Final report of methods of economic analysis topic building a prediction system for future price fluctuations of stock codes

Hình 1.2 Markov chains are used for several business applications, including predicting customer brand switching for marketing, predicting how long people will remain in their jobs for h

TRƯỜNG ĐẠI HỌC TÔN ĐỨC THẮNG

KHOA TÀI CHÍNH - NGÂN HÀNG

Final Report of Methods of Economic Analysis TOPIC: Building a prediction system for future price fluctuations of stock codes

B03013 - GROUP 6 Advised by:LÊ TUẤN BÁCH

LÊ THANH HOÀ

Student: NGUYỄN MINH HIẾU - B22H0103

ĐÀM TUẤN PHONG - B22H0013

NGUYỄN TRẦN TRƯỜNG THỊNH -B22H0133

Ho Chi Minh City, May 19, 2024.

Mục lục

PART 1 1

MARKOV 1

COBWEB 1

OPTIMIZATION 1

PART 2 2

1.INTRODUCTION 2

1.1/ History 2

1.2/ Definition 3

1.3/ Summary of some documents applying Markov models in Viet Nam 4

2.MOTIVATION FOR CHOOSING A MARKOV MODEL 5

3.APPLYING THE MARKOV CHAINS METHOD ON THE HCM STOCK EXCHANGE – HOSE 5

3.1/ Collecting Data 5

4 RESULT 8

4.1/ Statistics on profitability of 10 stock codes 8

4.2/ Describe the initial state of 10 tickers 10

4.3/ THE PROBABILITY OF THE FIRST NEXT DAY WHEN APPLYING THE MARKOV MODEL 10

4.4/ Describe the static state of 10 tickers 12

5.CONCLUSION 14

6 REFERENCES 15

PART 1

Markov

Cobweb

Optimization

PART 2

1.INTRODUCTION

1.1/ History

Markov chains are named after their creator, Andrey Andreyevich Markov, a Russian mathematician who founded a new branch of probability theory around stochastic processes in the early 1900s Markov was greatly influenced by his teacher and mentor, Pafnuty Chebyshev, whose work also broke new ground in probability theory

1.2/ Definition

A Markov process is a stochastic process that satisfies the Markov property (sometimes characterized as "memorylessness") In simpler terms, it is a process for which predictions can be made regarding future outcomes based solely on its present state and

—most importantly—such predictions are just as good as the ones that could be made knowing the process's full history In other words, conditional on the present state of the system, its future and past states are independent

Hình 1.1

A Markov chain is a type of Markov process that has either a discrete state space or a discrete index set (often representing time), but the precise definition of a Markov chain varies.For example, it is common to define a Markov chain as a Markov process

in either discrete or continuous time with a countable state space (thus regardless of the nature of time), but it is also common to define a Markov chain as having discrete time

in either countable or continuous state space (thus regardless of the state space)

Hình 1.2

Markov chains are used for several business applications, including predicting customer brand switching for marketing, predicting how long people will remain in their jobs for human resources, predicting time to failure of a machine in manufacturing, and forecasting the future price of a stock in finance

1.3/ Summary of some documents applying Markov models in Viet Nam

- Predicting stock price fluctuations: This study focuses on applying the Markov

model to predict stock price fluctuations of stock codes on the Vietnamese stock market It aims to provide insight into general market trends and assist investors in making investment decisions

- Predicting potential investment risks: Markov models can also be used to predict risks that may occur with investment projects, from which investors can consider and choose accurately

- Weather data analysis : Markov models can also be applied to predict fluctuations

in data time, such as rain, temperature and humidity in domestic regions This could be useful for industries such as agriculture, energy and tourism

-Projected traffic information: Use Markov models to analyze vehicle traffic data as

well as information at route locations in the city This work helps predict traffic jams and provide solutions to manage traffic outcomes

2.MOTIVATION FOR CHOOSING A MARKOV MODEL

In the context of the stock market growing and constantly fluctuating, one of the main problems of the stock market today is "Risk prediction and market domination", in other words, finding a method to predict price fluctuations and risks from which to receive benefits safest expected profit margin The appropriate method chosen for use here is the Markov model In this report, we focus on predicting the probability of price fluctuations of 10 stock codes in the securities industry, listed on the Ho Chi Minh City Stock Exchange Ho Chi Minh (HOSE)

In particular, we apply the Markov model in predicting random processes, to build a prediction system for future price fluctuations of stock codes The Markov model has proven to be a popular method for analyzing time series data, and we believe that applying this model will yield the necessary results In addition, practicing based on industry stock codes is the most important thing to complete to serve as a foundation for further study and become an essential skill for students in the future

3.APPLYING THE MARKOV CHAINS METHOD ON THE HCM STOCK EXCHANGE – HOSE

3.1/ Collecting Data

The data in this research topic is referred from the quite reputable website that regularly updates information on companies or banks in the financial sector –

investing.com The data that we collected for the stock codes is in the period from March 24, 2023 to March 25, 2024

The following table will show the names of stock codes representing the companies and banks that we have chosen to conduct this research project

Table 1 Ten companies were selected and their share prices were observed in the period from March 24, 2023, to March 25, 2024 (HOSE)

Based on the knowledge and foundation learned in class and with guidance from teachers, we have selected 10 stock codes from different companies and banks including commercial banks, securities companies, and even jewelry companies After being collected, data will be put directly into a separate Excel file to help viewers easily see and understand the data

The reason we chose these 10 stock codes is for 3 main reasons First, we selected different banks as well as companies to diversify to compare how stock prices changed over the past 1 year (specifically, about 250 observations)

N

o Company code Full name of the company

1 VCB Joint Stock Commercial Bank for Foreign Trade of Vietnam

Joint Stock Commercial Bank for Investment and Development of Vietnam

4 MBB Military Commercial Joint Stock Bank

5 PNJ Phu Nhuan Jewelry Joint Stock Company

6 TPB Tien Phong Commercial Joint Stock Bank

7 VPB Vietnam Prosperity Joint Stock Commercial Bank

8 SSI SSI Securities Corporation

The Corporation for Financing Promoting Technology Joint Stock Company

10 OCB Orient Joint Stock Commercial Bank

Second, finance in general and the banking industry are currently the solid pillar for country’s economy, so comparing many such companies and banks to provide accurate data to help investors invest in companies intelligently

Last but not least, we can see from the collected data that the financial industry is constantly developing in many different directions, contributing to promoting the development of the country's economy

Table 2 Ten companies stock code and distribution of returns in the period from March 24, 2023, to March 25, 2024 (HOSE)

Company

stock

code

Distribution of returns and initial vector values

Times Initial states

After gathering data, we will rely on the collected information to analyze the securities

It will ultimately determine which stocks to invest in and ensure that the circumstances under which investors should not invest avoid the worst possible consequences

The following is a small table explaining the states that occur when we make predictions according to the Markov model

State 1 - S1

State 2 - S2

Assumption

the returns lower than -0.5%

the returns in the interval [-0.5%,0.5%] obtained when the share prices at a given point in time

The probability distribution vector in the first next period and the steady state vector for each company

Choose stock The first next day The steady state

30.94% 34.95% 34.11% 30.93% 34.96% 34.11%

24.12% 46.58% 29.30% 24.14% 46.56% 29.30%

34.71% 32.24% 33.06% 34.68% 32.26% 33.07% BID

26.91% 34.08% 39.01% 26.92% 34.06% 39.02%

34.15% 36.93% 28.93% 34.14% 36.95% 28.92% PNJ

34.55% 29.75% 35.70% 34.57% 29.75% 35.68%

33.45% 30.11% 36.44% 33.45% 29.85% 36.70%

29.31% 28.47% 42.23% 29.30% 28.45% 42.25%

24.10% 41.77% 34.13% 24.11% 41.76% 34.13%

35.76% 28.90% 35.34% 35.79% 28.86% 35.35% OCB

4 Result

4.1/ Statistics on profitability of 10 stock codes.

The following are the results of the data processing process of calculating the profitability ratios of 10 stock codes, presented using the descriptive statistics method:

Table 1 Table calculating profit margins of 10 selected stock codes on HOSE according to descriptive statistics method

Stock

company

MIN -17.02% -17.14% 10.15% -13.50% -2.69% -31.67% -7.25% -7.15% -15.11% -36.56%

MEAN 0.03% 0.05% 0.06% 0.13% 0.08% -0.06% -0.04% 0.24% 0.15% -0.03%

MEDIAN 0.00% 0.00% 0.00% 0.24% 0.00% 0.00% 0.00% 0.29% 0.10% 0.00%

Source 1: Author.

Comment: The profit rate calculation table of 10 stock codes shows the viewer the highest (MAX), lowest (MIN), average (MEAN) and median (MEDIAN) profit rates

As we can see from the data from the table above, we can see that VCB, SSI and OCB respectively are the stocks with the highest profitability ratio exceeding 6% compared

to the remaining stocks In addition, it can be seen that the stock with the lowest profitability rate is OCB, only reaching -36.56% This shows that the risk is quite high when investing in OCB shares Besides, the stock code SSI has the highest average profit margin at 0.24% and the stock code TPB has the lowest average profit margin at -0.06%

4.2/ Describe the initial state of 10 tickers.

TABLE OF PROBABILITY MATRIX IN THE INITIAL STATE OF 10 STOCK

SYMBOLS

Initia

l

State

S1 30.80% 24.00% 34.80% 26.80% 34.40% 34.40% 33.60% 34.00% 24.00% 35.60%

S2 34.80% 46.80% 32.00% 34.00% 36.80% 30.00% 32.40% 28.40% 41.60% 29.20%

S3 34.40% 29.20% 33.20% 39.20% 28.80% 35.60% 34.00% 42.40% 34.40% 35.20%

After gathering data, we will rely on the collected information to analyze the securities

It will ultimately determine which stocks to invest in and ensure that the circumstances

under which investors should not invest avoid the worst possible consequences

4.3/ THE PROBABILITY OF THE FIRST NEXT DAY WHEN

APPLYING THE MARKOV MODEL.

(Unit: %).

The

first

next

day

S1 30.94 24.12 34.71 26.91 34.15 34.55 33.45 29.31 24.10 35.76

36.93

34.11 29.30 33.06 39.01 28.93 35.7 36.44 42.23 34.13 35.34

The graph below uses the MARKOV chain to explain the probability of price movements of the 10 different stocks in 3 states

Firstly, in state 1, it can be clearly seen that the stock that has the highest ratio is OCB with 35.76 percent, the second belongs to BIDV, one of the big 4 banks in Vietnam with 34,71 percent, the third is TP Bank with a probability of 34,55% In particular, it can be seen that FPT is the stock that has the lowest percentage with 24,1 This demonstrates that FPT is the best suited for the investor in the S1 model

Secondly, as can be seen from the graph, the probability in the S2 model has fluctuated quite a lot compared to S1, when FPT (24.1%) and ACB (24.12%) are the two stocks with the lowest probability in the model S1 Meanwhile, their probabilities are opposite

in model S2, accounting for the two highest rates respectively 41.77% (FPT) and 46.58% (ACB) Besides, they also found that SSI is the stock code with the lowest ratio of 28.47% SSI is followed by OCB - the stock code of Orient Bank with 28.9% Overall, we have two options for investors in case S2, there are SSI and OCB stock codes

Finally, the S3 model – is the most expected because of its return rate We can easily clarify the opposite position of rate in the S2 and S3 models of SSI stock After being the stock that had the lowest percentage in S2, SSI became the code that had the highest ratio in S3 with 42,23% Next to SSI is MBB in second place with 39,01% Furthermore, the lowest percent belongs to PNJ (28.93%) which has the potential to be

a suitable investment option in case S3 Besides, we also can consider about ACB (29.3%) and BID (33.06%)

4.4/ Describe the static state of 10 tickers.

THE MARKOV MATRIX TABLE DESCRIBES THE PROBABILITY OF PRICE MOVEMENTS OF 10 STOCKS LISTED ON HOSE IN A STEADY STATE

The

stead

y

state

S1 30.93

%

24.14

%

34.68

% 26.92

% 34.14

% 34.57

% 33.45

% 29.30

% 24.11

% 35.79

%

S2 34.96

%

46.56

%

32.26

% 34.06

% 36.95

% 29.75

% 29.85

% 28.45

% 41.76

% 28.86

%

S3 34.11

%

29.30

%

33.07

% 39.02

% 28.92

% 35.68

% 36.70

% 42.25

% 34.13

% 35.35

%

Looking at the MARKOV chart explaining the probability of price movements of the

10 stocks in Table S1, it can be seen that the fund with the highest probability of occurrence is OCB with a probability of 35.79%, followed by BID with a probability

of 34.68%, and then BID with a probability of 34.68% TPB with a probability of 34.57% In particular, it can be seen that the probability of the fund being FPT is the lowest at 24.11% This shows that the stock ticker is best suited to investing in the S1 model is FPT, followed by ACB stock code with a probability of 24.14%

Based on the matrix table above, it can be seen the probability of movement of 10 stocks in the S2 portfolio We can see that the fund with the highest probability is ACB with a probability of 46.56%, followed by FPT with 41.76% According to the table above, it can be seen that the fund with the lowest probability is SSI with 28.45%, and the fund with the second lowest probability is OCB with only 28.86% Thus, it can be seen that in the case of S2, the most invested stock code is SSI, followed by OCB which is also the stock code we can consider investing

Upon reaching the state of S3 and analyzing the MAKOV matrix table, a remarkable observation is made Although the ticker SSI seems to be the most profitable investment in the S2 state, it switches to high-risk stocks in the S3 state with an excess probability of 42.25% On the other hand, PNJ emerged as the stock with the lowest probability at 28.92%, indicating that it has the potential to be a favorable investment option in S3 The significant 13.33% difference between these two stocks further underscores PNJ's outperform Therefore, ACB shares with a probability of up to 29.30% will become a viable option for investment

PICTURE: COMPARE THE CHART MOVEMENT BASED ON S1 AND S3