Data Science Project: Predicting Credit Card Default Rates - PR1

This project, submitted by Ruhi Miglani, focuses on predicting credit card defaults using data science and financial technology. The project identifies the increasing issue of credit card defaults, exacerbated by factors like the COVID-19 pandemic, and proposes a startup, 'Default Prevention Pro Services,' to assist financial institutions. The methodology involves collecting and analyzing customer data, employing business intelligence, predictive analytics, R language, and SAS Enterprise Miner to generate insights and reports. The proposed business model includes revenue generation through service contracts and aims to offer valuable predictions to prevent financial losses for lenders. The project also emphasizes data ethics and governance, ensuring compliance with relevant laws and ethical standards to protect data and promote transparency. The budget outlines costs for IT infrastructure, office setup, and staffing, highlighting the financial aspects of the venture.

Submitted By: Ruhi Miglani
19828121
ASSIGNMENT-01(PR1)

Table of Contents
1. Background
2. Identified problem
3. Proposal
4. Design, Development and Deployment
5. Adopted Analytical Approaches
6. Data Driven Solutions
7. Business Model
8. Business Opportunities
9. Estimated Budget
10. Data Ethics & Governance

Background
Financial Technology is an economical sector made up of businesses leveraging technologies to
improve the performance of financial services and is often called Fintech. Its meaning is changing
drastically with the time that it is so difficult to classify the concept of fintech. Active sectors of
Fintech are Lending, Wealth-Management, Payments, Data and Information Management, Market
place Style solutions etc. However, among all sectors Payment and lending comes under the major
sector in which fintech is involved but with the time and with changing rules and regulations in
payment and lending sector various new finance firms came up which are offering better credit
services and rates to the customer in compare to banks. Moreover, credit card is considered or being
used as the best solution served by the banks and other credit providing firms (Anon, 2020). There
has been substantial growth in the number of lenders of cards and issuing banks and transactions with
credit card and this leads to serious concerns as upsurge of transactions via credit cards give rise to
more financial pressure on the borrower.
Currently the lending sector is facing too much of issues because of rise in credit card defaults.
Because of COVID 19 most of the peoples are jobless, they even don’t able to fulfill their daily basic
needs in such conditions it is not possible for them to pay credit card monthly instalments and even to
get rid from this situation people are opting various different sources to pay the dues and as a result it
is giving rise to more debt. However, multiple solutions are taken into consideration to predict the
default rate but still the condition is same.
Identified Problems
The major problem identified is credit card defaults and this is because of ease in availability of
credit. Now-a-days it becomes easy to get credit from any financial institution or credit provider firms
and this is creating a massive liability or debt. As credit is easily available around 16 million of
people are having credit cards. The credit cards are nothing but the loan taken by the customer from
financial institutions with implied obligation repayment in the form of monthly instalments borrower
have to pay to the lending institute along with that interest rate and penalties for not paying instalment
on time also contributing to a rise in high debts. Most of the people are using it in case of emergency
or it act a resource of finance in the case of any emergency for them and others are using it as a source
of gaining rewards like in flights booking they offer cashbacks and some discount if payment is done
via credit card but as per the report of ASIC users were overwhelmed with debt obligations on credit
cards, and the financial firms are owing almost $45 billion since June 2017 and Credit card defaults
are currently being pursued by 1.9 million people in Australia. Because of high debts people are
applying for other credit card so that they can be able to make payment for earlier dues and debts also

they took credits from peers this expands their burden. Additionally, digitalization people prefer to
use credit cards instead of cash this give rise to more problems for both the lenders and the borrowers.
Proposal
Default Prevention Pro Services
Considering the issue of credit default and all those problems mentioned above and the current
situation as COVID 19 impacts the financial market. It can be easily forecast with the current situation
that the number of credit defaults will go up as borrowers were not able to pay the interest, their
monthly instalments because they lost their jobs, most of the businesses are in loss and there is no
source of income but they have various obligations like their rent, various bills etc and currently there
is no update on when this ends and when people are able to get their jobs back, when this lockdown
ends. So, this is the perfect time to come up with the new start-up which provide help to the financial
institutions by predicting the future default rates so that they cannot bear too much loss because of the
credit defaults.
Design Development and Deployment
• First, we will collect the customer data from the banks and the credit lending firms the data is
nothing but historical transactions of the credit borrowers, their credit report which will help
to understand the spending pattern of the borrowers which will help to predict other factors.
• Then we will analyse the data with the help of various advanced analytical techniques in order
to produce meaning insights. With the help of the insights generated we will prepare the
report consisting all insights and required explanation
• Finally, we will deliver the report to the lending institutions so that for future loss or risk they
could do something in advance, or they can develop risk mitigation plan.

Analytical approaches adopted by the Default Prevention Pro Services are:
 Business Intelligence: It will help in utilizing the large volume of data effectively and
efficiently as with the help of BI it will become easy to convert the big data into actionable
insights. As, with this it will become easy to understand customer behaviour and accordingly
it will become easy to predict the credit default rate.
 Predictive Analytics: This is one of the most effective tools which directly helps the lending
sector to grow, this enables the study of historical data and based on that make predictions for
the future outcome. This will help the financial institutions to proactively manage the future
loss.
 R Language: This can be used in order to perform all crucial tasks related to finance. It helps
in performing credit risk analysis, using various codes we will be able to generate different
graphs, perform time series analysis. Additionally. Various R packages offers facility of data
mining with the help of various packages and R shiny is there to develop engaging
visualizations.
 SAS Enterprise Miner: This can be used as a major tool because it takes less time for data
preparation, quickly identifies the behaviour and existing relationship, creates the credit
scorecard which will help the lender to take wise decisions.
Solutions based on above mentioned techniques
 The data provided by the client is nothing but the historical transaction data of their
customers, with this data we can study the spending patterns and also able to get the clear
picture of what their customer prefer more, where they are spending too much money this will
help our client to understand the need of their client and accordingly they can offer the credit
services to them.
 With the help of R programming it is easy for us to generate different insights based on
different variables and this will help to understand which factor is more responsible for the
default.
 The credit report will give clear picture about the payment history of the customer, their
pending payments also how many times a person is not able to pay due instalment on time.
This will help to understand customer behaviour based on that we can predict the future
behaviour of the customer so that banks and lenders will not incur loss in future and
proactively handle the future debts. This we can done easily with the help of SAS.
 Financial model can also be created in python as the financial model will give clear picture
about the monetary impact of an individual based on the spending history of the customer.

 Same as above it will be easy to calculate the monetary impact of all customers as this will
allow us to identify who would be the riskier customer
 Combination of Past data and Machine learning techniques also helps to determine whether a
user is defaulting on his credit cards or not.
 Also, with predictive analytics it can be easy to identify the connections in between the
different variables and behavioral factors which helps to understand the various risks which
will directly help in making credit related decisions.
 Based on predictions now we have data of customers who can be defaulting and who cannot,
this will help to make decisions regarding how much credit limit is given to whom this will
help in preventing future losses.
 Customer segmentation can also be done with the help of data and further analysis can be
carried out to check various crucial factors giving rise to default.
All in all, the data provided by the banks and lenders helps to analyse the spending pattern of the
customers this will help to get clear picture about their monetary health based on that predictions
were made, and insights are generated. Past transactions and spending pattern help to identify
where the customer is spending more, what kind of the customer he is, based on that various
assumptions can be made which will help the lenders to make informed decisions.
Business Model
For Finance, we eagerly
looking for investors to
invest in our start-up
and major focus is on
Venture capital for the
finance. Target Market, our
target is Banks,
Financial Institutions
or all credit
providers.
We will generate the
insights based on given
data and identify the
potential defaulters
With the help of
Business Intelligence
and AI , we will study
the transaction patterns
and predict or calculate
default rate.

Business opportunities
Currently because of COVID 19 the economy of the world is in crisis, as people lost their jobs, they
don’t have any source of income. Even they are not able to fulfil their basic needs because they must
pay their rents, utility bills and many more things. This will automatically give rise to huge number of
credit card defaults. So, this kind of start-up is really required so that something could be done to
prevent huge losses. Here the firm has the competition with the various banks who internally do all
such predictions, but for them it hard to perform all the things so, our state-up can be a good option
for those who provides credit as their task is to provide data only rest all analysis will be performed
by us. This helps them to focus on other important aspects and it helps the lenders to take all possible
measures proactively.
Estimated Budget
Resources Cost
IT Infrastructure cost $60,000
Office set-up cost $100000
Hiring professional staff $120000
Total $280000
Revenue generation
• Initially, to attract more clients we will offer our services in less price this will directly help to
generate profit as well as to attract new clients.
• We also work on contract basis and different contracts have different prices and services
whosoever will opt the contract based on nature and type of contract we will offer our service
and generate revenue in return.
• After having a customer base or after building loyalty we will introduce more services based
on various advanced technology which will generate more clear predictions and insights also
change the price of our services in order to earn more.
Data Ethics and Governance

 The start-up will work according to the all applicable laws, code of conduct and will not try to
evade or sidestep their objective and will follow all policies framed for the businesses under
APRA.
 Data will be protected accordance to the GDPR and will used accordingly no data breaching
theft or misuse will conduct from our end.
 We will perform with Honesty and Integrity the interest of all the parties involve should not
be harmed.
 All the reports, description and advice given to the client will always for the best interest of
the client and doesn’t contain any wrong information.
 No advice and options suggested by us to the client is deceptive, it will not at all mis lead the
client.
 Ethical Behaviour must be promoted for the sake of protecting the interest of the public.
 Data science should be used by following all code of conduct and results produced with the
data science will be understandable by the human.
 We will work opting transparency with the client.
.