CHAPTER ONE

INTRODUCTION
1.1 Background of Study
In recent times, there has been a remarkable acceleration in adopting artificial intelligence
(AI) and Machine Learning (ML) in the economy. As a general-purpose technology, AI and
ML are reported to have, directly and indirectly, a strong impact in many sectors (Taddy,
2019). The financial sector is one of those industries where AI and ML has been more
intensively applied (Bredt, 2019; Biallas and O’Neill, 2020).

Since the early days of the computer, when automated computations could perform only a
few arithmetic operations in a single run of a program, the term “artificial intelligence” has
been coined to describe correct logical operations. With the advent of electronic computing in
the late 1940s, the AI’s potentiality was expected to grow within few decades beyond the
realm of fast calculations to reach the quality of human intelligence but working much faster
than the human mind. However, the high expectations soon fell in disillusion several times as
problems and limits were unexpectedly encountered following various stages of excitement.

During recent years, AI and ML services have reached almost all the main sectors of real life,
from commercial activities to health and security systems, military defense, finance, and
financial markets. However, as reported in the Technology Quarterly section of The
Economist (13th June 2020) entitled “An undersding of AI’s limitations is starting to sink in”.
It contains claims such as “after years of hype, many people feel AI has failed to deliver” and
“the same consultants who predict that AI will have a world-altering impact also report that
real managers in real companies are finding AI hard to implement and that enthusiasm for it
is cooling."

History has demonstrated that AI techniques can become more and more powerful, but they
remain limited and can be troublesome and challenging to deploy. Two sets of problems have
been identified. Machine learning used in contemporary AI is built on improved algorithms,
powerful computers, and big data on which they can learn. Strong obstacles plague all these

1
cornerstones. Firstly, algorithms, primarily based on the neuronal networks, “are not
intelligent in the way that most people understand the term” lacking many cognitive abilities
that biological brains take for granted. The result is that of an artificial “idiot savant” that can
excel at well-bounded tasks”: self-driving cars are already delayed, and language-dealing
systems, like chatbots, show frequent signs of misunderstanding. Secondly, computing power
can be expensive to achieve over an extended time. Thirdly, big data are seldom available,
and those that exist contain “hidden assumptions that can trip the unwary” (Sarvady, 2017).

According to a recent Financial Times’ article (January 25, 2021), private equity owners are
often in a rush to transform analogue enterprises into companies that have been thoroughly
modernized using technology. Thus, in their strategy to add value, they select digitally analog
companies and help them to use technologies such as artificial intelligence and machine
learning. This is not only to replace men with machines but also to discover what services
customers need and how to deliver them better. This rush to adopt artificial intelligence is
correlated with the boom in AI-related stocks, indicating investors' expectations from this
technology. Table 1 shows that the results of 36 Exchange Traded Funds (ETFs) of AI
securities on one database, indicating that the annual returns average 43% to 48%, depending
on whether the average is simple or weighted. These returns are far higher than those of any
market index. Assets' total value is 278 billion dollars, indicating that there is a sizable
attraction for this asset class. These results endorse the importance of research on AI in
general and the use of AI in finance and financial markets, in particular. If AI stocks are doing
so well, how is AI being used by financial managers and investment funds? Is the observed
performance of the traded AI Exchange Traded Funds rooted on a solid basis of expectations
about the promised future developments? These questions motivate this survey of the
literature on knowledge in a specific application field of AI and ML in Financial Markets.
This struggle for AI stocks reflects the increasing use of and great expectations from AI and
ML.
The use of technologies for services such as mobile payments has ushered in growth for the
economy and increased the poorest's financial inclusion and income (Lee et al., 2021;

2
Siddiqui and Siddiqui, 2020). Within these technologies, AI and ML offers continuous
innovation for the improvement and potential replacement of human tasks and activities with
a wide range of applications in finance, healthcare, manufacturing, retail, supply chain,
logistics, and utilities (Bredt, 2019; Dwivedi et al., 2019; Ben-Israel et al., 2020; Safdar et al.,
2020).

AI and ML are rapidly transforming the financial services industry. These technologies are
being utilized more and more to improve processes, boost efficiency, reduce costs, and
provide a superior customer experience. Recently, AI and ML have been widely applied in
various financial services, such as fraud detection, credit scoring, risk management, and
investment analysis. With the exponential growth of financial data, the value of AI and ML in
analyzing and interpreting this data is escalating rapidly. In this research, there will be
enumeration of the different use cases of AI and Machine learning in Financial services and
their implications while shining some light on the threats that these tools poses. Also, the
study shall discuss the different implications of these technologies in a micro and macro
financial scale in the Financial Markets.

1.2 Statement of the Problem

The quest to eliminate drudgery and inefficiency in the financial markets system has been on
an increasing demand. This has resulted in the adoption of artificial intelligence (AI) and
machine learning (ML) tools in the financial sector to help curb these human inabilities in the
system. Hence, the study provides a nontechnical background on the evolution and
capabilities of AI/ML systems, their deployment and use cases in the financial sector, and the
new challenges they present to financial sector policymakers.

1.3 Aim and Objectives of the Study

The aim of this research work is to investigate the use of AI and ML in Financial Markets.
The Objectives of the Study includes the following:
1. To investigate the evolution and capabilities of AI/ML systems in the financial markets.

3
2. To investigate the various finance functions in which AI/ML are being implemented.
3. To investigate the new challenges AI/ML techniques present to financial sector
policymakers.
4. To investigate how the implementation of AI/ML in the financial markets is benefitting
finance professionals, employees and business organizations.
5. To investigate the level AI/ML have been enforced in business organizations.

1.4 Significance of the Study

The study provides a comprehensive work on the various ways some AI/ML techniques are
deployed in the financial markets, their efficiency level to eliminate econometric challenges
and their potential to reveal standard econometric toolsets.

1.5 Scope of the study

The study seeks to carry out an investigation on the use and adoption of neueral network
(NN) and fuzzy logic (FL) as AI/ML techniques in the Financial sector, their applications and
exploration as regards stock market prediction, sales forecasting and market segmentation.

1.6 Research Questions

The research questions for the study include the following:
1. What is the evolution and capabilities of AI/ML systems in the financial markets?
2. What are the various finance functions in which AI/ML are being implemented?
3. What are the new challenges AI/ML techniques present to financial sector policymakers?
4. How has the implementation of AI/ML in the financial markets benefited finance
professionals, employees and business organizations?
5. What is the level AI/ML have been enforced in business organizations?

1.7 Expected Research Contribution

AI-powered automation and Machine Learning Techniques streamlines and optimizes
numerous financial processes, boosting operational efficiency. Tasks such as data entry,
reconciliation, fraud detection, and compliance monitoring can be automated, reducing

4
human errors, which are caused by extreme fluctuations, non-linearity and shifts in internal
and external environmental variables, improving scalability, and saving time.
The study gears towards exploring the various AI and Machine Learning Techniques that are
employed in the financial markets; their operational roles and areas of specializations, and
more importantly their effects in financial services in the financial markets.

1.9 Proposal Organization

Table 1: Proposal Organization
Title page
Background
Problem statement
Research objectives
Significance of study
Research questions
Research contribution
Literature review
Research methodology
Conclusion
Reference

1.10 Definitions of Terms

1.10.1 Artificial Intelligence (AI)
Artificial intelligence (AI) is the simulation of human intelligence processes by machines,
especially computer systems or computer-controlled robot to perform tasks commonly
associated with intelligent beings.

1.10.2 Machine Learning (ML)

Machine Learning is a branch of AI that deals with the use and development of computer
systems that are able to learn and adapt without following explicit instructions, by using

5
algorithms and statistical models to analyse and draw inferences from patterns in data.

1.10.3 Financial Markets

Financial markets refer broadly to any marketplace where the trading of securities occurs,
including the stock market, bond market, forex market, and derivatives market, among others.

1.10.4 Fuzzy Logic (FL)

Fuzzy logic is an approach to computing based on "degrees of truth" rather than the usual
"true or false" (1 or 0) Boolean logic on which the modern computer is based.

1.10.5 Neural Network (NN)

Artificial neural networks are a branch of machine learning models that are built using
principles of neuronal organization discovered by connectionism in the biological neural
networks constituting animal brains. It is a series of algorithms that seek to identify
relationships in a data set via a process that mimics how the human brain works.

6
 CHAPTER TWO
LITERATURE REVIEW
2.0 Introduction
AI is a wide term that relates to advancements that make machines "intelligent." John
McCarthy coined the term AI in 1956. There are numerous different terms related to AI, for
example, deep learning, ML, image recognition, natural language processing (NLP),
cognitive computing, cognitive augmentation, machine augmented intelligence, and
augmented intelligence. AI, as used here, includes all of these ideas.

Apart from developing and transforming the fields of aviation, healthcare, transportation,
education, medical diagnosis, electronic trading, remote sensing, robotics and many other
industries, AI is increasingly used by the financial services industry in data mining, market
analysis, personal financing, wealth and asset management, insurance, customer service,
credit scoring, retail lending, process automation and many more spheres to enhance
customer experience by providing him imaginative services with the assistance of the state of
the art technology. The banking industry is fast shedding its stodgy and dogmatic image in a
paradigm shift to a whole-hearted embrace of technology and the banks are accelerating their
innovation agenda to deliver smartly in future through novel solutions that will help
strengthen relationships with customers by offering customized, seamless and imaginative
banking in the digital age.

The evolution in Artificial intelligence (AI) and Machine Learning (ML) have come a very
long way over the years, and most of the large business organizations have started to
assimilate AI and ML in their everyday business hustle. Some of the core fields where AI is
being used comprehensively is finance, healthcare, manufacturing sector, Human resources
(HR), accounting and law. According to the recent report by PWC, 54 percent of the
executives say AI that has been executed in their businesses have already increased their
productivity significantly in comparison to when they were not using AI based technologies
(AI Predictions, 2018). Another, report by Forbes said that 95 percent of business executives

7
who states they are skilled in using big data technologies, to solve business related issues also
tend to use AI technologies (Press, 2020). Based on the above statistics it is pretty evident
that business organizations are emphatically taking into account AI for their business
activities. Based on preceding articles and reports some of the core sectors where AI is
extensively being implemented are illustrated below.

2.1 AI and ML Capabilities

2.1.1 Forecasting.
Machine learning algorithms are used for forecasting and benefit from using large data sets.
They usually perform better than traditional statistical or econometric models. In the financial
sector, this is used in such areas as credit risk scoring, economic and financial variables
forecasting, risk management, and so on.

2.1.2 Natural language processing.

Artificial intelligence systems can communicate by understanding and generating human
language. Boosted by deep learning and statistical models, natural language processing has
been used in the financial sector in such applications as chat bots, contract reviewing, and
report generation.

2.1.3 Image recognition.

Facial and signature recognition is being used by some financial institutions and financial
technology companies to assist with carrying out certain anti-money laundering/combating
the financing of terrorism (AML/CFT) requirements (for example, the identification and
verification of customers for customer due diligence process), and for strengthening systems
security.
2.1.4 Anomaly detection.
Classification algorithms can be applied to detect rare items, outliers, or anomalous data. In
the financial sector, insider trading, credit card and insurance fraud detection, and AML/CFT
are some of the applications that leverage this capability (Chandola, et al 2009).

8
2.2 Financial Markets
Early research indicated that NNs could be used to predict characteristics of high performing
stocks based on historical technical and fundamental data (Quah and Srinivasan, 1999).
Investors in financial markets want to maximize returns based on acceptable risk. Of course,
the mean-variance problem can consider only downside risk by taking semi-variance, but all
these calculations are complex. A rule-based expert system that considers the investor’s
attitude to risk and returns can usually provide better returns than traditional mean-variance
analysis (Rödder et al., 2010).
One study shows that Evolutionary recurrent NN with a linear output prediction model can be
used to profitably trade in the currency market, taking investment risk into account
(Maknickienė and Maknickas, 2013). Different investment profiles can be created, and the
portfolio can be devised for each risk profile. Another study constructed a Least Square
Support Vector Machine model to predict Stock movement. It found that this technique
performed better in predicting stock movements than Probabilistic NN (PNN) and
Discriminant Analysis in both training accuracy and predicting accuracy for the Chinese
stock market (Wang and Shang, 2014). NNs may not always outperform standard linear
models: it depends on the kind of NN model and the kind of market (Li and Mei, 2020).
One problem in forecasting is the noise created by too much information or the absence of
information. To treat this, we see that prediction error can be considerably reduced by the
development of a three-step hybrid intelligent prediction model. First, ensemble empirical
mode decomposition is applied to the data to improve model fitting to them. Second, NN and
support vector regressions are proposed individually for modelling the extracted features.
Third, a genetic algorithm to optimize and determine the weight is proposed for establishing a
unified prediction. It is found that the proposed model improves forecasts considerably,
compared to individual one-stage models, for the FTSE100 and Nikkei 225 indices (Alhnaity
and Abbod, 2020).
For fundamental analysis, investors consider many operating indicators and financial ratios,
often with a wide variety of data. Analysts need to be efficient in analyzing the data and come
to high-speed decisions for portfolio allocation. One study found that portfolio allocation

9
based on expert systems that rank the firms based on fundamental rule-based criteria before
applying an optimization procedure (maximizing returns or minimizing risk) provided better
than industry performance (Amin Naseri et al., 2020).
A comparison of case-based reasoning and NNs to assess financial markets' health, as
measured by the daily financial condition indicator, showed that case-based reasoning works
better than NNs (Oh and Kim, 2007). The case-based reasoning can be updated with a small
amount of data.
Most algorithm trading techniques are based on technical analysis and fundamental analysis,
but sentiments also drive very short-term movements. There are experiments with models
based on investor’s moods based on season, weather, moon, gender, and even sports results
that can influence markets. AI has difficulty in capturing emotions, but some systems are
finding ways to infer them. One such model uses natural language processing to analyze big
data, such as Twitter, press, television, and radio, for sentiments and incorporates the
sentiment in supervised learning algorithms. The algorithm classifies the news as positive,
negative or neutral, and then this classification is used to train the Bayesian network that
predicts whether the Spanish stock exchange index (IBEX35) would go up or down. Based
on this, algorithmic trading is done.
Researchers found that this algorithmic trading based on investor’s moods can predict the
IBEX 35 with an accuracy of more than 50 per cent while the returns from trading are higher
than market returns (Gómez et al., 2019).

2.2.1 Banking
In the banking sector, AI has become a weapon of choice in banks’ competitive struggle to
mine and feed increasing data volume and redefining operations (Padmaavathy, 2018). It is
also applied in investment banking and related services such as stock predictions and credit
rating without customer contact whereas AI’s presence in commercial banking is still limited
in the core operations that are typically concentrated on lending, processing of payments, and
management of deposits (Königstorfer et al., 2020). These authors contend that AI could be
beneficial by reducing lending losses, increasing security in processing payments, automating

10
compliance-related work, and improving customer targeting, managing deposits, and related
services. In this journal issue, Mor & Gupta (2021) demonstrate how the adoption of AI in
commercial banks can reduce inefficiency.
Kerkez (2020) considers that Artificial Intelligence can change the use of labour in banks.
She finds that AI used together with data analytics can lead to efficiencies by making routine
functions such as compliance reporting more efficient, leading to cost savings on labour,
which represents 57 per cent of the costs of banking. She finds that AI can reduce risk
through pattern detection, enabling people to focus on the essentials, but there may still be
new risks that past data cannot capture. Moreover, as the number of new regulations
increases, the time spent following all the alerts that the system generates, which are mainly
false positives, and managers need to continuously update the system to reduce the number of
matches that require manual review.

2.3 AI and ML in Finance and Financial Markets

Artificial Intelligence and Machine Learning (ML) in banking and financial markets are
significantly transforming the way we as human beings associate with money. AI is
extensively helping the finance sector to consolidate and optimize various processes ranging
from quantitative trading, algorithmic trading, risk management process and also financial
advisory services.
According to Forbes business magazine, 70 percent of finance companies around the globe
have already started implementing AI and 60 percent of the companies are using Natural
language processing (NLP).
Some of the projecting applications of AI in financial services industry can be counted as
under.

2.3.1 Specialized services to HNI’s (High Net Worth Individuals) wealth and portfolio
management
Major task of wealth and portfolio management services is to understand the risk return trade
off and to be able to advise which securities and assets will yield the highest returns. AI has
provided a cutting edge to the financial services companies by assisting them in giving

11
customized and accurate advice to their wealthy clients. The world’s largest investment group
– BlackRock with more than $6 trillion assets under management (AUM) has a dedicated AI
lab to assist its operations. Several other global organizations are in the process of embracing
AI to add value to their clients by improving their forecast. Swiss Bank UBS has recently
revamped its trading floor by introducing two new AI systems. One identifies trading patterns
after analyzing reams of market data and then formulates and advises trading strategies to the
bank’s clients for higher returns. The second one addresses the post trade allocation
preferences of its clients.
2.3.2 Automated customer support and virtual financial assistance through chatbots
and robo advisors
Banks are using AI assistant and other relevant apps like Revolut’s which provide instant
services to the customers using Smart chat technologies such as NLP (Natural Language
Processing) or they direct the customers’ enquiry to the relevant support staff. In 2016, Royal
Bank of Scotland installed AI assistant Luvo which responds to customers’ queries in general,
and hand them over to human staff when necessary. Such Robo advisors are redefining
customer experience and delight. Four leading commercial banks in India are using AI in the
form of Chatbots in collaboration with FinTech startups to improve the customer experience,
improve efficiency and reduce cost. One step further, in some cases, banks are using AI
powered smart cameras capable of capturing facial expressions of customers to provide
instant feedback on their experience (AI Applications in the top 4 Indian Banks, 2017).

2.3.3 Enhanced insurance experience

AI has a wide range of applications in the data driven insurance industry. Both in
underwriting a policy and at the time of claim settlement, insurance companies need to know
as much as possible about the education, health, lifestyle, character, etc. of the customer and
the circumstances of the event against which claim is lodged, respectively, which can better
be captured using AI algorithms. Using AI apps, as done by some US startups, an insurance
claim can be settled in less than 3 seconds by performing multiple back end processes and
checks simultaneously while interacting with the client at the front end. AI is all set to play a

12
disruptive role in the insurance industry.

2.3.4 Robotic process automation (RPA) - repetitive task automation

Various banking functions like withdrawal and deposit processing, statement generating,
cheque clearing, billing, etc. are repetitive and having monotonous front and middle office
processing. They can better be executed by AI software like RPA for cost saving, improved
efficiency and better time management. From industrial robots to self-driving cars, robotics
technology is increasingly replicating human intelligence and skills and may turn out to be a
game changer in the financial services industry. Investment into robotic sector is rising fast.
According to CB Insights, “the number of funding deals in robotics globally nearly doubled
from around $273 million in 2014 to $587 million in 2015. The investment growth in 2015
was 115%, as compared to 55% in 2014” (Anupan, 2019).

Fig. 1. Robotics (ex-drones): Yearly Global Financing History (How Robots are
changing the face of banking, 2018)

2.3.5 Credit scoring and predictive analysis through alternate data

There are millions of financially excluded individuals and SME’s who do not have access to
bank credit on account of little or no credit history. Banks find it challenging to lend to such
clients for want of adequate credit history. New FinTech startups are using AI to collect and
process alternate data like location, employment history, age, spending habits, educational
background, police record, social media, and other digital footprint to make lending decisions
in such cases. Predictive analysis, a derivative of AI, can help calculate credit score, prevent
bad loans and afford an insight in to what the customer’s credit requirement is and what he is

13
willing to buy
next (Ludwig, 2018). There is a host of FinTech companies using AI powered algorithms
disrupting the lending industry with AI solutions targeting such unbanked population in
emerging markets.

2.3.6 Regulatory compliance, prevention of money laundering and fraud prevention

& detection
Financial services firms and banks are under immense pressure of regulatory compliance and
risk management requirements particularly after the financial crisis of 2009. Basel Accords I,
II and III provide cumbersome capital adequacy compliance and risk management framework
including KYC and AML processes which are essential to protect the system against credit,
market and operational risks and fraudulent practices. The process is long and arduous
involving piles of paperwork and countless man hours. The real appeal of AI is in its ability
to sift through reams of data and identify trends and patterns in seconds. JP Morgan’s COIN,
which is AI based machine learning system can complete millions of hours of compliance in
minutes. The AI based anti-fraud products spot subtle variances from the predicted human
behavior and indicates suspicion. Real-time camera images and advanced AI techniques such
as deep learning can be used for image and face recognition at ATMs to detect and prevent
frauds & crimes (International Banker, 2018).

2.4 Artificial Intelligence Techniques

2.4.1. Artificial Neural Networks (ANN)
In the aspect of computing, machine learning and artificial intelligence, the developments
have seen a transition from developing powerful computing powers, programming closed
systems devoid of learning capability to attempts to model the human brain as far as its
learning and functioning is concerned (Roohi, 2013). Artificial neural networks are
computational structures which are based on simulation of the biological central nervous
system. The network is composed of a large number of highly interconnected processing
elements (neurons), working in parallel to solve a specific problem. The main advantages of
neural network sit suitability to handle incomplete, missing or noisy data, being a non-

14
parametric method and its capability to map any complex non-linearity and/or approximate
any continuous function (Vellido, et al, 1999). In the study, two types of ANN have been used
- Back Propagation Algorithm and Self Organizing Maps.

2.4.2. Fuzzy Logic

Fuzzy logic provides a method of modelling imprecise models of reasoning, such as
common-sense reasoning, for uncertain and complex processes. Fuzzy set theory resembles
human reasoning in its use of approximate information and uncertainty to generate decisions.
The main contribution of fuzzy control theory is its ability to handle many practical problems
that cannot be adequately handled by conventional control methods (Eksin and Erol, 2000).
Fuzzy models are premised on rule-based system identification, where if–then rules with
imprecise predicates are used to represent the relationships between variables. Fuzzy sets are
then used to define the variables more precisely and make the model operational. Fuzzy sets
are defined through their membership functions (denoted by μ) which map the elements of
the considered universe to the unit interval (Kolarik and Rudorfer, 1997).

2.4.3 Algorithms
Algorithms are a set of rules and steps used in mathematics or computer science to classify or
solve a problem. Therefore, starting from an input, they lead to an output, as shown in Figure
2. Typically, a problem is defined, a set of rules are formulated to solve it, and the solution is
tested over a wide range of examples. There are thousands of algorithms and different ways
of classifying them. Even within the research on finance and financial markets, there are
many kinds of algorithms being used. One example of such an algorithm is Classification and
Regression Tree (CART) analysis, an induction model that discovers rules from incomplete
data. The rules follow an “If…, then…” logic (Kim et al., 2010).

15
Figure 2: Depiction of a simplified algorithm (Kim et al, 2010)

A widely used machine learning algorithm is the Back-Propagation algorithm that uses
gradient techniques (Siddiqui and Abdullah, 2014). The algorithm starts from the last layer
and works backwards to the initial layers, and permits calculating the gradient based on the
network's weight rather than individual weights.
Algorithmic trading uses agent-based models that integrate genetic algorithms to process
information and to evolve the solution proposed because, with time, a fixed strategy stops
working (Yang et al., 2016).

16
2.5 Review of Related Studies
Arash, B. (2010) gives a comparative survey of artificial intelligence applications in finance:
artificial neural networks, expert system and hybrid intelligent systems. Results show
that the accuracy of such artificial smart methods in dealing with financial problems,
particularly nonlinear trends, is greater than traditional statistical methods.

Bjoern, K. et al. (2010) give a survey on Financial Time Series Forecasting with Machine
Learning Techniques. The outcome shows that there is agreement among researchers that
stock index forecasting is significant. The prevailing machine learning technology in this
field is known as Artificial Neural Network (ANN). They believe that stock index forecast is
significant. The core technology for computer education is named ANN.

Meryem, D. F. et al. (2009) survey on Assessing Bank Performance with Operational

Research and Artificial Intelligence Techniques. The research shows that for DEA research
banking and most experiments using a two-stage DEA, feasibility and power productivity
have been considerably reduced and their findings may be bias. Their viability and power
efficiency have significantly reduced and their results may be skewed in the case of DEA
study banking and in most DEA studies.

Tom, L. (2019) highlighted the study Artificial Intelligence, Finance, and the Law. The study
shows that The rise and growth in financial and other artificial intelligence would definitely
be one of the most important legal, business and social innovations of the years to come. The
early developments provided insights into the immense capabilities of financial artificial
intelligence and the opportunity.

Bonnie, G. B. (2019) carried out a research in Artificial Intelligence in Finance. The

research reveals that The financial services industry is still in the early stages of artificial
intelligence. Throughout banking, AI has become more omnipresent and there are more
problems, including political, cultural, economic and social hurdles. He remarked that there

17
is still artificial intelligence in the early stages of financial services industry. AI is more
omnipresent in banking and challenges, including financial, technological, economic and
social hurdles, are there.

Chi, C. et al. (2019) conducted research on Artificial intelligence applications in financial

services. The study shows that the use of AI in financial services has many advantages. It
can increase efficiency and productivity by automating; minimize psychological or
emotional errors; and improve the quality and conciseness of management knowledge by
identifying patterns and/or longer-term developments, which cannot be easily identified using
traditional monitoring techniques. The study also emphasizes that there are many drawbacks
to the use of AI in financial services. It can improve efficiency and output by
dynamically detecting anomalies and/or long-term changes that cannot be easily identified
through conventional tracking methods, mitigating psychological or emotional mistakes. It
can also improve quality and conciseness of management information.

Shorouq, F. E. et al (2010) on Neuro-Based Artificial Intelligence Model for Loan Decisions,

experimented that Artificial neural networks are a good tool to use in the Jordanian
commercial Banks loan application evaluation. Their research also reveals that Artificial
neural networks are a good tool for evaluating loan applications in Jordan commercial
banks.

Cüneyt, D. (2015) did a survey on The Impacts of Robotics, Artificial Intelligence on

Business and Economics. Developments in internet and mobile devices are now driving
mechatronic research, including advancements in computing, nanotechnology, advances in
medicine, safety and wireless apps. Primitive mechanization of production processes took
place in the industrial age that mankind joined with steam series long before.

Philippe, M. and Olivier, B. (2010) carried out a Generic Architecture for Realistic
Simulations of Complex Financial Dynamics. The research highlighted recent financial
crisis highlighted the need for new research tools capable of addressing the economic

18
world's high degree of complexity. The recent financial crisis demonstrated the need for
new tools to address the high degree of complexity in the field of the economy.

Alexandra, Z. (2017) carried out a comprehensive study on Artificial Intelligence in Finance:

Forecasting Stock Market Returns using Artificial Neural Network. He found out that The
best approach for minimizing prediction errors was identified for ARMA models, while
networks are often more accurate for predicting paths or symbols.

Ludwig, E. (2018) highlights the growing application of artificial intelligence (AI) in banking
and financial services industry. He describes how AI assists in analyzing customer data for
the purposes of credit scoring and arriving at loanable amounts. He also prescribes for
updating regulations to prevent misuse of AI.

Nunn, R. (2018) addresses the issue of potential bias in AI and advises that financial
institutions would better balance AI applications with their algorithmic bias against minority
population. He claims that increased diversity at workplaces can reduce the biases inherent in
AI applications.

Satell, G. (2016) broadly discusses ethical issues in artificial intelligence systems and
emphasizes on the need for designing an ethical learning environment, identifying bias in
algorithms, investigating ethical dilemmas, and setting higher moral standards.

Daks, M. (2018) focuses on the AI benefits for banks using the case study of the acquisition
of artificial intelligence company Layer 6 Inc. by TD Bank Group. He cites examples of
financial institutions using artificial intelligence model of third-party providers.

Guy A. M. (2017) discusses the integration of artificial intelligence in the digital ecosystems
of the financial services industry and examines the role of AI in business development and
providing customized advice and services to consumers.

19
Meinert, M. C. (2018) discusses the role of (AI) in combating cybercrime in the financial
services industry and in enhancing customer experience and also the likelihood of AI’s
potential involvement in increasing cyber-attacks.

Forrester Research Inc (White Paper) (2017), E-Book, Artificial Intelligence with the Human
Touch, contends that Artificial Intelligence can handle repetitive mundane but
cannot substitute human touch. The study explores how a blending of AI and human touch
can enhance customer experience.

Table 2: Summary of Review of Related Studies

SN Research Title Author(s) Research Outcome Research Limitation Year

1 A comparative Arash Results show that theThe results show that 2010
survey of artificial Bahrammirz accuracy of suchin solving financial
intelligence a ee artificial smartproblems, especially
applications in methods in dealingnon-linear patterns, the
finance: artificial with precision of such
neural networks, financial problems,artificial smart
expert system and particularly methods are greater
hybrid intelligent nonlinear trends, is than conventional
systems greater than statistical methods.
traditional
statistical methods.
2 Financial Time Bjoern There is agreement Researchers believe 2010
Series Forecasting Krollner, among researchers that stock index
with Machine Bruce that stock index forecast are
Learning Vanstone, forecasting is significant. The core
Techniques: A Gavin significant. The technology for
Survey Finnie prevailing machine computer education is
learning technology named ANN.

20
 in this field is known
as Artificial Neural
Network (ANN).
3 Artificial Forrester The study exploresArtificial Intelligence 2017
Intelligence with Research Inc how a blending of AIcan handle repetitive
Human Touch (White and human touch canmundane but cannot
Paper) enhance customersubstitute human touch
experience

4 Assessing Bank Meryem For DEA research Their viability and 2009
Performance with Duygun banking and most power efficiency have
Operational Fethi, experiments using a significantly reduced
Research and Fotios two-stage DEA, and their results may
Artificial Pasiouras feasibility and be skewed in the case
Intelligence power productivity of DEA study
Techniques: A have been banking and in most
Survey considerably DEA studies.
reduced and their
findings may
be bias.
5 Application of Ludwig, E. AI assists in Prescription for 2018
artificial analyzing customer
updating regulations to
intelligence (AI) in data for the
banking and purposes of credit prevent misuse of AI.
financial services scoring and arriving
industry at loanable
amounts.
6 Using a fuzzy G.T.S. Ho, The case study The case study 2012
association rule W.H. Ip, indicates that the reveals that the new
mining approach C.H. Wu, approach suggested solution helps
to Y.K. Tse is a realistic way of creditors logically in
identify the supporting investors actions without
financial data in decisions who understanding the
association cannot recognize underlying values of
the secret principles the Hang Seng index
between the Hang and the economic
Seng Index and indices.
other economic
indices.
7 Workforce Nunn, R. Financial Increased diversity 2018
Diversity Can institutions would at workplaces can
Help Banks better balance AI reduce the biases
Mitigate applications with inherent in AI

21
 AI Bias their algorithmic applications.
bias against
minority
population.
8 Teaching an Satell, G. Ethical issues in Ethical issues in 2018
Algorithm to artificial artificial intelligence
Understand Right intelligence systems systems.
and and the need for
Wrong. designing an ethical
learning
environment,
identifying bias in
algorithms,
investigating ethical
dilemmas, and
setting higher moral
standards.
9 Artificial Tom C.W. The rise and growth The rise and 2019
Intelligence, Lin in financial and development in
Finance, and the other artificial financial and other
Law intelligence would artificial intelligence
definitely be one of will undoubtedly be
the most important one of the biggest
legal, business and legal, economic and
social innovations social innovations of
of the years to the years to come.
come. The early The early advances
developments gave insight into the
provided insights enormous potential
into the immense and ability of
capabilities of financial artificial
financial artificial intelligence.
intelligence and the
opportunity.
10 Artificial Bonnie G. The financial There is still 2019
Intelligencein Buchanan services artificial intelligence
Finance industry is still in in the early stages
the of financial services
early stages of industry. AI is more
artificial omnipresent in
intelligence. banking and
Throughout challenges, including
banking, financial,

22
 AI has become technological,
more economic and social
omnipresent and hurdles, are there.
there
aremore
problems,
including
political,
cultural, economic
and
social hurdles.
11 Artificial Chi The use of AI in There are many 2019
intelligence financial services drawbacks to the use
applications in Chan, David has of AI in financial
financial services Nayler, many advantages. services. It can
Jayant It improve efficiency
Raman, can increase and output by
Matthew efficiency dynamically
Baker and productivity detecting anomalies
by and/or long-term
automating; changes that can not
minimize be easily identified
psychological or through
emotional errors; conventional
and tracking
improve the quality methods, mitigating
and conciseness of psychological or
management emotional mistakes.
knowledge by It can also improve
identifying patterns quality and
and/or longer-term conciseness of
developments, management
which cannot be information.
easily identified
using traditional
monitoring
techniques.
12 Neuro-Based Shorouq The results show Artificial neural 2010
Artificial Fathi Eletter, that networks are a good
Intelligence Model Saad Ghaleb Artificial neural tool for evaluating
for Loan Decisions Yaseen and networks are a good loan applications in
Ghaleb Awad tool to use in Jordan
Elrefae the commercial banks.

23
 Jordanian
commercial
Banks loan
application
evaluation.
13 The Impacts of Cüneyt In the industrial Primitive 2015
Robotics, Artificial Dirican age, which mechanization of
Intelligence On humanity entered production processes
Business and with steam series took place in the
Economics long ago, primitive industrial age that
production process mankind joined with
mechanization took steam series long
place. before.
Developments in
internet and mobile
devices are now
driving mechatronic
research, including
advancements in
computing,
nanotechnology,
advances in
medicine, safety
and wireless apps.
14 A Generic Philippe The recent The recent financial 2010
Architecture for Mathieu and financial crisis demonstrated
Realistic Olivier crisis highlighted the need for new
Simulations of Brandouy the tools to address the
Complex Financial need for new high degree of
Dynamics research complexity in the
tools capable of field of the
addressing the economy.
economic world's
high
degree of
complexity.
15 Banking on Daks, M. AI benefits for Financial institutions 2018
Technology: banks using the using artificial
Artificial case study of the intelligence model of
intelligence acquisition of third-party
helping banks get artificial providers.
smarter intelligence
company Layer 6

24
 Inc. by TD Bank
Group.
16 Artificial Alexandra ARMA models The best approach 2017
Intelligence in Zavadskaya were found to be for minimizing
Finance: the best way to prediction errors was
Forecasting Stock minimize prediction identified for ARMA
Market Returns errors, whereas models, while
using Artificial networks are often networks are often
Neural Networks more reliable to more accurate for
predict path or predicting paths or
symbol. symbols.
17 Artificial Guy A. M. The integration of The integration of 2017
Intelligence: The artificial artificial intelligence
Ultimate intelligence in the in the digital
Disrupter. digital ecosystems ecosystems of the
of the financial financial services
services industry industry.
and the role of AI in
business
development and
providing
customized advice
and services to
consumers.
18 Artificial Meinert, M. The role of AI in AI in combating 2018
Intelligence: The C. combating cybercrime in the
Next Frontier of cybercrime in the financial services
Cyber Warfare financial services industry
industry and in
enhancing customer
experience and also
the likelihood of
AI’s potential
involvement in
increasing cyber-
attacks

2.6 Research Gap

Although many studies have examined the application of AI and ML, AI techniques and use
in the health sector, government sector, retail sector, banking, etc., there has not been any

25
general, thorough and comprehensive research work to investigate the use of AI and ML in
the financial markets and its social implications the financial markets.

CHAPTER THREE

METHODOLOGY, RESEARCH PROCESS AND PLAN

3.0 Introduction

Models of AI, Knowledge Engineering and Operations research can be categorized as

generic, informal, semi-formal, or formal (Gavrilova et al., 2018). An introductory book to AI
divided AI into five fields: symbolic AI; artificial neural networks; evolutionary
programming; cellular automata; and dynamical systems (Boden, 2018).

A more specific review paper on AI divided the field into three categories: AI and decision-
making (e.g., neural networks, deep learning, algorithms); application domains (e.g., robotics,
healthcare, imaging) and data and information (e.g., big data, data visualization) (Dwivedi et
al., 2019). Nevertheless, since we are dealing with a specific domain of AI in finance &
financial markets, we are treating a sub-set of the literature with its features, some of them
coinciding with the larger AI literature and our division will depend on this

3.1 Research Design

There are three main perspectives of AI and ML in the Financial Markets: those describing
the technique (machine learning, expert system, neuralnetwork); those describing the
application (forecasting, decision-making); and those describing the sector of application
(financial markets, banking). Hence, the study tends to proffer analytical solutions in this
order:
1) AI techniques for finance or financial markets;
2) AI applications for finance and financial markets; and
3) sub-fields of finance and financial markets where AI is being used.

3.2 Population of the Study

To determine the population of study, an investigation was conducted in veritable databases

related to AI and financial or financial markets. In the EBSCO database investigated, 4,777
peer-reviewed references contained the term “AI” in the abstract. 55,780 references were
using the search term ‘finance or "financial markets"’ in the abstract. However, the
intersection of these two fields produced only 73 results. A similar search was done in

26
Science Direct and in Emerald databases and a corpus of 90 results was generated in the
Endnotes program after removing duplicates.

Also, some large banks and finance companies were involved in ascertaining the population.
Altogether, the population of the study consists of EBSCO, Science Direct, Emerald
Databases, large Banks and Finance Companies.

3.3 Sampling Techniques and Sample Size

The sampling technique adopted for this research is the Simple Random Sampling technique.

To determine the sample size, the EBSCO database, Science Direct database and the Emerald
database were investigated. A population of related studies was capped at 264 from all
databases out of which a random sample of 90 selected documents were made and a
questionnaire which included feedback from 100 finance professionals working in large
banks and finance companies.

Hence, the total Sample Size is 190.

3.4 Method of Data Collection

The above research is mainly focused on secondary data as there is no exhaustive primary
research conducted in the field of AI and the finance sector. Very comprehensive research
was orchestrated to identify and pick out articles from various research databases such as
EBSCO Database, Science Direct Database and Emerald Database. Data was also collected
from various grey literature sources such as websites, articles and magazines pertaining to
Artificial Intelligence and Machine Learning.
The study also incorporated an interview method with the help of a questionnaire which was
mainly focused on finance professionals working in large banks and finance companies. This
study also incorporated primary data collection from 100 finance professionals to understand
up to what extent Artificial intelligence (AI) has been implemented in the field of accounting
and Finance.

3.5 Method of Data Analysis

A systemic content/document analysis was used to evaluate related literature publications in

this study. A selection of papers, including posts, has been collected. This research focuses on
broad publications peer-reviewed, including Scopus and SSRN, which are listed in quality
and impact rankings. This selection of the highest-ranking papers not only guaranteed the
quality of papers that were most reviewed and validated but also provided the most up-to-date
research state during their publication periods. Some keywords are used to scan for artificial
intelligence papers, such as artificial intelligence and financial articles such as corporate
finance, artificial intelligence, digital finance, financial and artificial intelligence, etc.

27
3.6 Plan of Work and Time Schedule

ACTIVITY
Planning
Researching
Chaptering One
Chaptering Two
Chaptering
Three
Chaptering Four
Chaptering Five
Follow up
1-2 3 -4 5-6 7-8 9 – 10 11 - 13 - 15 - 17 -
12 14 16
NUMBER OF WEEK

Figure 1. Plan of Work and Time Schedule

3.7 Resources

The resources used are both within and outside traditional libraries. They consist of primary,
secondary and tertiary materials ranging from print media such as Theses, dissertations,
books, scholarly journal articles, article reviews on research works, Wikipedia to electronic
sources found on internet.

28
 CHAPTER FOUR
RESULTS AND DISCUSSION
4.0 Introduction
It's possible that financial services will benefit from AI and ML in the future.
As humanity becomes more and more dependent on digital operations and moves to online
platforms in response to a changing environment, artificial intelligence (AI) and machine
learning (ML) are changing how organizations operate and allocate resources. Using vast
amounts of data and cutting-edge statistical techniques, predictive pattern recognition with
AL and ML is able to provide the best estimate solution
to particular and conclusive problem sets. The analysis is based on information given to a
computer program rather than on intrinsic machine intelligence (Chan et al. 2019). Artificial
intelligence (AI) has enormous potential for good if businesses use it carefully and cautiously.
AI is already a potent tool that is extensively used in the financial services industry. The
study focuses on and discusses a number of particular technologies, including banking
algorithms, chatbots and automated assistants, risk management, and fraud detection.
4.1 Applications of Artificial intelligence in Finance functions
AI is taking over the financial sector by storm, most of the companies in the fin-tech sectors
have started to use Artificial intelligence (AI) in order to reduce costs, save a lot of time and
also add much-needed value to their products For instance, robot-advisors tends to track the
activity of account holders by using AI technology. AI is used to analyze and comprehend
how account holders invest their money, spend money from their account and make financial
decisions so, the companies can personalize the advice they give their clients. Finance
functions such as lending services, stock trading and financial fraud detection are some of the

29
many functions where Artificial intelligence has been successfully implemented.
4.1.1 Lending services
Financial institutions are provided by a number of macro- and microeconomic factors, and
numerous other factors are responsible for bringing these financial institutions closer to risk.
One of the main ways that banks and non-banking financial institutions (NBFC) generate
significant revenue is through lending loans, but loans cannot be approved for just anybody.
Many loan providers rely on different credit models, but ultimately, all of their models
are based on the fundamentals of the loan seeker's payment details and transaction details
from the past. In the early days of credit providing, lending decisions for clients and firms
heavily relied on the credit scores of the customers (Chawla, R., 2018). The older
models typically use statistical analysis, regression, and decision trees to determine the total
credit score from the limited data that is provided to them. Nonetheless, banks are continuing
to use a more thorough approach when giving consumers loans in the modern era. To
improve the accuracy of loan ratings, banks are also considering information from semi-
structured sources, such as text message and mobile phone usage, as well as social
media activity.
The current generation of credit scoring tools uses machine learning (ML) algorithms
to enable the assessment of even qualitative elements like the borrower's willingness
to repay the loan and their behavior. This tendency has given way to a more accurate credit
decision through the quicker and less expensive division of borrower quality. With
the application of AI, nearly 80% of people on the planet will likely have access
to credit. Artificial intelligence is used to provide significant access to credit. In addition to
being limited to producing a precise, segmented assessment of creditworthiness, machine
learning algorithms have the potential to significantly expand and expand credit availability.
For a potential customer to be deemed "scorable" under the majority of traditional credit
scoring models, they needed to have a sufficient amount of past credit history (Chawla, R.,
2018). When a person's credit application was denied because there was insufficient
prior credit scoring data, a credit score could not be created. However, with the advent of AI
technology, lenders are now able to determine a loan applicant's creditworthiness by gauging

30
their willingness and capacity to repay the loan. In this manner, a large number of people
would be able to obtain loans from banks.A shift in the prototype of the credit lending
process has been noticed in most of the banks after the implementation of AI. Over the years
a lot of fintech companies are rising and helping the customers who were not able to get
support from banks under traditional creditlending system.With the implementation of
Artificial intelligence, banks are nowable to investigate large amounts of data at a quick pace.
Thus, resulting in credit policies with the capacity to handle a very broad range of credit. It
also lowered the total cost of risk assessment for individuals and also significantly escalating
the number of people for whom credit risk can be accurately measured.

Introduction of AI will significantly reduce the total risk of lenders. Machine learning (ML)
algorithms can do what human beings usually fail to do. The best example for the above
statement is meticulously identifying rogue investors that are functioning across different
accounts. Machine learning usually does this by utilizing predictive analytics to large
amounts of data. Implementing AI in their struggle to digitalize credit risk processes will help
the banks on the nearer term gains while building key capability for overall transformation
(Chawla, R., 2018). Machine learning technology is also applicable in Early-warning system
(EWS) for instance bringing gaping insights at the desk from very large and devious data sets
without fixing the limits of standardized statistical analysis. With the use of EWS banks and
financial institutions get improvement in portfolio monitoring, automated reporting and
advice for any potential actions. In relation to the SME segment, financial institutions have
managed almost 70 per cent to 90 per cent enhancement in precisely forecasting any late
payments six or more months before late payment.

Artificial Intelligence is also being implemented in product matching and intelligent product
selection. Even though the customer is on board, AI technology can be used before the credit
check process to significantly improve the customer's overall experience (Rehfisch, 2019).
Customers in the present day and age usually tend to expect personalized offers that are
admissible to them. If the client chooses to agree, then the bank of their trust can recommend

31
a pre-selection of appropriate credit products by consolidating intelligent analysis tools. The
same technique can be used within application process usually when the customer of a bank
applies for a credit product 'A' online their transaction data must be checked to verify their
creditworthiness the incorporation of AI at this point will insert another step to the first check
if another credit product may be more suitable for them. In this case, the substitute product 'B'
can be automatically offered. From, the client’s perspective their application has an undue
advantage that they get better conditions and banks will also increase its chances of lending
them money.
AI application can be very useful in the case of structured financing. But sometimes it might
get complex for instance when financing is not concentrated towards the borrower but to an
investment object, then special purpose vehicles (SPV) or special purpose entities (SPE) are
usually set up in the case of financing a large project. Since historical data and the previous
years’ balance sheet play a secondary role in Special purpose vehicles (SPV) the banks’
lending decision, in this case, is based only on expected future cash flows and the
corresponding propensity of service debt. This is where AI comes into action it promises
more dependable predictions and also faster process time. AI technology can process a very
large amount of data in a very short period and at the same time improve their algorithms. For
instance, financing of large and expensive renewable energy projects such as solar panels and
wind farms can be made easy with AI. Through intricate simulation ofweather forecasts and
the effectiveness check ofthese results along with the customer, documentation will make it
easier for banks to make more dependable estimates on their future earnings.
4.1.2 Stock trading
An algorithm is a collection of a set of commands given to an Artificial Intelligence (AI)
program to read and learn on its own or quickly respond and deliver when a certain set of
events happen (Shah, 2020). The recent advancement in AI is providing a very good
opportunity for people that are going to benefit from them. Before the introduction of AI in
trading, buying and selling of stocks and shares was a very cumbersome process but with the
advent of AI, buying and selling of stocks have become a very easy process. The AI in stock
trading tends to use a set of algorithms which is also called algorithmic trading. Algorithmic

32
trading is known as a system of trading which provides decision-making transactions in the
stock markets using sophisticated mathematical techniques. It uses the most advanced
programming techniques to make decisions and transactions relating to buying and selling of
stocks and shares. The algorithms are designed in a way that it would govern the optimum
time to place an order which will cause the least amount of influence on the trade price.
Interference by a human is completely eliminated and the decisions made by the algorithm is
very quick and accurate. The algorithms will help to spot any huge possibilities to track profit
in the market (Shah, 2020). In algo-trading, computer software can decide on your sake to
buy and sell stocks. Algo-trading also tends to use very advanced high- alphanumerical
models that can deliver correct decisions at financial markets. Most of the companies started
to enforce these technologies especially the top investment banks.
Algorithmic trading usually uses three kinds of strategies that are performance-based,
statistical arbitrage and momentum investing. In performance-based strategy it is focused on
execution-based plans. Investors will use this kind of strategy only when generating bulk
purchases. In statistical arbitrage, it will identify cost differentials among devices that are
quoted in different markets which dispense predictable and known connections with each
other. The momentum investment strategy depends on the drive in the business, it helps stock
traders in the anticipation of market leaning that estimates vital action (Shah, 2020). The AI
in stock trading often uses deep learning techniques to recognize different patterns in stock
charts that human beings might not be able to do (Ponduru, 2019). For instance, the common
pattern that is known to stock traders is known as the ascending triangle, If the pattern
indicates the share price will go up human beings will usually not be able to identify this
particular pattern in a quick pace or sometimes they may not notice it at all but AI technology
will recognize this pattern quickly.AI will also use past data to learn how the market
reciprocated in the past events, based on the reactions they can perform in more predictive
ways thus making them more reliable and accurate in the process. In recent days even
National association of securities dealers automated quotation (NASDAQ) is using AI in its
United states stock markets to detect any rough and potential malevolent trading activity. The
newly launched proposal by NASDAQ will help to strengthen and revolutionize market

33
inspection through machine learning and other AI capacity.
4.1.3 Financial fraud detection
The finance sector is on the verge of a huge transformation and the ancillary force behind it is
AI. One such area that is growing tremendously is financial fraud detection, according to
McAfee cybercrime related to financial fraud costs the world around 600 billion US dollars,
that is equal to 0.8 per cent of the worlds Gross domestic product (GDP) (Dyzma, 2019).
Artificial Intelligence is showing to be very productive in conflicting financial frauds.
According, to the recent report by Forbes magazine 80 per cent of the financial fraud
detection specialists believe that AI technology will significantly help to reduce payment and
financial related frauds. AI's capacity to understand trend-based insights from machine
learning (ML) algorithms are reducing the prevalence of payments related to fraud
(Columbus, 2019). Artificial intelligence (AI) also uses predictive analytics and machine
learning procedure to identify any incongruity in very large data sets quickly, the more data
Machine learning (ML) model is fed the more authentic its predictive value. By scrutinizing
historical data from a large data network, Machine learning algorithms can tend to achieve
better predictability and accuracy at the same time. Big banks can also use predictive
analytics-based fraud detection software to identify frauds across different domains that are
involved in financial payment processing (Mejia, 2020). Banks can also use predictive
analytics to detect any fraud in mobile applications for banking or ordering and paying for
goods and services.
4.1.3 AI and ML used for the identification of phenomena
Pattern recognition facilitates the detection of comportments that vary from normal patterns.
For example, AI may be used to detect and send warnings to money laundering, security
threats, irregular financial arrangements, and illegal transactions. It is used to build maximum
investment strategies (Chan, Nayler, Raman, & Baker, 2019). There are now more and more
robot consulting services that automate portfolio management recommendations for
individual investors. AI's other use is in finance algorithmic commerce, systems that integrate
information on changing market conditions and price levels through the use of proprietary
algorithms, allowing automatic trades quite quickly. Trades are often conducted so fast that

34
the word 'high-frequency trading' has been taken up (Kraus & Palmer, 2018).
The use of AI in financial services has many advantages. It can increase efficiency and
productivity through automation; minimize mistakes induced by psychological or emotional
factors; and strengthen management information's accuracy or conciseness by detecting
patterns or longer-term developments that are not easily identified by existing monitoring
methods (Buchanan, 2019). Such requirements are particularly valid where legislation, like
the Financial Instrument Directive II for the European Union Markets (MiFID II), expands
senior management's obligations for analysis and takes greater data from the business into
account (Ho et al., 2012).
In contemporary times, Artificial Intelligence has made several advances that have allowed
applications for financial professionals to be developed, which could, or is likely to, disrupt
the financial sector. It is therefore believed that AI and ML will substitute not only human
capital in whole or in part but also improve performance beyond the human level (Fethi &
Pasiouras, 2009).
4.1.4 AI and ML used in the processing of text, news and semantic syntax.
AI uses automatically "interpreting" and evaluating texts such as papers, articles, social
media and material. It will be extremely important for the future development of financial
systems, because in a few seconds AI computers can absorb all relevant information and
news while people will take plenty of hours to handle all the details which can affect
inventory performance. Data mining facilitates estimation, predictive activity and price level
analysis in market data. Predictions and findings can also be included in legislative and
structural adjustments (Kraus & Palmer, 2018).
The use of AI technologies by Text mining is often viewed as a sub-set in a market sentiment
analysis. Market sentiment has recently grown with the extensive development of social
media platforms and the creation of large data sweatshops (Eletter, Yaseen, & Elrefae, 2010).
The growing pool of Big Data from the web and social networking interactions provides
interesting new ways to study the conduct of market participants. In this regard, the
comprehensive sentiment analysis of social media content has been found to proliferate the
predictability of future outcomes for a range of elections. In line with this, several other

35
studies linked the number of online searches for a specific topic to early economic activity
(Krollner, Vanstone, & Finnie , 2010).
4.2 How artificial intelligence and machine learning is benefitting business
organizations, employees and finance professionals
Artificial Intelligence (AI) plays a very consequential role in the different finance functions
which have favored not only the finance professionals but also organizations and employees
too.
4.2.1 Business organizations
AI is assured to have a huge impact on business organizations, it is no longer just about
codifying business judgment and process automation but AI will be used by most of the
business organizations around the world to gain deeper insights and also gain competitive
advantage simultaneously. According to a recent report by Gartner, Investment in AI-based
technology will be one of the companies’ top five priorities by 2020 (Victories, 2018). A
recent report by Harvard business review also says that Artificial Intelligence will add 13
trillion US dollars to the world economy over the next ten years (Fountaine, et al., 2019)
AI helps business organizations to automate most of its customer interactions. AI is helping
organizations to automate communications such as online chats, Emails, telephone calls by
analyzing data that was accumulated, it is also possible to program the computer software
algorithms in such a way that it will precisely respond to the customers and also deal with
any enquires they have (Victories, 2018). For instance, AI-based chatbots can connect with
unlimited customers at the same time and can both instigate communication and respond
whether it is an application or a website. Approximately by the year 2020, 85 per cent of
customer communications will be taken over by AI functions that can mimic and respond like
human beings.
Business organizations are now able to predict future outcomes based on the process of data
analysis. For instance, predictive analytics will be able to check patterns in customer data that
can show if the products at the moment on sale are expected to sell and what quantity. AI can
also predict when the demand for a particular product is declining. This will help the
company acquire the appropriate stock and in the right quantity. AI can also predict when the

36
demand for a particular product is declining. This will help the company acquire the
appropriate stock and in the right quantity. The propensity is not just useful in the retail
industry, but AI is also being used in many different areas, for instance, in the banking
industry it can forecast stock price fluctuations and predict currency. In the health care
sectors, it can also predict the upsurge of infections by scrutinizing social media posts.
AI will also help business organizations in the process of data unlocking. In the previous
years, the amount of data business was producing is less so the data was unstructured, so it
was easy to apprehend and store the data in large databases. Large business supervisors are
now able to extract deeper insights from the data propagated for their business requirements.
But in the present scenario, unstructured data tends to represent a larger part of the total
information that is available to us (The Power of AI for Business/Accenture, 2020). 80 per
cent of the data that is produced from online websites is unstructured data. Investigating
unstructured data will be the most significant factor for any business organization over the
future years.
4.2.2 Finance professionals
Artificial intelligence (AI) is assured to change the accounting industry with changes that will
abolish monotonous tasks and also free up time to prioritize any higher impact
responsibilities (Vordenbaeumen, 2019). In the recent digital era transformation, clients are
digitalizing most of the processes and multiplying the number of documents and spreadsheets
that accountants and audit professionals must analyze. The introduction of AI in the field of
accounting and auditing will significantly help finance professionals such as auditors and
accountants.
Artificial intelligence in the field of audit and accounting will help finance professionals to
streamline the process of data entry and analysis. AI will help finance professionals to stay on
top of the transactions amidst the system that is monotonous, time-consuming. Instead of
finance data being spread over multiple PDF's, documents and spreadsheets machine learning
technology will automatically classify it based on expenditure category and generates
financial reports for analysis in just one place (Vordenbaeumen, 2019). These compendious
financial reports will provide business organizations with smart perception to enhance the

37
process of financial planning. ML also extracts deeper insights as it churns data over a period
of time, meaning business organizations can gain a complete view into the long-term
expenditure patterns and the accounting department can contribute even greater value to
business organizations by recommending clients on optimistic budget forecasting.
AI technology can be used by finance professionals to extremely reduce the total time it takes
to discover any non-compliance issues in the financial data. Any travel bookings, order for
purchases, receipts from employees and credit-card related transactions are spontaneously
scrutinized for any purchases made outside the company’s policy thus helping audit
professionals to quickly rectify the identified error and assist to execute corporate policies to
the employees with improved precision into corporate spending patterns. Companies can also
identify which policies are functioning for the company as well as if certain policy
infringements are reasonable.
4.2.3 Employees
Nowadays most of the treasury and accounts receivable clerks strive to clear the invoice
payments, when clients usually combine all the invoices in one single payment, pay the
wrong amount or do not contain any invoice numbers with their payments (Zipperle, 2020).
In this, to clear the invoices the employee has to physically add the various invoices that may
match up the payment amount or communicate to the customer to resolve some information.
But with the introduction of AI, it would instantaneously help by proposing invoices that may
match the payment amount. On the basis of experienced thresholds, the AI system will
automatically help to clear any short payments by customers.
Artificial intelligence (AI) can help finance employees to conduct risk assessments. In
evaluating commercial proposition for projects, finance employees are assigned with the task
of estimating each project individually based on various customer attributes such as size,
industry, maturity, current system landscape as well as the intricacy of the products to be
implemented. To make this judgment employees of the organization often rely on finance
managers who have previous experience in dealing with a similar type of project. The
implementation of Artificial intelligence in this scenario will help the teams to access all the
data related to the projects the company has implemented till now. Using this past data, the

38
finance employees can map the planned project against all past projects and come up with a
better decision for risk assessment.

CHAPTER FIVE
CONCLUSION AND RECOMMENDATION
5.1 Conclusion
Artificial intelligence and machine learning in the financial markets is still in early days. AI
and ML will become more ubiquitous in finance, and with that comes more challenges
including legal, ethical, economic and social hurdles. They will also continue to bring new
complexities to the global financial ecosystem. As more and more data become available and
computing power increases, AI programs will become more complex.
Artificial Intelligence (AI) is the application of machines and algorithms to
enhance and simulate human intelligence. AI makes it possible to identify trends that
are adaptive across vast amounts of data and to apply contemporary statistical techniques to
a fixed, well-defined problem set. In essence, it is an optimizer. In the financial industry, the
use of machine learning (ML) and artificial intelligence (AI) systems will only increase. This
trend is propelled by notable advancements in modeling and use-case adaptations, along with
quick rises in processing power, data storage capacity, and big data. Financial service

39
providers will find AI/ML systems more appealing as a result of the COVID-19 pandemic's
acceleration of the transition to a more contactless environment and increasingly digital
financial services. According to the analysis, artificial intelligence and machine learning
have advanced significantly in recent years, making it possible to create applications
for finance professionals that may or may not cause disruptions in the industry. As a result,
it's thought that AI and ML could both surpass human performance standards and completely
or partially replace human resources. Its applications are found in
many companies across the globe.
It was discovered that phenomena identification involved the use of AI and ML. They
are employed in the development of successful investing plans. Algorithmic trading is
another use of AI and Ml in finance. Using proprietary algorithms, these systems use
information about shifting market conditions and price levels to make automated
trades quickly. The analysis of text processing, news, and semantic syntax is another
important application of AI and ML. Intelligent reading and interpretation of data, including
papers, articles, social media posts, and other content, is accomplished through the use of AI
and ML. If the AI and ML system can read in a matter of seconds while it takes several hours
to do so and still cannot provide all the details that
might affect the product's specific performance, then this is very significant. It will be crucial
for investment services to grow in the future. Data mining is an approach that makes it easier
to study market data, forecast behavior, and price level. Additionally, it can support model
performance and forecasts for structural and legislative development.
The use of AI and ML in market sentiment analysis emerges from earlier uses, such
as text mining, and is frequently regarded as a subset of it. Social networking
platforms have undergone significant development in recent times, and vast amounts of data
have been processed. The growing volume of "big data" resulting from personal online
experiences and different social networking platforms opens up exciting new avenues
for behavior research. For instance, a quantitative analysis of social
media content has revealed that the results of multiple upcoming elections are more likely to
be ascertained. Furthermore, a number of other studies concerning the quantity of internet

40
queries concerning a particular topic with a new company. Financial indicators like credit risk
calculation, credit rating and efficiency, bond rating, etc., may provide more efficient ways to
use AI and ML. According to some reports, the use of artificial neural networks boosts
business funding because their risk assessment and bankruptcy forecasts are more accurate.
ZestFinance, a cutting-edge machine learning platform that facilitates a better and more
effective borrower-lender relationship, was founded by former Google CIO Douglas Merrill.

Artificial intelligence (AI) is becoming a top business priority in the financial services sector,
even though not all businesses are prepared in areas like asset management, banking,
insurance, and other sectors. The industry as a whole is aware of the strategic importance of
AI and ML, and in order to keep up or compete, major players have already committed large
sums of money and resources to the field. In order to assist entrepreneurs in
navigating the challenges of implementing AI and ML systems and managing priorities
when doing so, this paper has offered a multi-stakeholder viewpoint. While AI has many
applications, the market needs to learn more about what it really entails.
Business executives will assess their objectives, challenge them, and decide how
important their AI and ML projects are outside of technologies in order to tell the
difference between truth and fiction. The use of AI and ML is expanding rapidly with both
clear advantages and even unexpected challenges. It forces executives to develop a deeper
comprehension of technology and to set a responsible, transparent path for
integrating AI successfully into their business models and overarching strategic objectives.

In consideration of the financial consequences, businesses can provide a sufficient

understanding of AI and ML and other technologies used in the industry by the senior
management and the board to provide effective control. This is particularly important because
the board members need to track significant issues that affect the long-term value of a
company. Throughout compliance with the Corporate Governance Code, the Board is'
required to determine the nature and scale of the major risks that it intends to take throughout
order to achieve its strategic objectives.

41
Global businesses should have strong internal control and risk management systems
to guarantee that a suitably updated risk framework is created, tracked, and shared in a proper
manner. AI decision-making, implementation, and use must all take place within a risk-
management framework that recognizes shifts in the business environment. Risk
identification, risk assessment, risk mitigation, and risk control are the four primary tasks that
it will entail, regardless of whether the system is centered on the ISO, the committee of
funding organizations, or some other standard. This plan should be backed by
early intervention, accident preparedness, crisis response plans, and preparation.
5.2 Recommendation
AI is poised to revolutionize how banks and financial companies operate, how they develop
cutting-edge products and services, and how they elevate the customer experience. It
has the potential to cut operating costs of the front, middle, and back office of the industry by
a staggering 20 to 25 percent (AI Applications in the top 4 Indian Banks, 2017). There is
no going back now that AI is here to stay. That being said, some things never really
change, and interpersonal communication between people is one of them. Technology
is not always the best choice, even when it is readily available. The spirit of AI adoption
and application might not extend to replace all tasks performed by people, who are
always the most qualified to build rapport and trust (E-Book: Artificial Intelligence with
Human Touch, 2017). One significant obstacle to the use of AI in any other sector for that
matter, is the fact that humans are the only ones involved. It is the dearth of experts with the
necessary machine learning science skills. In order for AI technology to be applied to
the banking and financial services sector in India quickly and widely, it is imperative that in-
house training programs be created to teach existing staff members AI skills. Additionally,
partnerships with universities and colleges must be established in order to produce
qualified data scientists who can work on real-world AI projects.

Also, any massive adoption of AI technology in banking sector to the neglect of the human
element is fraught with the risk of perceived redundancy of human skills of the bank
employees leading to large scale layoffs and resultant unemployment in the industry (AI

42
Applications in the top 4 Indian Banks, 2017). Equally important, there is an urgent need for
globally accepted legislative and regulatory framework for application of this technology in
an ethical manner not detrimental to the interest of targeted users in general and to prevent its
misuse in the hands of unscrupulous users (Ludwig, E., 2018).
Another argument in favor of combining artificial intelligence and human touch is
that, in order to prevent an impersonal and artificial system from having the final say in
financial transactions, the regulatory framework must allow for appeals to human authorities
against decisions made by AI algorithms when a customer feels unfairly treated or wronged.
In line with the global policy of "suitability and appropriateness" in
derivative transactions, regulatory measures may be imposed in the AI framework
for consumer protection in order to individually or collectively guarantee the
"suitability and appropriateness" of financial solutions or advice given to clients. Artificial
intelligence (AI) is superior to human intervention in handling complex personalized
requests, understanding sentiments, establishing trust, and creating an
emotional connection with the customer to win his attention and ensure his brand loyalty. AI
is good at handling simple, repetitive tasks and automated conversations. To improve the
customer experience, wisdom would dictate carefully balancing human intervention with AI
functionality when needed.

43
 Reference

2018 Artificial Intelligence (AI) Predictions. (2018). Retrieved 16 March 2020, from
https://www.pwc.com/us/en/services/consulting/library/artificial-intelligence-predictions.html

Ahn H and Kim K-j. (2009). Bankruptcy prediction modeling with hybrid case-based
reasoningand genetic algorithms approach. Applied Soft Computing 9(2):599-607.

AI Applications in the top 4 Indian Banks (2017) - https://www.techemergence.com/

(Accessed on 16.12.18)

Alexandra, Z. (2017). Artificial Intelligence in Finance: Forecasting Stock Market Returns

using Artificial Neural Networks (September 29, 2017). Available at:
https://helda.helsinki.fi/dhanken/bitstream/handle/123456789/170154/zavadskaya.pdf?
sequence=1

Alhnaity B and Abbod M. (2020). A new hybrid financial time series prediction model.
Engineering Applications of Artificial Intelligence 95: article 103873.

Amin Naseri M. R. et al. (2020). Modeling Portfolio Optimization based on Fundamental

Analysis using an Expert System in the Real Estate Industry. International Journal of Supply
& Operations Management 7(1): 39-50.

Anupan, M. (2019). Artificial Intelligence in Financial services – Need to Blend Automation

with Human Touch. Available at: https://www.researchgate.net/publication/334764043.

Arash, B. (2009). A comparative survey of artificial intelligence applications in finance:

artificial neural networks, expert system and hybrid intelligent systems. Available at:
https://www.researchgate.net/profile/Arash_Bahrammirzaee/publication/201829144_A_Com

44
parative_Survey_of_Artificial_Intelligence_Applications_in_Finance_Artificial_Neural_Net
works_Expert_System_and_Hybrid_Intelligent_Systems/links/5dc95bf8a6fdcc57503eda70/A

Biallas M and O’Neill F. (2020). Artificial intelligence innovation in financial services. EM

Compass Note 85, IFC, June, 1-8.

Bjoern, K. et al. (2010). Financial Time Series Forecasting with Machine Learning
Techniques: A Survey. Available at:
https://s3.amazonaws.com/academia.edu.documents/59922561/Financial_Time_Series_Forec
asting_with_Machine_Learning20190703-28941-ae1b16.pdf

Boden MA. (2018). Artificial Intelligence: A Very Short Introduction. Oxford University
Press: Oxford, UK.

Bredt, S. (2019). Artificial intelligence (AI) in the financial sector. Frontiers in Artificial
Intelligence 2: 1-5.

Chawla, R. (2018). How AI Supports Financial Institutions for Deciding

Creditworthiness.Retrieved 17 March 2020, from
https://www.entrepreneur.com/article/310262.

Columbus, L. (2019). How AI Is Protecting Against Payments Fraud. Retrieved 19 March

2020, from https://www.forbes.com/sites/louiscolumbus/2019/09 /05/how-ai-is-protecting-
against-paymentsfraud/#305acba44d29.

Daks, M. (2018). Banking on Technology: Artificial intelligence helping banks get smarter.
Njbiz, 31(7), 10

Dirican, C. (2015). The Impacts of Robotics, Artificial Intelligence On Business and

Economics. Available at: https://cyberleninka.org/article/n/926024.pdf

Dwivedi, Y. K. et al. (2019). Artificial Intelligence (AI): Multidisciplinary perspectives on

emerging challenges, opportunities, and agenda for research, practice and policy.
International Journal of Information Management:101994.

Dyzma, M. (2019). Fraud Detection with Machine Learning: How Banks and Financial
Institutions Leverage AI. Retrieved 19 March 2020, from
https://www.netguru.com/blog/fraud-detection-withmachine-learning-how-banks-and-
financial-institutionsleverage-ai.

E-Book- Artificial Intelligence with Human Touch, Forrester Research Inc, November 2017,
Available at:
https://genbin.genesys.com/old/resources/Artificial_Intelligence_With_The_Human_Touch_v
2.pdf (Accessed on 15.02.19)

Eksin, I. and Erol, O. K. (2000), “A Fuzzy Identification Method for Nonlinear Systems”,

45
Turk J ElecEngin, vol. 8, no. 2, pp. 125-135.

Fountaine, T., McCarthy, B., and Saleh, T. (2019). Building the AI-Powered Organization.
Retrieved 20 March 2020, from https://hbr.org/2019/07/buildingthe-ai-powered-organization.

FRPT- Finance Snapshot, 23–25. (2017), p23-25. 3p.No job losses due to chatbots, artificial
intelligence: Banks.

Gavrilova, T. et al (2018). Modeling methods for strategy formulation in a turbulent

environment. Strategic Change 27(4):369-377.

Giudici, Paolo and Fintech (2018). Risk Management: A Research Challenge for Artificial
Intelligence in Finance.Available at:
https://www.frontiersin.org/articles/10.3389/frai.2018.00001/full.

Gómez M. et al. (2019). Big Data Algorithmic Trading Systems Based on Investors' Mood.
Journal of Behavioral Finance 20(2):227-238.

Guy A. Messick. (2017). Artificial Intelligence: The Ultimate Disrupter. Credit Union Times,
28(38), 12.

How Robots are changing the face of banking [2018] – Available at

http://www.theasianbanker.com/updates-and-articles/ robots-are-changing-the-face-of-
banking (Accessed on 15.12.18)

International Banker (2018), How AI is disrupting the banking industry-

https://internationalbanker.com/banking/

Kerkez, N. (2020). Artificial Inteligence and Machine Learning Can Repurpose Humans, Not
Replace Them. American Bankers Association. ABA Banking Journal 112(6):30-32.

Kim, J., Won, C. and Bae J. K. (2010). A knowledge integration model for the prediction of
corporate dividends. Expert Systems with Applications 37(2):1344-1350.

Kolarik and Rudorfer, G. (1997) “Time series forecasting using neural networks, department
of applied computer science”, Vienna University of Economics and Business Administration,
no. 1090, pp. 2–6.

Königstorfer F & Thalmann S. (2020). Applications of artificial intelligence in commercial

banks – A research agenda for behavioral finance. Journal of Behavioral and Experimental
Finance 27: 1-15

Kou et al. (2014). Evaluation of clustering algorithms for financial risk analysis using
MCDM methods. Available at:
https://www.researchgate.net/profile/Gang_Kou/publication/292137095_Information_Scienc
es-Kou-2014-4/links/56a9715808ae2df82165333a/Information-Sciences-Kou-2014-4.pdf.

46
Kumba Sennaar (2019), AI in Banking – An Analysis of 7 Top American Banks, Available at
https://www.techemergence.com/ai-in-banking- analysis/ [2018] (Accessed on 15.02.19)

Lee J. N., Morduch J., Ravindran S, Shonchoy A and Zaman H. (2021). Poverty and
Migration in the Digital Age: Experimental Evidence on Mobile Banking in Bangladesh.
American Economic Journal: Applied Economics 13(1):38-71.

Li, W. and Mei, F. (2020). Asset returns in deep learning methods: An empirical analysis on
SSE and CSI 300. Research in International Business and Finance 54:101291.

Lo, A. W. (2016). Moore's Law vs. Murphy's Law in the Financial System: Who's Winning?
BIS Working Paper No. 564. Available at SSRN: https://ssrn.com/abstract=2789737

Ludwig, E. (2018). Regulators have their eye on AI. American Banker, 183(130), 1.

Maknickienė, N. and Maknickas, A. (2013). Financial Market Prediction System With

EvolinoNeural Network and Delphi Method. Journal of Business Economics & Management
14(2):403-413.

Meinert, M. C. (2018). Artificial Intelligence: The Next Frontier of Cyber Warfare? ABA
Banking Journal, 110(3), 43.

Mejia, N. (2020). AI-Based Fraud Detection in Banking – Current Applications and Trends |
Emerj. Retrieved 19 March 2020, from https://emerj.com/aisector-overviews/artificial-
intelligence-fraud-banking/

Meryem, D. F. et al. (2009). Assessing Bank Performance with Operational Research and
Artificial Intelligence Techniques: A Survey. Available at:
http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.473.4002

Nunn, Robin. 2018. “Workforce Diversity Can Help Banks Mitigate AI Bias.” American
Banker 183 (104): 1

Oh, K. and Kim, T. (2007). Financial market monitoring by case-based reasoning.

ExpertSystems with Applications 32(3):789-800.

Padmaavathy, P. A. (2018). Artificial intelligence: The weapon of choice in banks’ fight for
survival. International Journal of commerce and management research 4(5): 70-75.

Philippe, M. and Olivier, B. (2010). A Generic Architecture for Realistic Simulations of

Complex Financial Dynamics. Available at:
https://www.researchgate.net/profile/Philippe_Mathieu/publication/220988433_A_Generic_
Architecture_for_Realistic_Simulations_of_Complex_Financial_Dynamics/links/
0fcfd50806bbd55308000000/A-Generic-Architecture-for-Realistic-Simulations-of-Complex
_Financial_Dynamics.pdf

Ponduru, A. (2019). Artificial Intelligence: A New Age in the Stock Market. Retrieved 18

47
March 2020,from https://medium.com/datadriveninvestor/artificialintelligence-a-new-age-in-
the-stock-market991ac08aa11e.

Press, G. (2020). Artificial Intelligence Rapidly Adopted by Enterprises, Survey Says.

Retrieved 16 March 2020, from https://www.forbes.com/sites/gilpress/2016/07/20/artificial-
intelligence-rapidly-adopted-by-enterprisessurvey-says/#4e9531f512da

Quah T. S. and Srinivasan, B. (1999). Improving returns on stock investment through neural
Network selection. Expert Systems with Applications 17(4):295-301.

Rehfisch (2019). AI: Smarter Credit Processes Thanks to Artificial Intelligence. Retrieved 17
March 2020, from https://www.knowis.com/blog/smarter-credit-processes-thanks-to-
artificial-intelligence.

Rödder, W. et al. (2010). An entropy-driven expert system shell applied to portfolio selection.
Expert Systems with Applications 37(12):7509-7520.

Roohi, F. (2013), “Artificial Neural Network Approach to Clustering’, International Journal

of Engineering and Science, vol. 2, Issue 3, pp. 3338.

Safdar NM, Banja JD & Meltzer CC. (2020). Ethical considerations in artificial intelligence.
European Journal of Radiology 122:108768.

Sarvady, G. (2017). Chatbots, Robo Advisers, & AI: Technologies presage an enhanced
member experience, improved sales, and lower costs. Credit Union Magazine, 83(12), 18–22.

Satell, G. (2016). Teaching an Algorithm to Understand Right and Wrong. Harvard Business
Review Digital Articles, 2–5.

Shah, A. (2020). How AI Transforms the Future of Trading. Retrieved 18 March 2020, from
https://www.fusioninformatics.com/blog/how-aitransforms-the-future-of-trading/

Shorouq, F. E. et al (2010). Neuro-Based Artificial Intelligence Model for Loan Decisions

(2010). Available at:
https://www.researchgate.net/profile/Saad_Yaseen/publication/290614456_NeuroBased_Artif
icial_Intelligence_Model_for_Loan_Decisions/links/56ec38b608aea35d5b981f2b/Neuro-
Based-Artificial-Intelligence-Model-for-Loan-Decisions.pdf

Siddiqui, T. A. and Abdullah, Y. (2014). Non-Linear Relationship Between Macroeconomic

Variables and Stock Prices in India: An Artificial Neural Networks Approach. Management
Dynamics 14(2):52-59.

Siddiqui, T. A. and Siddiqui K. I. (2020). FinTech in India: An analysis on impact of

telecommunication on financial inclusion. Strategic Change 29(3):321-330.

Stahl, B.C. (2021). “Conclusion. In: Artificial Intelligence for a Better Future”.SpringerBriefs
in Research and Innovation Governance. Springer, Cham. https://doi.org/10.1007/978-3-030-

48
69978-9_8.

Taddy (2019). The technological elements of artificial intelligence. In Agrawal A, Gans J

&Goldfarb A. (eds), The Economics of Artificial Intelligence: An Agenda. Chicago,
IL:University of Chicago Press.

The Power of AI for Business | Accenture. (2020). Retrieved 20 March 2020, from
https://www.accenture.com/in-en/service-ai-for-business-transformation.

Tom, L. (2019). Artificial Intelligence, Finance, and the Law. Available at SSRN:
https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3480607

Vellido, A. et al. (1999), “Neural networks in business: a survey of applications’, Expert

Systems with Applications 17, Elsevier, pp. 51-70.

Victories, V. (2018). 7 Benefits of Artificial Intelligence for Business. Retrieved 20 March

2020, from https://www.eukhost.com/blog/webhosting/7- benefits-of-artificial-intelligence-
for-business/

Vordenbaeumen, H. (2019). 3 ways accountants can implement AI today. Retrieved 19 March

2020, from https://www.accountingtoday.com/opinion/3-waysaccountants-can-implement-ai-
today.

Wang, S. and Shang, W. (2014). Forecasting Direction of China Security Index 300
Movementwith Least Squares Support Vector Machine. Procedia Computer Science 31:869-
874.

Xing et al., (2017). Natural Language Based Financial Forecasting: A Survey. Available at:
https://dspace.mit.edu/bitstream/handle/1721.1/116314/10462_2017_9588_ReferencePDF.pd
f?sequence=2&isAllowed=y

Yang, G., Chen, Y. and Huang, J. P. (2016). The highly intelligent virtual agents for modeling
financial markets. Physica A: Statistical Mechanics and its Applications 443:98-108.

Zipperle, T. (2020). 7 ways artificial intelligence and machine learning will impact the
finance office. Retrieved 20 March 2020, from https://www.accountingtoday.com/list/7-ways-
artificial-intelligence-and-machine-learning-will-impact-the-finance-office.

49