vietnamese stock market. The previous
Frequency studies on other stock markets have shown that the effectiveness of
a candlesticks models depends very much on the particular market.
The purpose of this research is to examine whether candlestick
Figure 4: Frequency of occurrences of the patterns patterns can predict future trend and to determine whether they can
be used to achieve profit in short-term. There are 2 separate statistic
Shooting Star tests performed. The results from the first test determining the null
Hanging Man hypothesis stating that candlestick has no predictive power, cannot
Dark Cloud Cover be rejected. There is none of the 10 candlestick patterns has
fail predictive power when using on Vietnamese stock data. This test
Bearish Harami
Bearish Engulfing success result is similar to previous studies in different markets [6, 7, 8, 9,
10, and 12]. The second test is designed to determine if a trading
Piercing
strategy based on candlestick patterns can produce holding period
Inverted Hammer
returns, which are significantly greater than zero. The results show
Hammer that the bullish patterns basically produce higher returns than the
Bullish Harami bearish ones. This finding agrees with Lu (2015) [12]. The period
Bullish Engulfing of holding from 5-th to 10-th day shows the profitability (although
0.00% 50.00% 100.00% small). The result of the research can contribute to the investor's
decision-making process as well as existing researches on technical
Figure 5: Prediction Rate on individual patterns analysis.
This study is limited to the profit calculation is based on buying and Department of Economics NEKH01, Bachelor Thesis, Spring
selling at the exact opening and closing prices, this is unrealistic in 2016, Lund University.
the real-time transaction of investors. It could affect to holding- [11] Hércules A. do Prado et al : On the effectiveness of candlestick
period returns. We have been done experiments on only 10 candles chart analysis for the Brazilian stock market , 7th International
over many others. Others patterns may have additional value and Conference in Knowledge Based and Intelligent Information
the tested patterns may be valuable to investors when combined and Engineering Systems.
with other methods such as the different method of trend indicator.
[12] Lu, T., Y. Chen, and Y. Hsu. 2015. “Trend Definition or
Our future work may revolve around a number of key areas: Holding Strategy: What Determines the Profitability of
Evaluate other market-oriented approaches to avoid the Candlestick Charting?” Journal of Banking & Finance 61:
shortcomings that pivot-SMA (5) can bring. For example in the past 172–183. doi:10.1016/j.jbankfin.2015.09.009
Lu et al used MA (3) [12, 13] or F. Martinson et al [14] applied RSI
as the indicator of market trends. Another work is to test candlestick [13] Tsung-Hsun Lu & Yung-Ming Shiu (2016): Can 1-day
patterns in conjunction with other technical methods such as Fuzzy candlestick patterns be profitable on the 30 component stocks
based-methods [17, 18]. of the DJIA?, Applied Economics, DOI:
10.1080/00036846.2015.1137553
6. REFERENCES [14] F. Martinssson and I. Liljeqvist, ‘Short-Term Stock Market
[1] http://www.ssc.gov.vn, 14-May, 2018 Prediction Based on Candlestick Pattern Analysis’,
[2] Yuen, Raymond Wai Pong, Fundamental Analysis vs. Dissertation, 2017.
Technical Analysis (May 5, 2012). Available at [15] http://www.cophieu68.vn/export.php (24-May, 2018).
SSRN: https://ssrn.com/abstract=2051485 or http://dx.doi.org
/10.2139/ssrn.2051485 [16] FRANKLIN O. OCHOA, JR: PROFITING WITH PIVOT-
BASED MOVING AVERAGES, PivotBoss.com
[3] Elton, Edwin J. and Gruber, Martin J. and Brown, Stephen J.
and Goetzmann, William N. (2009): Modern Portfolio Theory [17] Takenori Kamo and Cihan Dagli : "Fuzzy Logic-Based
and Investment Analysis, 9th edition, John Wiley & Sons, Inc. Japanese Candlestick Pattern Recognition and Financial
Forecast" ,Intelligent Engineering Systems Through Artificial
[4] Nison Steve (1991): Japanese candlestick charting techniques: Neural Networks, Volume 17, DOI:
A Contemporary Guide to the Ancient Investment Techniques 10.1115/1.802655.paper41
of the Far East (1st edition, New York Institute of Finance).
[18] Qiujun Lan, Dan Zhang ,Longling Xiong : Reversal Pattern
[5] Morris, Gregory L. (2006): Candlestick Charting Explained: Discovery in Financial Time Series Based on Fuzzy
Timeless Techniques for Trading Stocks and Futures (3rd Candlestick Lines, Systems Engineering Procedia Volume 2,
Edition The McGraw-Hill Companies, Inc). 2011, Pages 182-190, DOI: 10.1016/j.sepro.2011.10.021
[6] Marshall, Young and Rose (2006): Candlestick technical
trading strategies: Can they create value for investors?,
Journal of Banking & Finance 30, pp 2303–2323
[7] Goo, Chen and Chang (2007): The Application of Japanese
ndlestick Trading Strategies in Taiwan, Investment
Management and Financial Innovations, Volume 4, Issue 4.
[8] Marshall, Young and Chang (2008): Are candlestick technical
trading strategies profitable in the Japanese equity market?,
Review of Quantitative Finance and Accounting 31, pp 191-
207.
[9] Horton, Marshall J. (2009): Stars, crows, and doji: The use of
candlesticks in stock selection, The Quarterly Review of
Economics and Finance 49, pp 283–294.
[10] Max Jönsson: The Predictive Power of Candlestick Patterns
An Empirical Test of Technical Indicators on the Swedish
Stock Market Using GARCH-M and Bootstrapping,
 Columns on Last Page Should Be Made As Close As
Possible to Equal Length

Authors’ background
Your Name Title* Research Field Personal website

*This form helps us to understand your paper better, the form itself will not be published.

*Title can be chosen from: master student, Phd candidate, assistant professor, lecture, senior lecture, associate
professor, full professor

View publication stats