A fuzzy logic model to forecast stock market

momentum in Indonesia’s property and real

estate sector
Cite as: AIP Conference Proceedings 1862, 030125 (2017); https://doi.org/10.1063/1.4991229
Published Online: 10 July 2017

H. K. Penawar, and Z. Rustam

ARTICLES YOU MAY BE INTERESTED IN

Application of machine learning on brain cancer multiclass classification

AIP Conference Proceedings 1862, 030133 (2017); https://doi.org/10.1063/1.4991237

Normed kernel function-based fuzzy possibilistic C-means (NKFPCM) algorithm for high-
dimensional breast cancer database classification with feature selection is based on Laplacian
Score
AIP Conference Proceedings 1862, 030143 (2017); https://doi.org/10.1063/1.4991247

Fuzzy Kernel k-Medoids algorithm for anomaly detection problems

AIP Conference Proceedings 1862, 030154 (2017); https://doi.org/10.1063/1.4991258

AIP Conference Proceedings 1862, 030125 (2017); https://doi.org/10.1063/1.4991229 1862, 030125

© 2017 Author(s).
 A Fuzzy Logic Model to Forecast Stock Market Momentum
in Indonesia’s Property and Real Estate Sector
H. K Penawar and Z. Rustama)

Department of Mathematics, Faculty of Mathematics and Natural Sciences (FMIPA),

Universitas Indonesia, Depok 16424, Indonesia
a)
Corresponding author: rustam@ui.ac.id

Abstract. The Capital market has the important role in Indonesia’s economy. The capital market does not only support
the economy of Indonesia but also being an indicator Indonesia’s economy improvement. Something that has been traded
in the capital market is stock (stock market). Nowadays, the stock market is full of uncertainty. That uncertainty values
make predicting stock market is all that we have to do before we make a decision in the stock market. One that can be
predicted in the stock market is momentum. To forecast stock market momentum, it can use fuzzy logic model. In the
process of modeling, it will be used 14 days historical data that consisting the value of open, high, low, and close, to
predict the next 5 days momentum categories. There are three momentum categories namely Bullish, Neutral, and
Bearish. To illustrate the fuzzy logic model, we will use stocks data from several companies that listed on Indonesia
Stock Exchange (IDX) in property and real estate sector.

INTRODUCTION
Indonesia as a developing country is in need of substantial capital which is comparable to growth rates that have
been targeted. In this case, the capital markets have an important role in the Indonesian economy. Because the stock
market is not only supporting the economy of Indonesia but also an indicator of economic progress in Indonesia. In
capital markets, one that has been traded is stock. Stock can be defined as a sign of ownership of a person or
company (entity) within a corporation or limited liability company.
In Indonesia, many people are investing in property and real estate because land prices tend to rise each year.
This is because the supply of land is fixed and the demand is increasing from public. In the property and real estate
sector, project funding is a major problem that must be solved. So the company in property and real estate sector
should seek the stock investor for the entire project to run smoothly.
In conducting transactions in capital markets, especially in the areas of property and real estate, stock investors
should gather information as much as possible. One of that information is the stock market momentum. Stock
market momentum can be described as the movement of the stock price. [1] The stock price is one indicator of the
success of a company. If a company's stock price always goes up, then potential investors can assess that the
company is successful. So that potential investors want to invest funds in the company. The stock market
momentum is divided into three categories, namely bearish, neutral, and bullish. Bearish stock can be interpreted
sellers dominate the market and push prices lower. Neutral means buyers and sellers of stock control markets and
prices are still within the limits. Bullish can be interpreted stock buyers dominate the market and push prices higher.
[2]
To find out information about stock market momentum in property and real estate sector in the future can be
used fuzzy set. Working system of fuzzy sets is detected uncertainty and lack of clarity in everyday’s life. Each
member of this fuzzy set will have a membership value, so it can use the fuzzy set with the help of fuzzy logic to
obtain conclusions. Fuzzy logic is defined as a system of logic that play a role in approximate reasoning.
Approximate Reasoning is a process that aims to get a possible conclusion. [3] Fuzzy logic has also been applied in
various fields including healthcare, transportation, and many other fields.

International Symposium on Current Progress in Mathematics and Sciences 2016 (ISCPMS 2016)
AIP Conf. Proc. 1862, 030125-1–030125-5; doi: 10.1063/1.4991229
Published by AIP Publishing. 978-0-7354-1536-2/$30.00

030125-1
 In this research, the use of fuzzy logic in predicting stock market momentum due to fuzzy logic can handle
incomplete data and incorrect. Fuzzy logic also uses the concept of a linguistic variable so it can easily be
understood. [4] This study used the data from several companies in property and real estate sector that listed in
Indonesia Stock Exchange. Data that needed in this research is 14 historic data of the companies’ high, low, open,
and close stock value in 2015. By using fuzzy logic output is expected that the result will be in stock market
momentum categories, which are bearish, bullish, and neutral for the next 5 days.

PROPOSED METHODOLOGY
In this paper, mathematical methods that will be used is part of fuzzy logic, which is membership function. Here
are the steps of fuzzy logic model to forecast stock market momentum.

Initializing Variables
In processing fuzzy logic model, there are variables that will be used. Table 1 describes all of the variables that
will be used in this fuzzy logic model.

Determination of Momentum Values

difference of open and close values in -th day is ∆ = − where = 1,2, … ,14. Momentum value is the value
With 14 historic data of open and close values, it can be calculated the difference between those values.The

of direction power of stock price movement. Momentum value of -th day is defined as = |∆ |.

Membership Function
Momentum values divided into several categories. There are three categories that will be discussed in general,
which are bearish, neutral, and bullish. In search of the membership function will be added in the form of hedge.
Hedge serves to modify the values of fuzzy. Hedge knew as an adjective which is combined with a noun. [1] So
there will be five categories momentum interval, including the Very Bearish, Bearish, Neutral, Bullish, and Very

as percentage values of maximum value of the momentum , which are = 15%, = 30%, = 45%, = 60%,
Bullish. The limits that will be used to limit the interval there is a, b, c, d, and e. The values of a to e are calculated

and = 75%. [2]

TABLE 1. Variables of Fuzzy Logic Model

No. Variable Definition

1 Open value of -th day
2 Close value of -th day
3 Lowest value of -th day

∆
4 Highest value of -th day
5 The difference of and in -th day

!"#$
6 Momentum value of -th day

!#$
7 Very bearish fuzzy momentum value of th day

!%
8 Bearish fuzzy momentum value of th day

!#&
9 Neutral fuzzy momentum value of th day

!"#&
10 Bullish fuzzy momentum value of th day

'
11 Very bullish fuzzy momentum value of th day

(
12 Fuzzy momentum value of -th day

)
13 Fuzzy momentum value that will be predicted in next 5 days
14 Fuzzy linguistic interpretation

030125-2
 FIGURE 1. Triangular and Trapezoidal Membership Function

Membership Function of fuzzy sets can be determined its shape according to what it wants as long as it meets the
rules of the conditions of membership function should be [4]. To get the momentum membership value, it will be
used triangular and trapezoidal shapes. This Fig. 1 is a picture of the triangular and trapezoidal membership function
that will be used.

0, ≤
Triangular and trapezoidal membership function is defined according to this Equation (1.1) :

1, ≤ . −
. ,
, − , ≤ ≤
!"#$ = !#$ = −
, ≤ ≤ −
- − - , ≤ ≤
, 0, ≥ , −
+
+ 1, ≥
0, ≤ 0, ≤
. . −
, −
, , ≤ ≤ , , ≤ ≤
!% = − !#& = −
−
- − , ≤ ≤ - , ≤ ≤
, −
, , −
+ 1, ≥ + 1, ≥

. 1, ≥
, −
!"#& = , ≤ ≤
(1.1)
- −
, 0, ≤
+

Fuzzy Momentum Values

To obtain fuzzy momentum values in each day can be defined as Equation (1.2) follows:

max5 !"#$ , !#$ , !% , !#& , !"#& 6, ∆ ≥ 0

' =1
Secondmax5 !"#$ , !#$ , !% , !#& , !"#& 6, ∆< 0
(1.2)

030125-3
 Estimated Value of Fuzzy Momentum
Estimated value of fuzzy momentum φ will be evaluated on 14 days of historical data. Estimated value of fuzzy

Open (>? ! ? ! Low (>? ! ? !

@A @A
TABLE 2. Stocks Data of Agung Podomoro Tbk 1 January 2015-20 January 2015
Date High (>B? Close (>CD
Jan 1, 2015 335,00 335,00 335,00 335,00
Jan 2, 2015 336,00 339,00 334,00 337,00
Jan 5, 2015 340,00 347,00 335,00 346,00
Jan 6, 2015 343,00 362,00 336,00 359,00
Jan 7, 2015 365,00 445,00 361,00 419,00
Jan 8, 2015 425,00 433,00 407,00 430,00
Jan 9, 2015 430,00 445,00 424,00 439,00
Jan 12, 2015 440,00 440,00 420,00 421,00
Jan 13, 2015 420,00 430,00 418,00 424,00
Jan 14, 2015 425,00 428,00 410,00 412,00
Jan 15, 2015 423,00 427,00 412,00 423,00
Jan 16, 2015 419,00 429,00 417,00 429,00
Jan 19, 2015 433,00 437,00 420,00 420,00
Jan 20, 2015 421,00 425,00 408,00 414,00

TABLE 3. Stocks Data of Agung Podomoro Tbk 21 January 2015-27 January 2015

Open (>? ! ? ! Low (>? ! ? !

@A @A
Date High (>B? Close (>CD
Jan 21, 2015 414,00 420,00 410,00 411,00
Jan 22, 2015 414,00 423,00 411,00 417,00
Jan 23, 2015 420,00 437,00 420,00 421,00
Jan 26, 2015 421,00 424,00 410,00 413,00
Jan 27, 2015 418,00 425,00 416,00 418,00

momentum is momentum that will predict market movements in the next 5 days. Fuzzy momentum value that is
estimated or can be expressed as φ be defined as Equation (1.3) follows :

(=
E∑IJ
HKI GH L
MN
. (1.3)

Linguistic Interpretation of Fuzzy Momentum Value

The important concept of fuzzy logic is linguistic variables. After getting the fuzzy values or degrees of
membership, it will be the interpretation of these values. The equation (1.4) are linguistic interpretations of the
estimated value of fuzzy momentum.

"Bearish", 0 ≤ ( ≤ 0,5
) = O"Neutral", 0,5 < ( ≤ 0,6
(1.4)

"Bullish", 0,6 < ( ≤ 1

IMPLEMENTATION
The company that will be used as an example is Agung Podomoro Tbk. Stock data that will be used is 14 days in
January 2015 to predict the next 5 days momentum. Table 2 is Agung Podomoro Tbk shares in January 2015. [5]
After the modeling process with the data table above, followed by linguistic interpretation. The benefits of this
interpretation is to determine categories of the fuzzy momentum values, are Bearish, Neutral, and Bullish. Because

030125-4
the values that obtained from Agung Podomoro Tbk in January 2015 is 0.84, then the lingustic intepretation is
Bullish.
To check whether it’s right for the next 5 days will be categorized into Bullish, it can be seen from the stock
values data when the open and close values for 21, 22, 23, 26, and January 27, 2015. Checking from the table below
that the close value on January 27, 2015 was 418.00 and open value on January 21, 2015 was 414.00. It can be seen
in the difference of the those values is 4.00. From this difference, we can draw conclusions that Agung Podomoro
Tbk on 27 January 2015 categorized as Bullish momentum. This is a table of data Agung Podomoro Tbk shares for
next 5 days after 14 days are used as historical data.

It can be considered by using fuzzy logic model, on December 23 and January 22, it can also be categorized
Bullish momentum. On January 23, close value was 421.00 and open value on January 21 was 414.00, resulting in a
difference of 7.00. On January 22, close value was 417.00 and open value on January 21 was 414.00, so the
difference is worth 3.00.

CONCLUSIONS
The conclusion that can be derived from this research is fuzzy logic model can be applied to predict the next 5-
day momentum of the stock market by using historical data for 14 days. In addition, the fuzzy logic modeling can
not only be applied to predict the stock market's momentum in the property and real estate sector but can also be
applied in other sectors listed in Indonesia Stock Exchange.

ACKNOWLEDGMENTS
The author are grateful to the referees for their valueable comments and suggestion. The authors also
acknowledgment that this research using data from Indonesia Stock Exchange.

REFERENCES
1. R. J. Teweles and E. S. Bradley, The Stock Market (John Wiley & Sons, Hoboken, New Jersey, 1998), pp. 3-1.
2. P. Roy, R. Kumar, and S. Sharma, Int. J. Adv. Eng. Res. Studies 98, 101 (2015).
3. P. Engelbrecht, Computational Intelligence: An Introduction (John Wiley & Sons, Hoboken, New Jersey,
2007), pp. 441-472.
4. J. S. R. Jang, C. T. Sun, and E. Mizutani, Neuro-Fuzzy and Soft Computing: A Computational Approach to
Learning and Machine Intelligence (Prentice-Hall, New Jersey, 1997), pp. 73-89.
5. Indonesia Stock Exchange (Bursa Efek Indonesia, Jakarta, 2015).

030125-5