=Paper=
{{Paper
|id=Vol-2393/paper_292
|storemode=property
|title=Forecasting Prices on the Stock Exchange Using a Trading System
|pdfUrl=https://ceur-ws.org/Vol-2393/paper_292.pdf
|volume=Vol-2393
|authors=Liubov Pankratova,Tetiana Paientko,Yaroslav Lysenko
|dblpUrl=https://dblp.org/rec/conf/icteri/PankratovaPL19
}}
==Forecasting Prices on the Stock Exchange Using a Trading System==
https://ceur-ws.org/Vol-2393/paper_292.pdf
Forecasting Prices on the Stock Exchange Using a
Trading System
Liubov Pankratova 1[0000-0002-1403-9454], Tetiana Paientko2[0000-0002-2962-308X], and Yaroslav
Lysenko 3
1
National University of Life and Environmental Science, 27 Heroiv Oborony st., Kyiv,
03041Ukraine
pankratova2105@gmail.com
2
Kyiv National Economic University named after Vadym Hetman, Peremohy avenu 54/1, Kyiv
02000 Ukraine
tpayentko109@gmail.com
1
National University of Life and Environmental Science, 27 Heroiv Oborony st., Kyiv,
03041Ukraine
zidfrih@gmail.com
Abstract. For successful trading on stock exchanges, it is important to use
trading tools that will ensure success in trading operations and provide competi-
tive advantages. The purpose of the article is to develop an algorithm for the
creation of a trading system and selecting a research object whose shares may
subsequently become the object of real trade. The basis of the developed trading
system is the consolidated mathematical model based on several models (multi-
pliers, neural network and discounted cash flows). The consolidated model es-
timates the stock price of NIKE Inc., which has a lower deviation from the ac-
tual price than the price is predicted by other mathematical models, including
linear regression models, etc. The results of the work also identified directions
for improving the trading algorithm: to extend the horizon of the forecast; to in-
clude TakeProfit at the predicted value; to form a stock portfolio; to cover more
factors in the model.
Keywords: stock exchange, trading system, forecast, consolidated mathemati-
cal model.
1 Introduction
Nowadays forecasting is being considered as one of the most important branches of
research in the economic and business fields and has been developing rapidly. Fore-
casting stock exchange prices by considering its dynamic factors is an important part
of a business investment plan. The confidence of investors in these markets has de-
clined and many negative problems in the world economy are present. This clearly
shows the strong relationship between uncertainty in financial markets and investor
confidence. Financial asset prices are influenced by numerous factors including peo-
ple behavior, political, economic, competition or other factors, so price forecasting
can be a difficult process.
Due to the development of stock trading, opportunities to receive a stock invest-
ment return exist. However, this is possible only with a properly selected trading
strategy and efficient trading system. Many traders diversify risks and increase profits
using several types of trading systems, which number more than a thousand today.
However, every trader or investor is trying to develop a unique trading system which
will allow anybody to successfully invest money by trading stocks with a correct price
forecast.
Forecasting prices allows not only individual financial asset price information to be
considered, but also financial and economic systems, and financial crises to be as-
sessed as to their possible scale in order to make appropriate economic decisions. At
the same time, the lack of a unified theory that would explain price fluctuations in
stock markets and a unified methodology for predicting prices for them determines
the expediency and necessity of further development of the methodology of forecast-
ing prices on stock exchanges.
The purpose of the article is to develop an algorithm for the creation of a trading
system and selection of a research object whose shares may subsequently become the
object of real trade. The paper is organized as follows. The next section explores the
theoretical background of forecasting prices in stock exchange. The third part de-
scribes the methodology of the research. The forth part is divided into three subsec-
tions. The first part analyzes the financial performance of the NIKE corporation and
shows correlations between the economic and financial indicators of this corporation.
The second part gives an assessment of the effectiveness of the developed forecasting
model. In the third part the forecast of stock prices is made with the help of the devel-
oped model.
The study has several limitations. First is the time period for forecasting. Secondly,
the testing was carried out using the shares of one corporation, not a portfolio, as an
example. Thirdly, the TakeProfit was not included at the predicted value.
2 Theoretical Background
Forecasting prices in stock markets is a matter of great interest both in the academic
field and in business. The forecasting of stock prices and stock returns is possible
using various techniques and methods. Many researchers study price trends in stock
markets with the help of artificial neural networks [1-2] or fuzzy-trends [3, 4]. The
application of artificial neural networks has become the most popular machine learn-
ing method, and it has been proven that such an approach can outperform most con-
ventional methods. The most popular neural network algorithm for financial forecast-
ing is the back-propagation algorithm. However, many articles have shown that the
artificial neural networks model, based on the back-propagation algorithm, has some
limitations in forecasting, and it can easily converge to the local minimum because of
the noise and complex dimensionality of the stock market data.
Many researchers use time-series models or other types of regressions [5-7]. Stock
market time series forecasting is an interesting and open research area. Artificial intel-
ligence algorithms are now mostly used to forecast time series. However, a highly
efficient stock exchange prediction model has not been designed yet.
Hybrid models have become more and more popular recently [8-9]. Kannan, Sekar,
Sathik and P. Arumugam in [10] used data mining technology to discover the hidden
patterns from the historic data that have probable predictive capability in their in-
vestment decisions. Usually, the rise or fall in an international stock market is caused
by some external factors. This means that stock exchange forecasting depends upon
local factors and international stock exchange markets. The robustness of forecasting
models remains an open research area that creates many approaches to design trade
systems for stock markets.
The trading system is based on a clear algorithm or, in other words, a clear set of
rules for generating trade signals (that is, the conditions for opening or closing a posi-
tion). The main difference between one trading system and another is its author's ap-
proach to the rules of trading signal generation [11]. Trading systems are based on
one or a limited number of algorithms. Fundamental and technical analyses are used
the most often in trading systems [12-15]. Also, genetic algorithms [16] and neural
networks and neuro-fuzzy computing [17] have become popular too. However, as
Kaufman mentioned, “most modeling methods are modifying cations of develop-
ments in econometrics and basic probability and statistical theory. They are precise
because they are based entirely on numerical data; however, they need trading rules to
make them operational. The proper assessment of the price trend is critical to most
trading systems” [18, p. 6]. A trading system, in the process of its operation, requires
constant debugging and analysis of completed transactions within a specified interval,
changing parameters for the following operations in order to maximally optimize the
intended trading strategy. Therefore, forecasting prices on the stock exchange with the
help of trading system of a trader will be a wide area for future research for a long
time.
3 Methodology
All trading systems operate according to their logic, that is, an algorithm that reports
to the system how to behave in different situations. The algorithms of trading systems
are developed based on the data obtained about events that previously occurred on the
stock market. The algorithm of creating a trading system includes the following stag-
es.
The first stage in the construction of a trading algorithm is the definition of a strat-
egy that will achieve a desired goal. The rules that formulate the strategy should be
set out consistently. The main rules are the rules surrounding entering and exiting
markets, that is, the terms of purchase and sale of stock commodities. Typically, a
trading strategy involves risk management by limiting the amount of risk capital. A
typical approach is to install a stop loss order that limits the maximum damage that is
allowed under the agreement. A trading strategy can also include revenue manage-
ment that protects the untapped profits generated during a lifetime position. A typical
approach to managing long-run profits is to establish a retractable stop-loss for a fixed
dollar value relative to the maximum of non-actualized profits. The purpose of our
strategy is to verify the correctness of the forecast of prices, and not real bargains, so
the stop-loss order was not used in our algorithm.
The best solution for the stock market is a strategy based on fundamental analysis,
which involves an analysis of the work of business entities, as well as external market
conditions. Two traditional forecasting models were used: Multiplier (M) and Dis-
counted Cash Flow (DCF). These are classic models that do not require complex cal-
culations or a large number of steps to calculate, although automation of calculations
of these models will save several days. Since it is impossible to fully evaluate the
effectiveness of stages such as testing or optimization, another model for forecasting
stock market prices, the mechanical neural network (NN), which is based on econom-
ic indicators of the enterprise, was used. Thus, our consolidated model for predicting
stock market prices (W) includes three models: multiplier, discounted cash flows and
a mechanical neural network. For comparison, the traditional linear regression model
(LM) is used (Fig. 1).
Consolidated
model
Mechanical
Discounted
Coefficients Neural
cash flows
Network
EV/EBITDA P/E
Fig. 1. Constituent methods of forecasting in the consolidated model
The second stage is to write an algorithm of action in the trading system in the pro-
gramming language R, using C ++ to extract data, to automate all processes and cal-
culations.
The third step in constructing the algorithm is testing the trading system. The test-
ing stage has two goals: the first one is to determine if the system performs the speci-
fied functions; the second one is to check the possibility of obtaining profits and the
risk of losses. The model should be moderately profitable with different price trends
and over several different time periods. Not necessarily every test should show profit,
but if each test is going to cause loss, then this system should be discarded.
Testing in various sectors of the economy has been used, which is necessary for the
possibility of wider uses of the algorithm. For testing, the sample was limited to 10%
of quarters, and at some periods of time it allows a better comparison of their reliabil-
ity. Periods were chosen randomly, by the algorithm, but in a way that the trend of
stock quotes and indices was versatile (downward, growing, lateral). This is necessary
for a better understanding of the efficiency of the algorithm in all types of market.
In the process of testing, an important step is to check the stability of the trading
system. A robust trading system will provide profits across a wide range of variables,
market segments, and market conditions. In other words, a sustainable model will
continue to show profitable results and in changing market conditions, which is an
extremely important result of trade. Thus, testing consists of two parts:
1) Selective manual check of various computer calculations of rules and formulas.
2) Investigation of the tested transactions and checking them for deviation from
the theory.
The first trading system test is the calculation of profit and loss on a segment of
price history of significant duration, for example, on an annual basis. This first test
gives a preliminary idea of profit and risk. The main rule is to proceed from the ex-
pectations of annual profits at the level necessary for trading in this market.
The fourth stage of work is the choice of the subject of testing and the definition of
the study period. The best market for research is the US stock market, as it is liquid
and has a long history that is needed for analysis. The main criterion for selecting
companies is the availability of electronic reporting and more than 20 years of quota-
tions on the stock exchange. Seven companies were selected by sectors of the econo-
my and their financial performance for 95 quarters (30.11.1994 - 30.08.2018) was
analyzed. These are the following companies:
1) AT&T Inc. (technology);
2) WALMART Inc. (wholesale and retail trade);
3) ECOLAB Inc. (means for water, hygiene, health, etc.)
4) BIOGEN Inc. (health care)
5) WELLS FARGO & COMPANY (finance)
6) NIKE Inc. (consumer goods)
7) CATERPILLAR Inc. (production goods).
To demonstrate the results of the analysis, NIKE Inc. was selected. The company's
revenue structure is simple in scope, but difficult in geography, that is, its financial
performance is influenced not only by the situation in the US but also in the world.
The fifth stage is the optimization of the trading system, which is carried out on the
same principles as testing, but the main task is to make the use of the trading system
most effective. In a practical sense, optimization is a process of calculating the indica-
tors of many different tests of this trading system on the same segment of price data.
According to certain criteria, the best test results, which provide maximum profit
potential in real trade, are selected, and they will be the basis for the optimization of
the trading system. The object of optimization is the coefficient of reliability of the
model, with which it is possible to achieve the best consolidated forecast.
The optimization has five components: (1) selection of model parameters; (2) set-
ting the ranges of their scanning; (3) determination of the sample size; (4) determina-
tion of criteria for evaluation, selection of a better model; (5) determine the criteria for
evaluating the test forecast as a whole. In the process of optimization, we should use
the model parameters that have the most impact on its effectiveness. If the parameter
has a small effect on efficiency, there are no reasons to make it a candidate for opti-
mization. Instead, it should be assigned a fixed value (constant) for optimization time.
If optimization shows improved results, it is time to move to the final step of the
testing process, namely, forward analysis. Forward analysis evaluates the effective-
ness of the trading system solely on the basis of post-optimization trading or test data
that are not part of the optimization sample. This level of testing answers two of the
most important questions about our trading system: 1) the correctness of the forecast
of prices 2) the possibility of profit after optimization.
The sixth stage is an assessment of the ratio of real trade indicators with projected
indicators. If the real figures differ much from the test ones for no clear reasons, then
a need to return to step number three is warranted.
The mathematical formalization of the processes embedded in the trading system
algorithm and designations used in the study are as follows:
1. On the basis of the current financial report, forecasts are made for three models
(NN, DCF, M)
2. The consolidated forecast price is based on formula (1):
Pr_Price = Knn * Pr_NN + Kdcf * Pr_DCF + Km * Pr_M, (1)
Pr_Price is the consolidated forecasted price,
K - coefficient of reliability of the model,
Pr_ - forecasted price by model
NN - model of mechanical neural network,
DCF - Discounted Cash Flow Model
M - model of multipliers.
3. Determination of projected income by formula 2:
Pr_Prof = | Pr_Price - Now_Price | / Now_Price, (2)
Pr_Prof is a projected income,
Pr_Price is the consolidated forecasted price,
Now_Price is the actual current price.
4. The decision to enter the market based on the assessment of the appropriateness
of investment, which is calculated by the formula 3:
Pr_Prof - Km-Slip> RF_Rate / 4, (3)
Km - commission for the opening and closing of a position,
RF_Rate - without a risky interest rate
Slip - slippage
RF_Rate is a risk-free investment rate.
5. Closing a position on the day on which the forecast was made.
The concept of exit from the market implies the absence of StopLoss and
TakeProfit, since the receipt of real profit is not the main goal, but only one of the
indicators for checking the efficiency of the trading system. The main goal is to de-
termine the price trend and price value for the planned closing date of the position.
Next, it is necessary to detail the information on one of the models, namely the
model of the mechanical neural network, which entered the consolidated model and
contains the largest number of economic indicators. Indicators that will be analyzed in
the neural network were selected based on the main indicators of financial reporting,
namely:
From the report on financial results:
1) Rev (Revenue) - Revenue;
2) Inc (Net Income) - net profit;
3) Div (Dividends declared per share (in dollars per share)) - dividends on ordinary
shares (in dollars per share).
From the balance:
1) Ast (Total assets) - aggregate assets;
2) Ldbt (Long-term debt) - long-term liabilities;
3) Sheq (Total shareholders' equity) - share capital;
4) Curl (Total current liabilities) - current liabilities.
From the Cash Flow Statement
1) Cash_op (Cash provided by operations) - cash flow from operating activities;
2) Cash_inv (Cash used by investing activities) - cash used in investment activities;
3) Cash_fin (Cash used by financing activities) - cash flow used in financial activi-
ties.
Additional indicators were also used such as:
1) Price_1 is a stock price at the time of the report's release;
2) S & P 500 is the stock index in which Nike is located;
3) Qw_1 / 2/3/4 - quarters of the marketing year of Nike;
4) Price_2 is a stock price for three days before payment of dividends for the fore-
casted quarter.
Additional indicators are needed to better understand the environment:
"Price_1" is required for the neural network to be able to track the price change and
understand what indicators have influenced it more,
"The S & P 500 Index" is needed to understand the situation in the economy and
the US stock market.
Quarters as indicators needed to understand the cyclicity algorithm present in this
market, as established by research. The above indicators are independent variables,
the only dependent variable in this model will be Price_2.
4 Efficiency Estimation Procedure
4.1 NIKE Inc. Financial and Economic Indicators Analysis
According to the New York Stock Exchange (NYSE), the shares of the company
grew more than 33 times (Figure 2) over the past twenty-four years, that is, the rate of
growth - an average of 16% annually. However, the highest growth rates have been
since 2009, when the company's products have gained popularity and spread around
the world.
Fig 2. The price dynamics of NIKE Inc., 11.1994 - 03.2018, USD. US / share (Source::
NYSE)
Interaction of Financial Results Indicators with NIKE Inc. depicted in Fig. 3.
Fig. 3. Correlation coefficients, scatter plot, and distribution histogram between revenue,
profit, dividends and share price of NIKE Inc.
Explanations to the figure 3:
Coefficient of corre- Correlation coeffi- Coefficient of
Revenue lation between revenue cient between revenue correlation between
and profit and dividends revenue and price
Coefficient of corre- Coefficient of
Scatter plot between
Profit lation between profit correlation between
revenue and profit
and dividends profit and price
Correlation coef-
Scatter plot between Scatter plot between
Dividends ficient between
earnings and dividends profits and dividends
dividends and price
Scatter plot between Scatter plot between Scatter plot between
Price
revenue and price profit and price dividends and price
As can be seen from the scatter plot, the revenues are positively interdependent
with net profit and stock price, with the profit being a linear dependence of dispropor-
tionate growth. On the contrary, with the price of the stock, the interdependence is
nonlinear, and the more accelerated rate of growth of prices from the growth of reve-
nue. The net profit also correlates positively with the company's price, this depend-
ence is also nonlinear and has hyperbolic acceleration function.
The Pearson correlation coefficients show that the largest share price correlates
with income (0.94), less correlated with net profit (0.7) and has a slight correlation
with dividends. In general, dividends moderately correlate with all indicators and are
placed on the chart rather chaotic.
A distribution histogram is located on the central diagonal. Here it is necessary to
note the full asymmetry on the right side (net profit and share price of the company)
due to the long-term presence of the company in the medium-sized business, divi-
dends and earnings are asymmetrical to the right, but they are more evenly arranged.
Next to be considered is how interrelated indicators balance with the predicted
price in Figure 4. The aggregate assets have a strong proportional relationship with all
the analyzed indicators; this is evident in the plot of scattering of the sample elements,
as well as by the coefficient of correlation, which is higher than 0.8. Scattering in
almost all cases is based on a linear function. There is a very interesting scatter plot
between long-term capital and equity capital; initially the values take hyperbolic ac-
celeration, then after 70% of the sample changes to the function of the hyperbolic
cosine region, that is, the inverse hyperbola, which in our opinion is associated with
an increase in the interest rate by the Fed of 2%. This influenced the decision on how
to raise funds for financing the range; the absence of less costly loans prompted inves-
tors to find investors in the stock market.
Fig. 4. Correlation coefficients, scatter plot and histogram of distribution between assets,
share capital, liabilities and share price of NIKE Inc.
Explanations to the figure 4:
Coefficient of
Coefficient of Correlation coeffi- Coefficient of
correlation be-
correlation be- cient between correlation be-
Assets tween assets and
tween assets and assets and current tween assets and
long-term liabili-
equity liabilities price
ties
Coefficient of Coefficient of Coefficient of
Scatter plot
correlation be- correlation be- correlation be-
between assets Long-term liabili-
tween long-term tween long-term tween long-term
and long-term ties
liabilities and liabilities and obligations and
liabilities
share capital current liabilities price
Scattering Scatter plot be- Correlation coeffi- Correlation coeffi-
Scale between tween long-term cient between cient between
Equity
Assets and liabilities and share capital and share capital and
Equity share capital current liabilities price
Scatter plot Scatter plot be- Scatter plot be- Coefficient of
between assets tween long-term tween current correlation be-
Current liabilities
and current liabilities and liabilities and tween current
liabilities current liabilities current liabilities liabilities and price
Scatter plot be-
Scatter plot Scatter plot be- Scatter plot be-
tween long-term
between assets tween current tween current Price
obligations and
and price liabilities and price liabilities and price
price
As in the figures in Figure 3, distribution histograms have right-side asymmetry
(Figure 4), which indicate that the company is growing.
Spearman's correlation calculations (Figure 5) showed a close correlation between
the price of a company's shares and independent variables, in particular the S & P 500
market index. This suggests that external factors are also influenced by the price of
NIKE shares.
Fig. 5. Spearman correlation coefficient between the investigated parameters
4.2 Estimation of Efficiency of the Developed Model of the Forecast of Prices
One of the objectives of the study is to evaluate the effectiveness of the model, and
one of the best methods of evaluation is a comparison with the traditional model. The
classic method of forecasting, which is widely used in many spheres, is linear regres-
sion. Therefore, a regression model of stock price forecast was constructed.
Further work was aimed at constructing, testing, optimizing, testing the reliability
of the consolidated model, which included several models: neural, multiplicative and
discounted cash flows with the help of computer equipment. In Table 1, projected
prices for different models and the actual price at the end of the projection period can
be compared. The date of forecasting is chosen independently by the program.
The final step in optimization is to determine the weight of each model in the con-
solidated forecast model, that is, at the forecasted price. The values of the mean-
square deviation are as follows: for the neural network - 0,778; for multipliers - 0.134;
for discounted cash flows - 0.088. This indicates the greatest impact of the neural
model on the consolidated weighted price. This model, after optimization, gave the
most accurate forecast.
Table 1. Actual and forecasted stock price figures for NIKE Inc. during test periods,
US $ / share
Data Pr_P_NN Pr_P_M Pr_P_DCF PR_P_LM Pr_Price Now_Price
31.05.2001 5,175 5,358 6,403 4,553 5,307 6,313
31.05.2002 6,186 6,555 8,241 6,036 6,416 5,719
28.02.2003 7,180 5,679 6,548 6,051 6,923 6,498
31.05.2007 13,366 12,797 14,820 11,208 13,417 13,53
30.11.2007 15,731 14,777 18,244 17,081 15,824 15,175
29.02.2008 16,753 13,561 15,816 15,965 16,243 16,23
30.11.2009 15,830 14,847 16,230 15,729 15,733 16,143
30.11.2010 20,236 21,092 21,269 21,346 20,442 22,073
30.11.2015 63,028 66,428 66,328 67,843 63,774 62,46
Figure 6 illustrates the deviation of the consolidated forecast price (Pr_Price) and
the price calculated by the linear model (Pr_P_ LM) from the actual price
(Now_Price). As can be seen from Figure 6 and Table 1, the forecasted prices for the
model we have developed are deviating less from the actual price line than the fore-
casted prices constructed according to the linear model, which indicates the undenia-
ble advantages of the first model.
-030 -020 -010 000 010 020
Deviation Pr_Price from Now_Price
Deviation Pr_P_ LM from Now_Price
Fig. 6. Comparison of forecasted prices based on the consolidated model and the linear
model with respect to actual prices for shares of NIKE Inc.,%.
A trading system based on a weighted model correctly identified the direction of
price movement in 44% of cases, while 22% of trade signals differed from real price
dynamics directly opposite, and 34% of trading signals differed slightly from the real
price movement. It should also be noted that 66% of the predicted values indicated the
presence of a bear market, while the actual values in only 33% of observations in the
next period showed a falling trend.
For comparison, the trading system based on a linear model correctly predicted the
direction of prices only in 33% of the cases, another 33% of trading signals differed
directly opposite from the real price dynamics, and the remaining 34% had a slight
deviation from the real movement of prices. In one case, the trading system recom-
mended not entering the market while during this period there was a bullish trend, so
earnings opportunities would have been lost, and in two cases the system recom-
mended taking a short position, when in fact the market during the quarter was in a
side movement.
4.3 Forecast of Prices Using the Developed Trading Model
Since the testing did not give a precise assurance of the efficiency of the model, a
forward test had to be carried out, that is, an analysis of the forecasted price within the
actual time period that was not investigated nor considered. This is best done on the
broker's demo account using the whole sample of data (95 quarters) to predict, not the
10% as used during testing.
For the forecast, a period was chosen that was not used in the model (31.08.2018-
30.11.2018) and a quarterly forecast of stock prices of NIKE Inc. was made. (Figure
7, Table 2).
Pr_P_LM
90.427
Now_Price Pr_P_DCF
85.587
75.12
83.979
78.928
Pr_P_NN Pr_P_M
80.191
Pr_Price
Fig. 7. Estimated price models and actual share price for NIKE Inc. as of 30.11.2018,
USD US / share
Table 2. Comparison of NIKE Inc.'s forecasted price models and actual prices. as of
30.11.2018, USD US / share
Actual price Estimated Actual price
Rejection of
Designation (Now_price) price (Now_price)
projected prices
of models (as of as of as of
from actual
31.08.2018 30.11.2018 30.11.2018
Pr_P_DCF 82,20 85,59 75,12 10,47
Pr_P_M 82,20 83,98 75,12 8,86
Pr_P_NN 82,20 78,93 75,12 3,81
Pr_Price 82,20 80,19 75,12 5,07
Pr_P_LM 82,20 90,43 75,12 15,31
Thus, the biggest difference between the forecast and actual price as calculated
based on linear regression model was $15.31 per share. The smallest deviation from
the projected price from the actual demonstration was the model based on the neural
network at $3.81 per share. This model has the greatest impact on the consolidated
weighted price, so the deviation from the actual price was $5.07 per share.
Both the neural network model and the consolidated model predicted lower prices
at the end of the examined quarters than at the beginning. However, all other models
predicted a bullish trend, which proved to be false. The joint impact of model multi-
pliers (M) and the discounted cash flow model (DCF) had a negligible effect on the
consolidated model.
During the period in which the forward analysis was conducted there were no sig-
nificant changes in the company's policy nor strategy, and the projected figures for the
following year were not revised. However, there was a negative marketing impact
when the company’s major advertising face, Cristiano Ronaldo, was accused of per-
sonal income tax evasion. This news alone could have provoked a downward trend in
prices that we could predict using a trading system based on our consolidated model,
which showed better results than the linear regression model, and the two models,
neural network and discounted cash flows, which had been tested separately
5 Conclusions
Consequently, the created trading algorithm is capable of allowing predictions to be
made of the company's share price with a fairly high accuracy using the consolidated
model. The largest influence on the company's share price is made by indicators such
as revenue, net profit and aggregate assets for which the correlation coefficient is
more than 0.95. The company has a sound dividend policy, so dividend changes have
little effect on price dynamics.
The results of the work also identified directions for improving the trading algo-
rithm.
1) In further research, we plan to extend the horizon of the forecast, as the stock
market is one of the best environments for long-term investment. It will also simplify
the calculations, namely, deviating from the analysis of quarterly indicators to annual.
2) In many cases, test tests, if we put TakeProfit at the predicted value, we could
profit before, and sometimes even avoid losses. Therefore, in the future, we plan on
using two models to determine the expediency of buying shares, and the neural net-
work to determine when it is best to buy or sell shares.
3) In order to diversify risks, it is expedient to use not only shares of one company
for forecasting but also to form a stock portfolio for investing funds.
4) The broader coverage of the fundamental factors that will be included in the
model for analysis will only improve the results of the work.
Improving the trading system will allow more accurate forecasts and, accordingly,
more effective investments in the stock market.
References
1. Strzelchyk, A., Strzelchyk, Ar.: Trends in the stock market and their price forecasting us-
ing artificial neural networks, Central and Eastern European Journal of Management and
Economics Vol. 1, No. 2, 155-164 (2013).
2. Lin, Y., Guo, H., Hu, J.: An SVM-based approach for stock market trend prediction. In
Proceedings of the 2013 International Joint Conference on Neural Networks (IJCNN), Dal-
las, TX, USA, 4–9 August 2013; IEEE: Piscataway, NJ, USA (2013).
3. Guan, H., Dai, Z., Zhao, A., He, J.: A novel stock forecasting model based on high-order-
fuzzy-fluctuation trends and back propagation neural network. PLoS ONE 13 (2018).
4. Kodogiannis, V., Lolis, A.: Forecasting financial time series using neural network and
fuzzy system-based techniques. Neural Comput. Appl. 11, 90–102 (2002).
5. Gong, X., Si, Y.-W., Fong, S.: Biuk-Aghai, R.P. Financial time series pattern matching
with extended UCR Suite and Support Vector Machine. Expert Syst. Appl, 55, 284–296
(2016).
6. Wen, Q., Yang, Z., Song, Y.: Jia, P. Automatic stock decision support system based on
box theory and SVM algorithm. Expert Syst. Appl, 37, 1015–1022 (2010).
7. Wang, L., Wang, Z., Zhao, S., Tan, S.: Stock market trend prediction using dynamical
Bayesian factor graph. Expert Syst. Appl, 42, 6267–6275 (2015).
8. Huang, C.-F.: A hybrid stock selection model using genetic algorithms and support vector
regression. Appl. Soft Comput, 12, 807–818 (2012).
9. Chiang, W.-C., Enke, D., Wu, T., Wang, R.: An adaptive stock index trading decision sup-
port system. Expert Syst. Appl, 59, 195–207 (2016).
10. Kannan, К.S., Sekar, S., Sathik M. M. and Arumugam, P.: Financial stock market forecast
using data mining Techniques, Proceedings of the international multiconference of engi-
neers and computer scientists (2010).
11. Leigh, W., Hightower, R. and Modani, N.: Forecasting the New York Stock Exchange
Composite Index with Past Price and Interest Rate on Condition of Volume Spike, Expert
Systems with Applications, Vol 28, pp. 1-8 (2005).
12. Tseng, K-C., Kwon, O., and Tjung, L.C.: Time Series and Neural Network Forecast of
Daily Stock Prices, Investment Management and Financial Innovations, Vol 9, No 1, pp
32-54 (2012).
13. Klassen, M.: Investigation of some technical indexes in stock forecasting using neural
network. Proceedings of World Academic of Science, Engineering and Technology 5:75-
79 (2005).
14. Reznik N. and Pankratova L.: High-Frequency Trade as a Component of Algorithmic
Trading: Market Consequences. CEUR Workshop Proceedings, vol. 2104, P. 78-83..
Available: http://ceur-ws.org/Vol-2104 (2018).
15. Westerhoff, F. H.: Multi-Asset Market Dynamics, Macroeconomic Dynamics, 8/2011, pp.
596—616 (2011).
16. Thompson, J.R., Wilson J.R. and Fitts, E. P.: Analysis of market returns using multifractal
time series and agent-based simulation, Proceedings of the Winter Simulation Conference
(WSC '12). Winter Simulation Conference, Article 323 (2012).
17. Lento, C.: A Combined Signal Approach to Technical Analysis on the S&P 500, Journal of
Business & Economics Research, 6 (8), pp. 41–51 (2008).
18. Vezeris, D., Kyrgos T. and Schinas Ch.: Take Profit and Stop Loss Trading Strategies
Comparison in Combination with an MACD Trading System J. Risk Financial Manag. 11,
56, 2-23. doi:10.3390/jrfm11030056 (2008).
19. Mendes, L., Godinho P. and Dias, J.: A Forex trading system based on a genetic algorithm,
Journal of Heuristics 18 (4), pp. 627-656 (2012).
20. Badawy, O. and Almotwaly, A.: Combining neural network knowledge in a mobile col-
laborating multi-agent system, Electrical, Electronic and Computer Engineering, ICEEC
'04, pp. 325, 328, DOI: 10.1109/ICEEC.2004.1374457 (2004).
21. Kaufman, P. J.: Trading systems and methods / Perry J. Kaufman. — 5th ed (2013).