Key Indicators Influencing BRICS Countries' Stock Price Volatility

through Classification Techniques: A Comparative Study

NURSEL SELVER RUZGAR

Ted Rogers School of Management,

Toronto Metropolitan University,

350 Victoria Street, Toronto, ON M5B 2K3,

CANADA

Abstract: - The stock market is crucial for a country’s economy. It reflects the economic health and investment

status of a country. While it has attracted the interest of many scholars, the volatility of stock prices and the

indicators influencing this volatility has not been extensively studied, particularly using classification

techniques. This study aims to fill this gap in the literature by identifying an effective classification technique to

classify the data of BRICS countries using eight classification techniques via WEKA software from 2000 to

2021. Additionally, the study seeks to explore the common indicators that significantly impact stock price

volatility in BRICS countries. Findings reveal that tree algorithm-based techniques performed well in terms of

accuracy and reliability, although no single common classification technique was identified. Among the eight

techniques, Random Tree classified the data of BRICS countries with high accuracy, except for India, where

the J48 technique was more efficient. Furthermore, the study indicates that there are no common indicators

affecting stock price volatility, as these indicators vary across countries due to the distinct economic and

sociopolitical structures of BRICS countries. These findings provide valuable insights for investors and

policymakers to better understand and manage stock market dynamics in BRICS countries.

Key-Words: - Stock price volatility, Classification, Data Mining, BRICS countries, Indicators, WEKA

Received: September 28, 2023. Revised: April 29, 2024. Accepted: May 25, 2024. Published: June 27, 2024.

1 Introduction

The stock market plays a crucial role in the

country’s economy and is extremely sensitive to

social, economic, and political news events around

the world. It works as a mirror to reflect the

country’s socio-economic and political structures.

In existing literature, a myriad of works studied

the stock markets of emerging economies in

different aspects, however, the indicators

influencing the stock price volatility (SPV) have not

been extensively explored. To fill the gap in the

literature, this study examines indicators influencing

the SPV of BRICS countries from 2000 to 2021. As

emerging economies, BRICS countries have been

selected because they have diverse political,

demographic, and economic structures, [1]. The

acronym BRICS is the abbreviation of five fast-

growing markets in the universe of emerging market

economies, [2]. The story of BRICS started in the

early 2000s and Goldman Sachs analyst Jim O’Neill

coined the term BRIC for the group of Brazil,

Russia, India, and China to describe four fast-

growing countries, [3]. In 2010, South Africa joined

that group and formed the acronym BRICS, [4]. In

terms of economic structures, Brazil has a

liberalized and market-driven economic structure,

Russia, India and China have dominant government-

controlled economic structures and South Africa has

a driven, structured, and open economic structure,

[5]. BRICS countries have more than 40% of

world's population, 28% of the world's massive land

24% of the global GDP, and more than 16% of

global commerce, [6]. They established the New

Development Bank in 2015 to mobilize resources

for infrastructure and sustainable development

projects, [7]. Being major recipients of foreign

investments, they play an important role in the

current pattern of global investments, [8].

The rest of the paper is structured as follows:

Section 2 presents the literature review; Section 3

describes the data and methodology of the study;

Section 4 presents the findings and discussion of

empirical results; Section 5 presents the

conclusions, research, and recommendations.

2 Literature Review

The economic structures of countries, particularly

emerging countries such as BRICS, have become a

focal point for many scholars. Research on this topic

WSEAS TRANSACTIONS on BUSINESS and ECONOMICS

DOI: 10.37394/23207.2024.21.122

Nursel Selver Ruzgar

E-ISSN: 2224-2899

1494

Volume 21, 2024

has been approached from a various perspective.

Scholars have searched bank development and stock

market performance [1] and the correlation between

oil volatility and stock markets, [9], [10].

Additionally, studies have explored the interplay

between economic growth, financial development,

and income inequality, [11], as well as the

relationship between exchange rates and stock

market indices, [12], [13], [14], [15]. Other areas of

investigation include volatility indices [14], the

impact of crises on economic growth [13], [16],

[17], and the effects of innovations on economic

growth, [18]. Moreover, scholars have examined

stock market efficiency [19], [20], [21], [22], green

finance and climate change, [23], [24], [25] and the

relationship between tourism and economic growth,

[26]. Additionally, some scholars focus on the

effects of exchange rates, [20], [27], the

interconnectedness between Bitcoin and equity

markets, [28], the impacts of private credit shocks

on economic growth, [29] while some others have

studied the relationship between entrepreneurial

activity and economic growth [30], the effects of

financial shocks [31] and the impacts of exports and

imports on economic growth, [32]. Other than those,

some scholars have worked on the relationship

between trade openness and economic growth, [33],

[34], the impact of market crashes on sector indices

and volatility [35] and stock market efficiency

before and after the COVID-19 pandemic, [36].

Stock markets and banks have been key drivers

of economic growth and financial system

development. Bank development is measured by

credit facilities to the private sector relative to GDP,

while stock market development is assessed by

market size and liquidity, [1]. In existing literature,

the bank and stock market indicators have not been

studied together. This study addresses this gap by

using both indicators to explore which common

indicators impact stock price volatility in BRICS

countries. One scholar studied in cross-country and

panel form the interactions of bank development,

stock market development, and global equity index

for the BRICS countries from1990 to 2018, [1].

Their findings indicated that the models for bank

development and market performance respond

differently in the short term compared to the long

term. Furthermore, they concluded that the growth

of the global stock market is predominantly

influenced by the global financial situation rather

than the development of banks within BRICS

countries.

Among the BRICS countries, Russia and Brazil

are net oil exporters, while India, China, and South

Africa are net oil importers. Consequently, their

stock returns react oppositely to changes in oil

volatility. One scholar examined the quantile

dependence and directional predictability from oil

volatility to stock returns in BRICS countries using

the cross-quantilogram model, [37]. The findings

reveal that low quantile oil volatility has a minimal

impact on stock returns, whereas high quantile oil

volatility amplifies losses in stock returns.

Moreover, the influence of oil volatility on stock

returns varies depending on whether a country is a

net exporter or importer of oil. Oil volatility is not

included in this study because its impact on the SPV

of BRICS countries varies depending on whether

they are net oil exporters or net oil importers.

The BRICS countries have experienced years of

rapid trade and economic growth, now accounting

for nearly a quarter of the global economy, [8].

These factors make the BRICS countries key

players in global investment, as they are major

recipients of foreign direct investment and

increasingly significant for outward investors, [8].

The exchange rate is a crucial factor for investors.

Several scholars have explored the relationship

between exchange rates and stock market indices

within BRICS countries. Reference [12], assessed

the impact of exchange rates on stock market returns

using the auto-regressive distributed lag (ARDL)

method, concluding a significant effect of exchange

rates on stock market indices returns. Similarly, the

information linkages of the volatility index, a

forward-looking measure of volatility, across

BRICS countries were examined using a

multivariate generalized autoregressive conditional

heteroscedasticity model, [14]. This research

highlighted varying degrees of connectedness

among BRICS countries over the study period.

Similar to the reference [12], another scholar

worked on the dynamic linkages between exchange

rates and stock market returns in a regime-switching

environment across BRICS countries, [15]. The

findings suggested that stock markets have more

influence on exchange rates during both calm and

turbulent periods. In the existing literature, scholars

have used various methods, such as the auto-

regressive distributed lag model [12], the dynamic

five-factor parametric model, the multilayer feed-

forward neural network [8], and the multivariate

generalized autoregressive conditional

heteroscedasticity model [14], to perform their

analyses. However, there is a lack of comparison

among the data mining (DM) methods and

classification techniques. To address this gap, this

study compares DM classification techniques to

identify a common method for classifying the SPV

data of BRICS countries.

WSEAS TRANSACTIONS on BUSINESS and ECONOMICS

DOI: 10.37394/23207.2024.21.122

Nursel Selver Ruzgar

E-ISSN: 2224-2899

1495

Volume 21, 2024

The efficiency of stock markets has a significant

interest among researchers. Reference [14],

conducted a comparative study on the efficiency of

stock markets across BRICS countries, evaluating

the profitability of four trading rules: Simple

Moving Average, Relative Strength Index, Moving

Average Convergence Divergence, and Momentum,

spanning from 1995 to 2008. The findings revealed

that these indicators were most profitable in the

Russian stock market. Additionally, they identified

the Brazilian stock market as the most efficient

among the BRIC countries. In 2020, Panda

examined the financial structure of BRICS countries

and the USA was examined to focus on investment

opportunities within BRICS from 1997 to 2017,

[20]. Considering factors such as market depth,

market microstructure, portfolio weights, and

various macroeconomic indicators, it was found that

all BRICS countries, except Brazil, exhibited high

investment rates. The financial and macroeconomic

indicators suggest that BRICS countries are

attractive destinations for investors, offering

substantial economic value. These findings

contributed to the findings in [17]. They also

highlighted high investment rates in BRICS

countries, excluding Brazil, [17]. In another study,

the relationship between gold and stock markets in

BRICS countries was explored using weekly data

from 2000 to 2014, [21]. The research indicated that

dynamic conditional correlations between gold and

stock markets were generally low to negative during

significant financial crises, suggesting that gold

could serve as a safe haven during times of extreme

market volatility.

Several scholars have examined individual stock

markets within the BRICS countries during financial

crises. For instance, the financial contagion effects

on African stock markets, including South Africa's,

during global financial crises such as the European

debt crisis, Brexit, and the COVID-19 pandemic

were studied, [38]. The findings indicated that the

regional impact of these crises varied based on their

nature. Financial contagion was observed to

increase with country-level risk, market

capitalization, and export-to-GDP ratio, but

decreased with lower corruption levels. Moreover,

financial interconnectedness was investigated by

analyzing volatility spillovers and movements

between equity and foreign exchange markets in

BRICS countries from 1997 to 2018, [39]. That

research showed that shocks originating from equity

markets had a stronger impact on foreign exchange

markets at the individual level. Conversely, foreign

exchange markets had a greater influence on their

corresponding equity markets. Interdependencies

were found between the equity and foreign

exchange markets of most BRICS countries, except

for China, which displayed relative isolation. Brazil

was identified as the main source of volatility

spillovers to other BRICS markets, while South

Africa showed the highest level of integration

within the BRICS countries. These findings show

the diverse social, economic, and geopolitical

structures within BRICS countries. Under the

consideration of diverse economic and sociopolitical

structures in BRICS countries, two questions arise

regarding the analysis techniques and the indicators

affecting their stock price volatility (SPV): Is there a

common classification technique that can classify

the SPV data for all BRICS countries? And is it

possible to identify common indicators that affect

the SPV of BRICS countries? This paper aims to

answer these questions by applying eight data

mining (DM) classification techniques to twenty-

seven indicators.

3 Aim and Methodology

The BRICS countries hold strategic significance in

the global economy. As a group of fast-growing

emerging economies, they account for 26% of the

global GDP and 40.8% of the world's population.

Numerous scholars have studied their economies,

stock markets, and trade volatilities using various

methods. However, there is a lack of comparative

studies on stock price volatilities using DM

classification techniques, as well as a lack of

studies identifying common indicators affecting the

SPV of BRICS countries. To address these gaps,

this study aims to identify a common classification

technique for classifying the SPV data of BRICS

countries using eight distinct DM techniques

through the open-source software WEKA.

Additionally, the study seeks to determine which

common indicators have the greatest impact on

SPV in BRICS countries based on the

classification outcomes. To achieve these

objectives, the following hypotheses are

formulated:

H1: There are common indicators that significantly

influence the SPV of BRICS countries.

H2: A common classification technique effectively

classifies the data of BRICS countries.

3.1 Data

The annual data from 1994 to 2022 were collected

from The World Bank Indicators and the OECD

Data Bank, [5]. To ensure data quality, only

indicators with minimal missing values were

WSEAS TRANSACTIONS on BUSINESS and ECONOMICS

DOI: 10.37394/23207.2024.21.122

Nursel Selver Ruzgar

E-ISSN: 2224-2899

1496

Volume 21, 2024

included in the statistical analyses, so the data from

2000 to 2021 were utilized to perform the analysis.

27 indicators were chosen based on availability for

each country. Table 1 presents these 27 indicators

along with their definitions and SPV. The dependent

variable in this analysis is SPV. The SPV data were

initially collected in US dollars. However, since

many classification techniques require binary data in

WEKA, the SPV data were transformed into a

categorical format by calculating the difference

between the current year's SPV and the previous

year's SPV. A positive difference was denoted by

"P", and a negative difference was denoted by "N".

Table 1. Indicators and their definitions

As highlighted in Section 2, the BRICS

countries are key players in global investment, both

as major recipients of foreign direct investment and

as significant outward investors, [8]. Trade is also

vital for these economies. Brazil depends on

exporting commodities and agricultural products.

Similarly, Russia's economy is predominantly based

on natural resources and commodity exports, [20].

China and India benefit from cheap labor and

resources, exporting manufactured goods,

agricultural products, technology, and services.

South Africa's economy is diverse, relying on

mineral resources like gold, platinum, and

diamonds, as well as agriculture, tourism, and

financial services. Given this context, the

independent variables, X1-X27, were meticulously

selected from financial and economic indicators that

influence SPV and are available for each BRICS

country. These include ten economic and seventeen

financial indicators. The economic indicators are

Exports of Merchandise (Customs) in current US

dollars (X1), GDP at market prices in current US

dollars (X2), Imports of Merchandise (Customs) in

current US dollars (X3), Nominal Effective

Exchange Rate (X4), Total Reserves (X5), Real

Effective Exchange Rate (X6), Stock Market

Capitalization to GDP percentage (X24), Stock

Market Return percentage (X25), Stock Market

Total Value Traded to GDP percentage (X26), and

Stock Market Turnover Ratio percentage (X27). The

remaining indicators are financial, as detailed in

Table 1.

3.2 Classification Techniques

Classification in Data Mining (DM) is an important

technique that groups the data into distinct

categories by employing mathematical and

statistical techniques. The main objective of

classification is to accurately predict the target class

for each data, [40]. It enables accurate predictions

and better decision-making. In literature, various

classification methods have been proposed. In this

study, WEKA implementation software is used to

classify SPV data of BRICS countries. WEKA is a

DM implementation software program developed by

the University of New Zealand under the General

Public License, [41]. In WEKA, there exist seven

classification algorithms, namely bayes, functions,

lazy, meta, misc, rules, and trees, each including

various sub-classification modules. WEKA employs

many classification techniques depending on the

data type, nominal ordinal, or interval. Based on the

data type, a suitable algorithm can be chosen to

effectively extract information from the dataset. For

this study, all applicable classifications were utilized

on the dataset, but only eight of them demonstrated

effective classification. These are Naïve Bayes

(NB), Simple Logistic (SL), Meta-Bagging (MB),

Classification via Regression (CVR), Decision

Table (DT), Decision Stump (DS), J48, and Random

Tree (RT).

The NB is a sub-classification module of the

bayes algorithm. The NB technique is based on the

Bayesian theorem of probability. The Bayesian

Network Classifier efficiently computes the most

likely output based on the input. In the NB, the

presence of a particular attribute is considered

independent of the presence of any other attribute

when the class variable is given. Bayesian networks

are directed acyclic graphs (DAGs) where nodes

represent random variables, [42]. The edges denote

conditional dependencies, while unconnected nodes

represent independent variables. Each node is

associated with a probability function that provides

the probability of the variable it represents. For the

WSEAS TRANSACTIONS on BUSINESS and ECONOMICS

DOI: 10.37394/23207.2024.21.122

Nursel Selver Ruzgar

E-ISSN: 2224-2899

1497

Volume 21, 2024

NB, numeric estimator precision values are chosen

based on the analysis of the training data. This

algorithm can be used when the variables are binary

class, missing class values, and nominal class. The

SL is a sub-classification module of the functions

algorithm. For the SL, LogitBoost with simple

regression functions as base learners are used for

fitting the logistic model. In SL classification, the

model predicts the probability of each class. The

logistic function transforms the linear combination

of the input features into a probability value

between 0 and 1. The logistic function is defined as:

󰇛  󰇜

 (1)

where 󰇛  󰇜 represents the probability of the

dependent variable being 1 given the values of the

independent variables X and z is the linear

combination of the independent variables and their

coefficients.

        (2)

where  represents the intercept,      are

the coefficients corresponding to the independent

variables     , respectively, [43]. This

algorithm can be used when the dependent variable

is binary and analyzes the data after employing

discretization of the continuous variables. The MB

is a sub-classification module of meta algorithm, it

can do classification and regression depending on

the base learner. Similar to Naïve Bayes, it is used

when the variables are binary class, missing class

values, nominal classes or numeric classes. The

CVR is also subclassification module of the meta

algorithm. It is binarized and one regression model

is built for each class value. It is used when the

variable binary class, missing class values, and

nominal class. In addition, the DT is a sub-

classification module of the rules algorithm,

whereas the DS is a sub-classification module of the

tree algorithm. The DT is usually used in

conjunction with a boosting algorithm. The goal of

DT is to create a model that estimates the value of a

target variable based on several input variables. The

DT is used when the variables are binary class,

missing class values, nominal class, or numeric

class. Meanwhile, DS does regression based on

mean-squared error, and missing is treated as a

separate value. It is used when the variables are

binary class, missing class values, nominal class, or

numeric class. The J48 is also a sub-classification

module of the tree algorithm and it uses the rules of

the C4.5 algorithm. Similar to DS and J48, the RT is

also a sub-classification module of the tree

algorithm. It performs no pruning. It also has an

option to allow estimation of class probabilities.

They are capable use of using different data types.

NB, SL, CVR, and J48 manage binary and nominal

datasets with missing values, whereas MB, DT, DS,

and RT manage binary, and nominal numerical data

sets with missing values.

3.2.1 Classifier Performance Measures

Classifier performance is measured by using the

accuracy of correctly classified instances, mean

absolute error (MAE), root mean square error

(RMSE), and various performance metrics such as

precision, recall, F-statistic, ROC Area, and

confusion matrix. The confusion matrix presents a

visualization of the classification performance based

on a table that contains columns representing the

instances in a predicted class and rows representing

the instances in an actual class. Classification

accuracy refers to the ratio of correct predictions to

the total number of predictions made. It is calculated

as the percentage of correctly predicted instances

over the total number of instances. In literature, 80%

is assumed as the threshold point, [44]. If it is close

to 100% the accuracy rate is an overwhelming

situation to say that the data are perfectly classified.

Another criterion is the kappa statistic, which

measures the agreement between observed and

expected classification outcomes. It varies from -1

to +1, [45]. The kappa statistic value can be

interpreted as follows: values ≤ 0 as indicating no

agreement and 0.01-0.20 as none to slight, 0.21-0.40

as fair, 0.41- 0.60 as moderate, 0.61-0.80 as

substantial, and 0.81-1.00 as almost perfect

agreement, [46]. ROC curve is another performance

measure. It measures the overall performance of the

classifier according to the area under the curve. It

can be used to compare two or more class

performances. The area under the curve is the highest

and the best classifier. The range of values for the

area under the curve changes from 0 to 1. 1 indicates

the classifier is perfect. A ROC curve can be used to

select a threshold for a classifier that maximizes

the

true

positives

while

minimizing

the

false

positives, [47]. Precision quantifies the number of

correct positive predictions made whereas recall

quantifies the number of incorrect positive

predictions made from all positive predictions that

could have been made. Maximizing precision will

minimize the number of false positives, whereas

maximizing the recall will minimize the number of

false negatives. Like precision and recall, a poor F-

Measure score is 0.0 and a best or perfect F-

Measure score is 1.0, [47].

WSEAS TRANSACTIONS on BUSINESS and ECONOMICS

DOI: 10.37394/23207.2024.21.122

Nursel Selver Ruzgar

E-ISSN: 2224-2899

1498

Volume 21, 2024

3.3 BRICS Countries

The BRICS countries have a significant role in the

world economy and they are the most powerful and

fastest-growing emerging markets, [48]. The BRICS

countries stand as significant recipients of global

investment streams and are major consumers of

commodities worldwide, [1]. According to World

Bank data, the BRICS countries have different

demographic and economic structures. Table 2

illustrates some demographic and economic

indicators in 2022 for BRICS countries and the

world, [5].

Table 2. Demographic and economic indicators of

BRICS countries and the World in 2022

In terms of population size, China and India lead

with vast populations of approximately 1.5 billion

each, while Brazil, Russia, and South Africa have

smaller populations (Table 2). Population growth

rates differ among these countries. South Africa has

the highest growth rate at 0.841% and India has the

second highest growth rate at 0.680%. Brazil's

growth rate is modest at 0.46%, and Russia's is

much lower at 0.074%. Meanwhile, China

experiences a slight decline with a growth rate of -

0.013%, [49]. In terms of land area, Russia is the

largest, followed by China, Brazil, India, and South

Africa, [50]. Economically, China leads with the

highest GDP of approximately $17.96 trillion US

dollars, followed by India at $3.4 trillion US dollars.

Brazil and Russia lag behind with $2.24 trillion US

dollars and $1.92 trillion US dollars, respectively.

South Africa has the smallest GDP at $405 billion

US dollars. Among the BRICS countries, Russia is

the only country with a negative GDP growth rate,

primarily due to sanctions following its invasion of

Ukraine in February 2022, [50]. India boasts the

highest growth rate at 7.24%, followed by China,

Brazil, and South Africa at 2.99%, 2.90%, and

1.91%, respectively. Globally, the population is

nearly 7.95 billion with an annual population growth

rate of 0.79% and a total area of approximately 140

million square kilometers. The global GDP stands at

about $101.33 trillion USD, with a growth rate of

3.087%. In terms of demographic composition, the

BRICS countries cover 28.25% of the world's land

area and account for over 40% of the global

population, [5]. The BRICS countries show diverse

economic and socio-political structures, each with

distinct natural resources. Their economies depend

on factors such as industrialization, commodities,

trade openness, exports, and imports. These lead to

distinct growth drivers for each country, [12]. For

instance, Brazil has rich natural resources, such as

iron ore, soybeans, and oil. It is a major exporter of

agricultural products, minerals, and manufactured

goods, [12]. Russia also has rich natural sources,

particularly crude oil and gas, [12]. It exports

mainly oil, gas, various commodities, metals, and

military equipment. India has cheap laborlabor and

exports commodities, like textiles, chemicals,

machinery, and software services. Similarly, China

benefits from a large labor pool and is a leading

exporter of electronics, machinery, textiles, and

other goods. South Africa has diverse resources, like

gold, platinum, and diamonds, along with strengths

in agriculture and tourism. It exports agricultural

products, as well as gold, platinum, and diamonds,

[12], [17].

3.4 Stock Price Volatilities of BRICS

Countries

The stock market is a mirror of the economy and

wellness of a country. The dynamic of stock price

volatility depends on many factors, such as

exchange rates, oil prices export and import

volatilities, trading, and assets. Risk is another

factor highly influential on stock price volatility. In

highly risky situations, domestic or foreign investors

do not invest which leads to a volatility decrease.

For this study, annual SPV data of BRICS

countries in US dollars were received from the

World Bank and OECD Data from 1994 to 2022. To

account for volatility variations among countries

and missing data for some countries before 2000,

the data were standardized to percentages, with the

year 2000 serving as the reference point for

aggregation. Figure 1 illustrates the SPV of BRICS

countries from 2000 to 2021. In 2000, due to

aggregation, the SPV of BRICS countries showed

the unique economic conditions and market

responses in each country.

The SPV of BRICS countries demonstrated

fluctuations during the study period from 2000 to

2021 due to three economic and financial crises, the

2007-2009 global financial crisis, the 2014-2016

crisis, and the 2019-2020 COVID-19 pandemic

crisis. The economic structure and stock markets of

countries face significant impacts from global

economic disturbances and financial market

instability. These disruptions lead to fluctuations in

various sectors, such as declines in SPV and

WSEAS TRANSACTIONS on BUSINESS and ECONOMICS

DOI: 10.37394/23207.2024.21.122

Nursel Selver Ruzgar

E-ISSN: 2224-2899

1499

Volume 21, 2024

production, increases in unemployment and

inflation. The SPV of BRICS countries has also

been affected by these crises. The SPV of Brazil

started at 100% in 2000, decreased to 79.72% in

2001, and fluctuated moderately until a peak of

118.89% in 2009. This is coming from the effects of

the global crisis. After 2009, the SPV of Brazil

generally declined until 2018 and then slightly

increased to 82.95% in 2021. During this period,

political scandals (Operation Car Wash), and

unemployment increases affected seriously its

economy. The SPV of Russia shows a more volatile

pattern. It increased to a peak of 193.59% in 2004,

then dropped significantly after 2008. It reached

49.12% in 2010 and stabilized in the 30-60% range

afterward. The war between Russia and Georgia,

economic sanctions due to in 2014 annexation of

Crimea, and the ongoing Russia-Ukrainian war

drastically affected Russia’s economy and SPV. As

a big commodity exporter, oil price decreases also

affected its economy seriously. The SPV of India

shows a similar trend to Russia until 2016, then

increased until 2020. It had a pick value 184.91% in

2004. Between 2005 and 2011, India's SPV showed

a general declining trend, stabilizing around 66-68%

then it has a pick at 121.62% in 2020. From 2000 to

2004 ease China’s SPV showed a gradual decrease,

then increased to the pick value of 147.74% in 2009.

It fluctuated between 62-106% after 2015. From

2000 to 2004, South Africa's SPV remained stable

with minor fluctuations, then reached a low of

70.46% in 2005. It displayed moderate volatility

until 2009 and peaked at 175.88% and 119.95% in

2009 and 2020, respectively. It generally stabilized

around 70-120% after 2009. These trends highlight

how each country's unique economic, political, and

market conditions influence their SPV.

Fig. 1: Stock price volatility percent at constant

2000 prices, [5]

4 Findings and Discussion

The following subsections present detailed

outcomes from eight classification techniques via

WEKA.

4.1 Naïve Bayes

The initial classification technique employed is the

Naïve Bayes (NB), utilizing the full training dataset

for evaluation. Table 3 displays the results of NB for

BRICS countries. The accuracy of correctly

classified instances for BRICS countries ranged

from 67.86% to 89.28%. India shows the highest

accuracy, followed by South Africa, Russia, China,

and Brazil in descending order. The Kappa statistic,

which measures the agreement between observed

and expected classification results, shows that

India's classification has the highest agreement at

0.7813, followed by South Africa at 0.7083. Russia

has a moderate agreement at 0.6%, whereas China

and Brazil demonstrate low agreements at 0.43%

and 0.27%, respectively. Additional performance

metrics include the mean absolute error (MAE) and

root mean squared error (RMSE). These metrics

measure the average deviation of predicted values

from actual values in classification, with lower

values indicating better classifier performance. India

and South Africa show lower MAE and RMSE

compared to the other countries. The true positive

(TP) rate and false positive (FP) rate provide

information about the sensitivity and specificity of

the classification model.

Table 3. Classification results of Naïve Bayes

aCCI:Correctly classified instances, bICI: Incorrectly classified

instances, c MAE: Mean absolute error, dRMSE: Root mean

squared error, eTP: True positive, hFP: False positive: increase

in SPV, gN: decrease in SPV

WSEAS TRANSACTIONS on BUSINESS and ECONOMICS

DOI: 10.37394/23207.2024.21.122

Nursel Selver Ruzgar

E-ISSN: 2224-2899

1500

Volume 21, 2024

In Table 3, TP and FP rates were presented for

Positive (P) and Negative (N) SPV data. India,

South Africa, and Russia have high-levellevels of

TP rates for Positive and Negative SPV data. The

performance of the classifier is evaluated using

metrics such as Precision, Recall, F-measure, and

ROC Area, which measure correctness and

completeness. A high precision suggests fewer false

positives, while a high recall indicates fewer false

negatives. The F-Measure shows a balance between

precision and recall, and a high ROC Area signifies

good classifier performance.In comparing these

metrics, India shows the highest performance,

followed by South Africa and Russia. The

performance of NB is less satisfactory for China and

Brazil. The classifier effectively classifies SPVs of

India, South Africa, and Russia, while classifying

SPV of China moderately. Brazil shows the lowest

correctly classified instances and generally

underperforms in comparison to the other countries.

4.2 Simple Logistic

The Simple Logistic (SL) is the second

classification technique that uses the full training

data. Table 4 shows the results of SL across BRICS

countries. In terms of correctly classified instances,

South Africa demonstrates a high accuracy of

96.43%, and closely Russia follows with 93.99%

accuracy. For Brazil, the classifier achieved a good

accuracy of 85.71%. Meanwhile, India and China

exhibit moderate accuracies of 78.57% and 67.86%,

respectively.

Table 4. Classification results of Simple logistic

aCCI:Correctly classified instances, bICI: Incorrectly classified

instances, c MAE: Mean absolute error, dRMSE: Root mean

squared error, eTP: True positive, hFP: False positive: increase

in SPV, gN: decrease in SPV

Similar to accuracies, Kappa statistic is high for

South Africa and Russia. Brazil and India show

moderate Kappa statistics, while Kappa statistics for

China are comparatively low. The TP rate and FP

rate for South Africa and Russia are excellent in

distinguishing positive and negative instances.

Brazil and India also show good TP and FP rates,

but China demonstrates low TP and FP rates. The

performance metrics for SL range from strong to

weak across the countries in the following order:

South Africa, Russia, Brazil, India, and China.

Based on the results in Table 4, South Africa stands

out as the best-performing country, followed by

Russia and Brazil. The SL generated two sets of

equations based on the output: one for positive SPV

values and another for negative SPV values. The

main difference between these sets is in the signs of

the coefficients of the indicators. The following

logistic equations were achieved for positive values

of SPV.

         

       

    

(3)

         

       

   (4)

     (5)

         

       

 (6)

        

       

     (7)

According to the logistic equations (3)-(7), Bank

capital to total assets percent (X7) emerges as a

common indicator for BRICS countries, with the

exception of China. Bank capital to total assets

percent (X7) has a positive effect on Russia but a

negative effect on Brazil, India, and South Africa.

Additionally, Non-life insurance premium volume to

GDP percent (X17) and Pension fund assets to GDP

percent (X19) are shared indicators for Brazil, India,

and South Africa. Both of the indicators positively

impact these three countries. For Brazil, the most

influential indicators, whether positive or negative,

include Non-life insurance premium volume to GDP

percent (X17) and Remittance inflows to GDP

percent (X23), Nominal Effective Exchange Rate

WSEAS TRANSACTIONS on BUSINESS and ECONOMICS

DOI: 10.37394/23207.2024.21.122

Nursel Selver Ruzgar

E-ISSN: 2224-2899

1501

Volume 21, 2024

(X4), Pension fund assets to GDP percent (X19),

and Bank capital to total assets percent (X7). In

Russia, the key indicators are Gross portfolio debt

liabilities to GDP percent (X12), Bank capital to

total assets percent (X7), Gross portfolio equity

liabilities to GDP Percent (X13), and Mutual fund

assets to GDP percent (X16). India's key indicators

include Non-life insurance premium volume to GDP

percent (X17), Bank capital to total assets percent

(X7), Insurance company assets to GDP percent

(X14), Pension fund assets to GDP percent (X19),

and Mutual fund assets to GDP percent (X16).

Lastly, for South Africa, the primary indicators are

Non-life insurance premium volume to GDP percent

(X17), Remittance inflows to GDP percent (X23),

Bank capital to total assets percent (X7), and

External loans and deposits of reporting banks vis a

vis the nonbanking sectors percent of domestic bank

deposits (X9).

The results indicate that key indicators vary

across the countries. The variations in indicators

depend on the political and socio-economic

structures of the countries, as well as ongoing

expected and unexpected global events like the

Russia-Ukraine war and the COVID-19 pandemic.

These events have altered the profile of indicators

influencing the SPV. The results of SL indicate that

no common indicators were identified across all

BRICS countries. Consequently, this finding rejects

the hypothesis H1: “There are common indicators

that significantly influence the SPV of BRICS

countries.”

4.3 Meta Bagging

The Meta Bagging (MB) utilized the full training

dataset for evaluation. Table 5 presents the

classification results of MB across BRICS countries.

According to the classification results of MB, the

correctly classified instances (CCI) range from

78.57% to 92.86% accuracies. South Africa showed

the highest value of CCI, while Brazil and China

demonstrated the lowest values. India and Russia

fall in between. The Kappa statistic is highest for

South Africa (0.8511) and lowest for China

(0.5435).

Although MAE and RMSE are lowest for South

Africa, suggesting better overall model

performance, they are very close to South Africa for

the other BRICS countries. When considering TP

and FP rates, South Africa achieves the highest

value while Brazil and China show a wider range.

This indicates higher sensitivity and a higher FP

rate. India and Russia show good performances.

Although precision, recall, F-measure, and ROC

area vary across the countries, they generally

perform well. South Africa achieves the highest

values compared to the other countries, followed by

India, Russia, China, and Brazil. Based on the

classification results of MB in Table 5, South Africa

demonstrates strong classification results, followed

by India, Russia, China, and Brazil. Therefore, the

results show that MB classified the SPV data for

BRICS countries effectively.

Table 5. Classification results of Meta Bagging

aCCI:Correctly classified instances, bICI: Incorrectly classified

instances, c MAE: Mean absolute error, dRMSE: Root mean

squared error, eTP: True positive, hFP: False positive: increase

in SPV, gN: decrease in SPV

4.4 Classification via Regression

The fourth classification technique is Classification

via Regression (CVR). Full training data and M5

pruned model tree were used for evaluation. Table 6

illustrates the classification results of CVR.

In terms of CCI, India and South Africa achieved

the highest percentages with both 85.71% accuracy.

Russia and Brazil follow with 80% and 75%

accuracy, respectively. China has the lowest CCI

(71.43%). Kappa statistic shows that India has the

highest agreement between observed and expected

accuracy with 0.7068. South Africa and Russia

follow India with the Kappa statistic values of

0.6957 and 0.600, respectively. However, the

Kappa statistics for Brazil (0.4615) and China

(0.3600) are very low, which indicates a low

agreement between observed and expected

accuracy. MAE and RMSE are the lowest for South

Africa and low for India, which implies better

overall performance, compared to the other BRICS

countries. When considering the TP rate and FP

WSEAS TRANSACTIONS on BUSINESS and ECONOMICS

DOI: 10.37394/23207.2024.21.122

Nursel Selver Ruzgar

E-ISSN: 2224-2899

1502

Volume 21, 2024

rates, South Africa, India, and Russia demonstrate

higher rates, whereas Brazil and China display

lower rates. In terms of the performance metrics,

South Africa shows the highest values, closely

followed by India. China shows the lowest values in

these metrics. When considering the classification

performance of CVR, it shows varied performances

for BRICS countries.

Table 6. Classification results of Classification via

Regression

aCCI:Correctly classified instances, bICI: Incorrectly classified

instances, c MAE: Mean absolute error, dRMSE: Root mean

squared error, eTP: True positive, hFP: False positive: increase

in SPV, gN: decrease in SPV

4.5 Decision Table

The fifth classification technique employed is the

Decision Table (DT). It uses both the full training

data and the M5 pruned model tree for evaluation.

Table 7 illustrates the classification results of DT.

Russia and South Africa show the highest

percentage of CCI, with accuracies of 80% and

78.57%, respectively. Brazil’s accuracy is the lowest

(60.71%), while China’s and India’s fall in between.

The Kappa statistic ranges from 0.4011 to 0.6000,

indicating fair to low agreement between observed

and expected accuracy. Russia has the highest

Kappa statistic, whereas China has the lowest. MAE

and RMSE metrics show the average magnitude of

errors for all. Russia has the lowest values,

indicating good performance, whereas Brazil has the

highest.

In terms of TP rate and FP rate, China

demonstrates the best performance, followed by

Russia and India. Precision, recall, and F measures

for positive and negative instances demonstrate

unbalanced performances. While Brazil and Russia

show high performance in positive instances, India,

China, and South Africa show high performance in

negative instances. ROC area values for the BRICS

countries vary. When considering the positive and

negative instances Russia, China, and India

demonstrate strong discriminatory abilities, while

Brazil and South Africa display weaker

performance. Comparing the classification of DT

with other classification techniques, it appears that

other techniques outperform the DT. Although the

results of the DT vary across countries, Russia and

India demonstrate better classification results

compared to other BRICS countries.

Table 7. Classification results of Decision Table

aCCI:Correctly classified instances, bICI: Incorrectly classified

instances, c MAE: Mean absolute error, dRMSE: Root mean

squared error, eTP: True positive, hFP: False positive: increase

in SPV, gN: decrease in SPV

4.6 Decision Stump

Decision Stump (DS) is the sixth classification

technique which uses both the full training data and

the M5 pruned model tree for evaluation. Table 8

shows the results of DS across BRICS countries.

The results in Table 8 show that Russia achieves

the highest accuracy of 80%, followed closely by

South Africa at 78.57%. Brazil, India, and China all

have the same accuracy of 75%. The Kappa statistic

indicates moderate to low agreement between

observed and expected accuracy, ranging between

0.4615 and 0.6002. Among the BRICS countries,

Russia demonstrates the highest Kappa statistic,

whereas Brazil shows the lowest. The MAE and the

RMSE show the average magnitude error. Russia

WSEAS TRANSACTIONS on BUSINESS and ECONOMICS

DOI: 10.37394/23207.2024.21.122

Nursel Selver Ruzgar

E-ISSN: 2224-2899

1503

Volume 21, 2024

has the smallest value, and Brazil has the highest

value. When considering TP Rate and FP Rate for

positive and negative instances, Russia and China

show high rates. However, there is an unbalanced

gap between the TP rate and FP rate for Brazil,

India, and South Africa. Precision, Recall, and F-

Measure demonstrate reasonable performance.

Russia, Brazil, and China show better performances

in these metrics than India and South Africa. ROC

Area values vary across the countries. South Africa

and Russia show better classification ability than the

other countries for both positive and negative

instances. In terms of the classification ability of the

DS, Russia and South Africa were categorized with

high performance compared to the other countries.

However, when compared with previous

classification techniques, other techniques exhibit

better performances than DS, similar to the findings

with DT.

Table 8. Classification results of Decision Stump

aCCI:Correctly classified instances, bICI: Incorrectly classified

instances, c MAE: Mean absolute error, dRMSE: Root mean

squared error, eTP: True positive, hFP: False positive: increase

in SPV, gN: decrease in SPV

4.7 J48

The seventh classification technique is the J48

which utilizes full training data for evaluation. Table

9 presents the results of the J48 across BRICS

countries.

According to the findings in Table 9, the CCI

ranges between 92.86% and 75%. India leads with

the highest percentage, performing well at 92.86%,

followed by Brazil at 89.29%. China and Russia

follow with accuracies of 82.14% and 80%,

respectively. South Africa demonstrates the lowest

CCI at 75%. The Kappa statistic displays very good

agreement for India, followed by Brazil. While

China and Russia show moderate agreement levels,

South Africa shows the least agreement between

observed and expected accuracy. The MAE and the

RMSE indicate the average magnitude of errors.

India and Brazil show the better performance,

followed by China and Russia. India, Brazil, and

Russia demonstrate very high TP Rate and FP Rate,

while China also shows good rates. However, South

Africa shows the lowestperformance in these

metrics. In terms of Precision, Recall, and F-

Measure metrics, the performance rankings from

highest to lowest are as follows: India, Brazil,

China, Russia, and South Africa. Additionally, the

ROC Area metric indicates strong performance

across the countries. India and Brazil show the

highest ROC Areas, followed by China, Russia, and

South Africa.

Table 9. Classification results of Decision Stump

aCCI:Correctly classified instances, bICI: Incorrectly classified

instances, c MAE: Mean absolute error, dRMSE: Root mean

squared error, eTP: True positive, hFP: False positive: increase

in SPV, gN: decrease in SPV

Based on the performance metrics of BRICS

countries obtained by J48 as shown in Table 9, this

classification technique demonstrated impressive

performance compared to the previous classification

techniques. According to the outputs of J48, the key

indicators affecting SPV vary for the countries. In

Brazil, Financial system deposits to GDP percent

(X10), Gross portfolio debt liabilities to GDP

percent (X12), and Non-life insurance premium

volume to GDP percent (X17) are influential. For

Russia, Gross portfolio debt liabilities to GDP

WSEAS TRANSACTIONS on BUSINESS and ECONOMICS

DOI: 10.37394/23207.2024.21.122

Nursel Selver Ruzgar

E-ISSN: 2224-2899

1504

Volume 21, 2024

percent (X12), Pension fund assets to GDP percent

(X19), and Stock market turnover ratio percent

(X27) play significant roles. For India, Gross

portfolio equity liabilities to GDP percent (X13) and

Non-life insurance premium volume to GDP percent

(X17) are influential. The SPV of China is impacted

by External loans and deposits of reporting banks

vis a vis the nonbanking sectors' percent of domestic

bank deposits (X9) and Mutual fund assets to GDP

percent (X16). Lastly, the Nominal Effective

Exchange Rate (X4), Liquid liabilities to GDP

percent (X15), Provisions to nonperforming loans

percent (X22), and Stock market capitalization to

GDP percent (X24) influence the SPV of South

Africa.

4.8 Random Tree

Random Tree (RT) is the eighth classification

technique employed in this study. RT uses the full

training data for evaluation. Table 10 displays the

classification results of RT for BRICS countries.

According to the classification results in Table

10, Brazil, Russia, China, and South Africa achieve

perfect percentages of CCI, indicating perfect

accuracy. Conversely, India shows a lower

percentage of 71.4%, with 8 instances incorrectly

classified. The Kappa statistics show perfect

agreement between observed and expected accuracy

for all BRICS countries, except India. India shows

fair agreement. The MAE and the RMSE metrics are

very low for South Africa, Brazil, Russia, and

China. This indicates that they have an

approximately perfect performance. However, these

metrics are higher for India. Similarly, TP Rate and

FP Rate show perfect true positive rates and no false

positive rates for all countries except India.

Precision, Recall, and F-Measure metrics indicate a

strong balance between precision and recall for most

countries, whereas India demonstrates low values.

ROC Area values are excellent for all countries

except India.

Based on the results in Table 10, the RT appears

to be a powerful classifier for the SPV of BRICS

countries compared with the previous classification

techniques. The outputs from the RT show

significant indicators affecting the SPV of BRICS

countries. In Brazil, the indicators such as Exports

Merchandise Customs current US dollars millions

not seasonally adjusted (X1), GDP at market prices

current US dollars millions of seasonally adjusted

(X2), Nominal Effective Exchange Rate (X4), Total

Reserves (X5), Central bank assets to GDP percent

(X8), External loans and deposits of reporting banks

a vis the nonbanking sectors percent of domestic

bank deposits (X9) and Stock market turnover ratio

percent (X27) demonstrate influence. Similarly, in

Russia, the influential indicators include Exports

Merchandise Customs current US dollars millions,

not seas adj (X1), GDP at market prices current US

dollars millions of seasonally adjusted (X2),

Nominal Effective Exchange Rate (X4), Total

Reserves (X5), Real Effective Exchange Rate (X6)

and Private credit by deposit money banks and other

financial institutions to GDP percent (X20). For

China, the indicators affecting SPV are Exports

Merchandise Customs current US dollars millions

not seas adj (X1), Imports Merchandise Customs

current Us dollars millions not seasonally adjusted

(X3), Central bank assets to GDP percent (X8),

Financial system deposits to GDP percent (X10),

Liquid liabilities to GDP percent (X15) and Stock

market return percent year on year (X25). In the

case of India, Non-life insurance premium volume

to GDP percent (X17) influences SPV, while for

South Africa, the influential indicators are Liquid

liabilities to GDP percent (X15), Remittance inflows

to GDP percent (X23), and Stock market total value

traded to GDP percent (X26).

Table 10. Classification results of Random Tree

aCCI:Correctly classified instances, bICI: Incorrectly classified

instances, c MAE: Mean absolute error, dRMSE: Root mean

squared error, eTP: True positive, hFP: False positive: increase

in SPV, gN: decrease in SPV

4.9 Comparison

Kappa statistics, accuracy, and RMSE are used to

compare the classification techniques. To identify an

effective classification technique for BRICS

countries, these metrics serve as the basis for

comparison.

WSEAS TRANSACTIONS on BUSINESS and ECONOMICS

DOI: 10.37394/23207.2024.21.122

Nursel Selver Ruzgar

E-ISSN: 2224-2899

1505

Volume 21, 2024

Table 11. Comparison of classification techniques

bRMSE: Root mean squared error,

Table 11 shows an overview of these evaluation

metrics. The top-performing and the second-best

performing classification techniques were

highlighted in blue and orange, respectively. When

accuracy exceeds 80%, the classification technique

demonstrates strong data classification.

Approaching or achieving 100% accuracy indicates

nearly perfect classification. Similarly, A Kappa

statistic nearing 1 displays strong agreement

between observed and expected classification

results. The RMSE measures the average deviation

of predicted values from actual values in

classification, and a low value is expected for

optimal classifier performance. Based on the

evaluation criteria, for Brazil, RT shows the perfect

accuracy, then J48 follows it with the accuracy of

89.29%. SL also performs well, achieving 85.71%

accuracy. In Russia, RT demonstrates perfect

performance with 100% accuracy, followed by SL

with 93.99% accuracy. For India, J48 leads with a

high accuracy of 96.86%, followed by NB with an

accuracy of 89.28%. In the case of China, RT

achieves perfect accuracy, while J48 follows with

82.14% accuracy.

Lastly, for South Africa, RT demonstrates the

perfect accuracy and SL follows RT with 96.43%

accuracy. The results indicate varying effectiveness

of classification techniques across countries. While

RT achieves perfect classification for Brazil, Russia,

China, and South Africa, J48 performs best for India

with an accuracy of 96.86%. There is no common

classification technique, therefore hypothesis H2:

“A common classification technique effectively

classifies the data of BRICS countries.” is rejected.

5 Conclusion

The results of eight classification techniques reveal

that there is no common classification technique

categorizing the SPVs of BRICS countries. Among

these techniques, RT provided promising results and

perfectly categorized the SPV of BRICS countries

except India which was effectively classified by J48.

When ranking the classifiers based on their

performance from high to low, the order is RT, J48,

SL, MB, and CVR. This variability can be attributed

to the diverse economic, social, and political

structures of the countries. For instance, Brazil has

been dealing with high inflation, political

fluctuations, and slow industrialization. The

ongoing war between Russia and Ukraine led Russia

to be the most sanctioned country in the world,

which is impacting its economy, GDP growth, and

stock market. Meanwhile, South Africa faces

challenges, such as high inflation, unemployment,

reduced trade and fluctuation in financial flows, and

increasing public expenditure, [51].

By analyzing the outcomes of the top three

effective classification techniques, RT, J48, and SL,

the pivotal indicators significantly impacting the

SPVs of BRICS countries were identified. While no

common indicator was identified, “Exports

Merchandise Customs current US dollars millions

not seasonally adjusted” (X1) is prominent for

Brazil, Russia, and China based on the outputs of

RT. Additionally, “GDP at market prices current US

dollars millions seasonally adjusted” (X2) has an

impact on the SPVs of Brazil and Russia, while

“Central bank assets to GDP percent” (X8) is

common to the SPV of Brazil and China. “Liquid

liabilities to GDP percent” (X15) and “Stock market

return percent, year on year” (X25) are common

indicators for China and South Africa. In addition to

the commonly shared indicators, several others play

significant roles in individual countries' SPV.

In Brazil, key indicators include “External loans

and deposits of reporting banks vis a vis the

WSEAS TRANSACTIONS on BUSINESS and ECONOMICS

DOI: 10.37394/23207.2024.21.122

Nursel Selver Ruzgar

E-ISSN: 2224-2899

1506

Volume 21, 2024

nonbanking sectors percent of domestic bank

deposits” (X9) and “Stock market turnover ratio

percent” (X27). Russia's key indicators include

“Nominal Effective Exchange Rate” (X4), “Real

Effective Exchange Rate” (X6) and “Private credit

by deposit money banks and other financial

institutions to GDP percent” (X20). For China, the

important indicators are “Imports Merchandise

Customs current US dollars millions not seasonally

adjusted” (X3) and “Financial system deposits to

GDP percent” (X10). Lastly, for South Africa, the

key indicators are “Remittance inflows to GDP

percent” (X23) and “Stock market total value traded

to GDP percent” (X26). The J48 revealed fewer

common indicators influencing the SPV compared

to the RT. “Gross portfolio debt liabilities to GDP

percent” (X12) is common for Brazil and Russia,

while “Non-life insurance premium volume to GDP

percent” (X17) is common for Brazil and India. The

remaining influential indices vary for each country.

For India, “Non-life insurance premium volume to

GDP percent” (X17) is consistent with both RT and

J48 results. Regarding the outputs of the SL, “Bank

capital to total assets percent” (X7) is a shared

indicator for all countries except China.

Trade, import, and export play important roles in

the economies of BRICS countries. Brazil and

Russia are net oil exporters, while China, South

Africa, and India are net oil importers. Fluctuations

in oil prices significantly affect the economic

balance and currency. The classification outcomes

from RT indicate that nominal and real exchange

rates have an impact on Brazil’s and Russia’s SPV

which contributes to the literature [14], [51], [52],

[53].

This paper has several limitations. Firstly, due to

the unavailability of daily, monthly, or quarterly

data for some countries, annual data were used for

analysis. Secondly, missing values were observed in

the datasets from 1994 to 2022. To address this

issue annual data from 2000 to 2021 were used for

analysis. To maintain the integrity of the original

data during the analysis process, random variables

or means were not assigned to the missing values.

Although there is no standardized technique for

classifying the SPV data of BRICS countries and no

common indicators have been identified, the

findings will assist investors and policymakers in

understanding market conditions, especially during

periods of fluctuation, and managing stock market

dynamics.

5.1 Future Research and Recommendations

The inclusion of new five members, Egypt,

Ethiopia, Iran, Saudi Arabia, and the United Arab

Emirates (UAE), effective from January 1st, 2024,

BRICS countries are likely to attract many scholars’

attention in the future. According to 2022 World

Bank Data, with this extension, the new population

of BRICS countries became 45.50% of the world

population and their land area represents 32.61% of

the world’s land area. In the “Situation Report”, it

was stated that BRICS countries with the new

members represent 28.1% of the global economy

and the expanded group holds more than 43% of

global oil production, [7].

Future research could enhance comparisons by

including additional indicators, such as oil prices,

trade volatility, and unemployment. Utilizing the

high frequency data is strongly recommended for

further studies. Replicating this study with an

expanded set of BRICS countries could provide

valuable insights. Additionally, alternative methods

such as ARCH/ GARCH models, multiple

regression, and structural analysis could be

employed and their results could be compared.

References:

[1] E. B. Kalu, C. A. Augustine, E. U. O. Okoro,

F. I. Onaga, and C. A. Felix, A cross-country

and country specific modelling of stock

market performance, bank development and

global equity index in emerging market

economies: A case of BRICS countries, PLoS

ONE, Vol. 15, No. 11, 2020. e0240482,

https://doi.org/10.1371/journal.pone.0240482.

[2] W. Mensi, H. Shawkat, D. K. Nguyen, and S.

H. Kang, Global financial crisis and spillover

effects among the U.S. and BRICS stock

markets, International Review of Economics

and Finance, Vol. 42, 2016, pp. 257–276,

http://dx.doi.org/10.1016/j.iref.2015.11.005.

[3] J. O'neill, Building better global economic

BRICs, Global Economic Papers, Vol. 66,

2001, pp. 1-16.

[4] I.V. Dimitrios, and G. Lakshmi, Market risk

of BRIC Eurobonds in the financial crisis

period, International Review of Economics

and Finance, Vol. 39, 2015, pp. 295–310,

http://dx.doi.org/10.1016/j.iref.2015.04.012.

[5] Data: World Development Indicator, Database

of World Development Indicators, World

Bank national accounts data, 2022, org/public

-licenses#were identifiedcc-by, [Online].

https://data.worldbank.org/indicator/NY.GDP.

MKTP.KD?locations=BR-CN-IN-RU-ZA and

https://datacatalog.worldbank (Accessed Date:

December 16, 2023).

WSEAS TRANSACTIONS on BUSINESS and ECONOMICS

DOI: 10.37394/23207.2024.21.122

Nursel Selver Ruzgar

E-ISSN: 2224-2899

1507

Volume 21, 2024

[6] World Economic Outlook, imf.org. April 16,

2024, [Online].

https://www.imf.org/en/Publications/WEO

(Accessed Date: February 19, 2024).

[7] A. G. Ross, Will BRICS expansion finally end

western economic and geopolitical

dominance?, Situation Reports, January 26,

2024, [Online].

https://www.geopoliticalmonitor.com/will-

brics-expansion-finally-end-western-

economic-and-geopolitical-dominance/

(Accessed Date: February 3, 2024).

[8] O. Castello, and M. Resta, Modeling the yield

curve of BRICS countries: Parametric vs.

machine learning techniques, Risks, Vol. 10,

No. 2, 2022, pp.1–18,

https://doi.org/10.3390/risks10020036.

[9] G. C. Aye, M. Balcilar, R. Gupta, N.

Kilimani, A. Nakumuryango, and S. Redford,

Predicting BRICS stock returns using

ARFIMA models, Applied financial

economics, Vol. 24, No: 17, 2014, pp. 1159–

1166,

https://doi.org/10.1080/09603107.2014.92429

[10] X. Chen, X. Sun, and J. Wang, Dynamic

spillover effect between oil prices and

economic policy uncertainty in BRIC

countries: A wavelet-based approach,

Emerging Markets Finance and Trade, Vol.

55, No: 12, 2019, pp. 2703–2717,

https://doi.org/10.1080/1540496X.2018.15649

04.

[11] M. Younsi, and M. Bechtini, Economic

growth, financial development, and income

inequality in BRICS countries: Does Kuznets’

inverted U-shaped curve exist?, Journal of the

Knowledge Economy, Vol. 11, 2020, pp. 721–

742.

[12]

M. Mroua, and L. Trabelsi,

Causality and

dynamic relationships between exchange rate

and stock market indices in BRICS countries

Panel/GMM and ARDL analyses, Journal of

Economics, Finance and Administrative

Science, Vol. 25, No. 50, 2020, pp. 395-412

[Online].

https://www.emerald.com/insight/content/doi/10

.1108/JEFAS-04-2019-0054/full/html

(

Accessed Date: February19, 2024).

[13] N. S. Ruzgar, and C. Chua-Chow, Behavior of

banks’ stock market prices during long-term

crises International Journal of Financial

Studies, Vol. 11, No. 31, 2023,

https://doi.org/10.3390/ijfs11010031.

[14] G. Sharma, P. Kayal, and P. Pandey,

Information linkages among BRICS countries:

Empirical evidence from implied volatility

indices, Journal of Emerging Market Finance,

Vol. 18, No. 3, 2019,

https://doi.org/10.1177/0972652719846315.

[15] W. Chkili, and D. K. Nguyen, Exchange rate

movements and stock market returns in a

regime-switching environment: Evidence for

BRICS countries, Research in International

Business and Finance, Vol. 31, 2014, pp. 46–

56,

https://doi.org/10.1016/j.ribaf.2013.11.007.

[16] B. D. Simo-Kengne, and S. Bitterhout,

Corruption’s effect on BRICS countries’

economic growth: a panel data analysis,

Journal of Economics, Finance and

Administrative Science, Vol. 28, No. 56, 2023,

pp. 257-272, 2077–1886,

https://doi.org/10.1108/JEFAS-04-2021-0041.

[17] I. G. Radulescu, M. Panait, and C.Voica,

BRICS countries challenge to the world

economy new trends, Procedia Economics

and Finance, Vol. 8, 2014, pp. 605–613,

https://doi.org/10.1016/S2212-

5671(14)00135-X.

[18]

S. Gyedu, T. Heng, A. H. Ntarmah, Y.

He, and E. Frimppong, The impact of

innovation on economic growth among

G7 and BRICS countries: A GMM style

panel vector autoregressive approach,

Technological Forecasting & Social

Change, Vol. 173, 121169, 2021,

https://doi.org/10.1016/j.techfore.2021.12116

[19] T. T. L. Chong, S. H. S. Cheng, and E. N. Y.

Wong, A comparison of stock market

efficiency of the BRIC countries, Technology

and Investment, Vol. 1, 2010, pp. 235–238,

https://www.scirp.org/html/2-

900045_3209.htm.

[20] P. Panda, Stock Markets, Macroeconomics

and financial structure of BRICS countries

and USA, Prajnan, Vol. 49, No. 2, 2020, A.

21.

[21] W. Chkili, Dynamic correlations and hedging

effectiveness between gold and stock markets:

Evidence for BRICS countries, Research in

International Business and Finance, Vol. 38,

2016, pp. 22–34,

http://dx.doi.org/10.1016/j.ribaf.2016.03.005.

[22] K. Kalu, and H. Wang, Financial market

structure and finance-growth relationships in

the BRICS, International Journal of Business

and Economics, Vol. 4, No.1, pp. 58–70,

WSEAS TRANSACTIONS on BUSINESS and ECONOMICS

DOI: 10.37394/23207.2024.21.122

Nursel Selver Ruzgar

E-ISSN: 2224-2899

1508

Volume 21, 2024

2019,

https://doi.org/10.1371/journal.pone.0240482.

t001.

[23] M. A. Nawaz, U. Seshadri, P. Kumar, R.

Aqdas, A. K. Patwary, and M. Riaz, Nexus

between green finance and climate change

mitigation in N-11 and BRICS countries:

empirical estimation through difference in

differences (DID) approach, Environmental

Science and Pollution Research, Vol. 28,

2021, pp. 6504–6519,

https://doi.org/10.1007/s11356-020-10920-y.

[24] E. Zerbo, and O. Darné, On the stationarity of

CO2 emissions in OECD and BRICS

countries: A sequential testing approach,

Energy Economics, Vol. 83, 2019, pp. 319–

332,

https://doi.org/10.1016/j.eneco.2019.07.013.

[25] N. Apergis, M. Pinar, and E. Unlu, How do

foreign direct investment flows affect carbon

emissions in BRICS countries? Revisiting the

pollution haven hypothesis using bilateral FDI

flows from OECD to BRICS countries,

Environmental Science and Pollution

Research, Vol. 30, 2023, pp. 14680-14692,

https://doi.org/10.1007/s11356-022-23185-4.

[26] H. Rasool, S. Maqbool, and Md. Tarique, The

relationship between tourism and economic

growth among BRICS countries: a panel

cointegration analysis, Future Business

Journal, Vol. 7, No. 1, 2021,

https://doi.org/10.1186/s43093-020-00048-3.

[27] M. Balcilar, D. Roubaud, O. Usman, and M.

E. Wohar, Moving out of the linear rut: A

period-specific and regime-dependent

exchange rate and oil price pass-through in

the BRICS countries, Energy Economics, Vol.

98, 2021, pp. 105249,

https://doi.org/10.1016/j.eneco.2021.105249.

[28] D. A. Mohamed, F. Mahat, B. A. A. Noordin,

and N. H. A. Razak, Dynamic connectedness

between Bitcoin and equity market

information across BRICS countries:

Evidence from TVP-VAR connectedness

approach, International Journal of

Managerial Finance, Vol. 16, No. 3, 2020,

pp. 357–371, 1743-9132,

https://doi.org/10.1108/IJMF-03-2019-0117.

[29] S. Nahla, and A. M. Kutan, Private credit

spillovers and economic growth: Evidence

from BRICS countries, Journal of

International Financial Markets, Institutions

and Money, Vol. 44, 2016, pp. 56-84,

http://dx.doi.org/10.1016/j.intfin.2016.04.010.

[30] A. K. Gaba, and N. Gaba, Entrepreneurial

activity and economic growth of BRICS

countries: Retrospect and prospects, The

Journal of Entrepreneurship, Vol. 31, No. 2,

2022, pp. 402–424,

https://doi.org/10.1177/09713557221097160.

[31] B. B. Lumengo, Uncovering equity market

contagion among BRICS countries: An

application of the multivariate GARCH

model, The Quarterly Review of Economics

and Finance, Vol. 67, 2018, pp. 36-44,

https://doi.org/10.1016/j.qref.2017.04.009.

[32] R. Chandrashekar, and K. R. Chittedi, Is there

an export or import led growth in emerging

countries? A case of BRICS countries,

Journal of Public Affairs, Vol. 20, No. 3,

2020, e2074, https://doi-

org.ezproxy.lib.torontomu.ca/10.

1002/pa.2074. Accessed on March 3, 2024.

[33] R. Ritu, and N. Kumar, On the causal

dynamics between economic growth, trade

openness and gross capital formation:

Evidence from BRICS countries, Global

Business Review, Vol. 20, No. 3, 2019, pp.

795–812,

https://doi.org/10.1177/0972150919837079.

[34] B. M. Ahad, A. Jamal, and M. N. Beg, Trade

integration and export aspiration: Evidence

from India’s trade in goods with BRICS

countries, Organizations and Markets in

Emerging Economies, Vol. 13, No. 2, 2022,

pp. 490–514.

[35] M. B. Mudiangombe, and J. W. M. Mwamba,

Impacts of U.S. stock market crash on South

African top sector indices, volatility, and

market linkages: Evidence of copula-based

BEKK-GARCH models, International

Journal of Financial Studies, Vol. 11, No. 77,

2023, https://doi.org/10.3390/ijfs11020077.

[36] M. L. d. Santos, Brazilian stock-market

efficiency before and after COVID-19: The

roles of fractality and predictability, Global

Finance Journal, Vol. 58, 100887, 2023,

https://doi.org/10.1016/j.gfj.2023.100887.

[37] Z. Zhou, J.Yong, L.Yan, L. Ling, and L.Qing,

Does international oil volatility have

directional predictability for stock returns?

Evidence from BRICS countries based on

cross-quantilogram analysis, Economic

Modelling, Vol. 80, C, 2019, pp. 352–382,

https://doi.org/10.1016/j.econmod.2018.11.02

[38] W. Bello, The BRICS: Competition and crisis

in the global economy, Rosa-Luxemburg,

Stiftung, 2015, [Online].

WSEAS TRANSACTIONS on BUSINESS and ECONOMICS

DOI: 10.37394/23207.2024.21.122

Nursel Selver Ruzgar

E-ISSN: 2224-2899

1509

Volume 21, 2024

https://www.rosalux.de/en/publication/id/4047

/the-brics-competition-and-crisis-in-the-

global-economy (Accessed Date: March 1,

2024).

[39] N. Tsepiso, and S. A. Khumalo, Volatility

spillovers in equity and foreign exchange

markets: Evidence from emerging economies,

Journal of Economic and Financial Sciences,

Vol. 15, No. 1, 2022, A713, pp. 1-15,

https://doi.org/10.4102/jef.v15i1.713.

[40] Wei Fan, Albert Bifet. Mining Big Data:

Current Status, and Forecast to the Future,

SIGKDD Explorations, Vol. 14, No. 2, 2013,

pp. 1-5,

https://doi.org/10.1145/2481244.2481246.

[41] WEKA software, [Online].

https://sourceforge.net/projects/weka/

(Accessed Date: December 13, 2023).

[42] N. I. Hussain, B. Choudhury and S. Rakshit,

A Novel Method for Preserving Privacy in

Big-Data Mining, International Journal of

Computer Applications, Vol. 103, No. 16,

October 2014.

[43] M. L. McHugh, Interrater reliability: the

kappa statistic, Biochemia Medica, Vol. 22,

No. 3, 2012, pp. 276–282.

[44] B. Mile, Evolving efficiency of stock returns

and market conditions: The case from Croatia,

Montenegrin Journal of Economics, Vol. 19,

No. 1, 2023, pp. 107-116. DOI:

10.14254/1800-5845/2023.19-1.9.

[45] C. Laurier, O. Meyers, J. Serra, M. Blech, P.

Herrera, X. Serra, Indexing music by mood:

design and integration of an automatic

content-based annotator, Multimedia Tools

Applications, Vol. 48, 2010, pp. 161–184.

[46] W. W. Cohen, Fast effective rule induction,

in: Proceedings of the Twelfth International

Conference on Machine Learning, Tahoe

City, California, July 9–12: pp. 115–123.

1995, https://doi.org/10.1016/B978-1-55860-

377-6.50023-2.

[47] H. He, Y. Ma, Imbalance learning:

foundations, algorithms, and applications,

Wiley, pp. 13.

[48] X. Fan, L. Hong, Y. Wang, Y. Wan, and D.

Zhang, Models of internationalization of

higher education in developing countries - A

perspective of international research

collaboration in BRICS countries,

Sustainability, Vol. 14, No. 20, 2022, pp.

13659, https://doi.org/10.3390/su142013659.

[49] France 24. 2023. Size, population, GDP: The

BRICS nations in numbers. Issued on:

22/08/2023, [Online].

https://www.france24.com/en/business/20230

822-size-population-gdp-the-brics-nations-in-

numbers (Accessed Date: February 18, 2024).

[50] R. M. Nelson, The economic impact of Russia

sanctions, Focus. Congressional research

service, 2022. Updated December 13, 2022,

[Online].

https://crsreports.congress.gov/product/pdf/IF/

IF12092 (Accessed Date: February 27, 2024).

[51] N. Steytler, and D. Powell. The impact of the

global financial crisis on decentralized

government in South Africa, Dans L'Europe

en Formation, Vol. 358, 2010, pp. 149- 172,

https://doi.org/10.3917/eufor.358.0149.

[52] A.A. Salisu, J. Cuñado, K. Isah, and R. Gupta,

Stock markets and exchange rate behaviour of

the BRICS, Journal of Forecasting, Vol. 40,

2021, pp. 1581-1595,

https://doi.org/10.1002/for.2795.

[53] P. Panda, W. Ahmad, and M. Thiripalraju.

Better to give than to receive: A study of

BRICS countries Stock markets, Journal of

Emerging Market Finance, Vol. 22, No. 2.

2023,

https://doi.org/10.1177/09726527231154100.

Contribution of Individual Authors to the

Creation of a Scientific Article (Ghostwriting

Policy)

Nursel Selver Ruzgar organized, analysed and wrote

the paper.

Sources of Funding for Research Presented in a

Scientific Article or Scientific Article Itself

No funding was received for conducting this study.

Conflict of Interest

The authors have no conflicts of interest to declare.

Creative Commons Attribution License 4.0

(Attribution 4.0 International, CC BY 4.0)

This article is published under the terms of the

Creative Commons Attribution License 4.0

https://creativecommons.org/licenses/by/4.0/deed.en

_US

WSEAS TRANSACTIONS on BUSINESS and ECONOMICS

DOI: 10.37394/23207.2024.21.122

Nursel Selver Ruzgar

E-ISSN: 2224-2899

1510

Volume 21, 2024