mots2791@gmail.

com
7OPR3VT6YI

This file is meant for personal use by [email protected] only.

Proprietary content. ©Great
Sharing Learning.
or publishing theAll RightsinReserved.
contents part or fullUnauthorized use
is liable for legal or distribution
action. prohibited.
 DSBA CURRICULUM DESIGN
CORE DOMAIN
FOUNDATIONS COURSES APPLICATIONS

[email protected]
7OPR3VT6YI

This file is meant for personal use by [email protected] only.

Proprietary content. ©Great
Sharing Learning.
or publishing theAll RightsinReserved.
contents part or fullUnauthorized use
is liable for legal or distribution
action. prohibited.
 Learning Objective of This Module
• Different Types of Risk

• Credit Risk
[email protected]
7OPR3VT6YI

• Market Risk

This file is meant for personal use by [email protected] only. 3

Proprietary content. ©Great
Sharing Learning.
or publishing theAll RightsinReserved.
contents part or fullUnauthorized use
is liable for legal or distribution
action. prohibited.
 What is Risk?
• Risk may be defined as an exposure to a transaction
with loss, which occurs with some probability and
which can be expected, measured and minimized.
[email protected]
7OPR3VT6YI
• Is Risk good, bad or ugly?

This file is meant for personal use by [email protected] only.

Proprietary content. ©Great
Sharing Learning.
or publishing theAll RightsinReserved.
contents part or fullUnauthorized use
is liable for legal or distribution
action. prohibited.
 Different Sources of Risk
Product Risk
Business Risk Macroeconomic Risk
Technological Risk

Firm wide Risk

Legal Risk
Reputational
[email protected] Risk
7OPR3VT6YI Event Risk
Disaster Risk

Political Risk
Non Business
Risk
Market Risk

Credit Risk
Financial Risk
Liquidity Risk
Operational
Risk

This file is meant for personal use by [email protected] only.

Proprietary content. ©Great
Sharing Learning.
or publishing theAll RightsinReserved.
contents part or fullUnauthorized use
is liable for legal or distribution
action. prohibited.
 Credit Risk
• A credit risk is the risk of default on a debt that may arise from a borrower failing
to make required payments.
• The risk is that of the lender and includes lost principal and interest, disruption to
cash flows, and increased collection costs.
[email protected]
7OPR3VT6YI
• The loss may be complete or partial.

This file is meant for personal use by [email protected] only.

Proprietary content. ©Great
Sharing Learning.
or publishing theAll RightsinReserved.
contents part or fullUnauthorized use
is liable for legal or distribution
action. prohibited.
 Examples of Credit Losses
• A consumer may fail to make payment due on a mortgage loan, credit card, or
any other loan
• A company is unable to repay asset-secured fixed or floating charge debt.
• An insolvent insurance company does not pay a policy obligation
[email protected]
7OPR3VT6YI

• An insolvent bank does not return the funds to a depositor

• A government grants bankruptcy protection to an insolvent consumer or business

This file is meant for personal use by [email protected] only. 7

Proprietary content. ©Great
Sharing Learning.
or publishing theAll RightsinReserved.
contents part or fullUnauthorized use
is liable for legal or distribution
action. prohibited.
 Assessment of Credit Risk

• Probability of Default(PD) – What is the likelihood that the counterparty will

default on its debt obligation.
• Exposure at Default(EAD) – In the event of a default, how large will the
[email protected]
7OPR3VT6YI
outstanding obligation be when the default occurs.
• Loss given default (LGD)- This is the amount of money a bank or other
financial institution loses when a borrower defaults on a loan, depicted as a
percentage of total exposure at the time of default.

This file is meant for personal use by [email protected] only.

Proprietary content. ©Great
Sharing Learning.
or publishing theAll RightsinReserved.
contents part or fullUnauthorized use
is liable for legal or distribution
action. prohibited.
 PD Models - Statistical Models in Credit
Risk Measurement
PD Models

[email protected]
7OPR3VT6YI
Linear Models Non-Linear

Linear
Discriminant
Probability Logit/Probit
Analyses
Models

This file is meant for personal use by [email protected] only.

Proprietary content. ©Great
Sharing Learning.
or publishing theAll RightsinReserved.
contents part or fullUnauthorized use
is liable for legal or distribution
action. prohibited.
 Case-1
• Prediction of probability of default on payment towards credit card for individual
cardmembers.
• The dataset captures credit card customer’s demographics such as age, gender,
marital status, education etc.
[email protected]
7OPR3VT6YI
• Some additional information shared as limit available on card , fico score, balance
amount and payment made for the last three months.
• There are 10000 observations listed and you are asked to build a probability
model to deploy and predict the risk of default.

This file is meant for personal use by [email protected] only.

Proprietary content. ©Great
Sharing Learning.
or publishing theAll RightsinReserved.
contents part or fullUnauthorized use
is liable for legal or distribution
action. prohibited.
 Companies’ PD Models – Linear Models
Altman’s Z Score Model
• Z (Public) = 1.2X(1) + 1.4X(2) + 3.3 X(3) + 0.6 X(4) + 1.0 X(5)
• Where Z > 2.99 is healthy
• Z < 1.81 is unhealthy
[email protected]
7OPR3VT6YI • 1.81 < Z < 2.99 is indeterminate

Financial Ratios :
• X(1) = Working Capital / Total Assets
• X(2) = Retained Earnings / Total Assets
• X(3) = Earnings Before Interest and Taxes (EBIT) / Total Assets
• X(4) = Market Value of Equity / Total Assets
• X(5) = Net Sales / Total Assets

This file is meant for personal use by [email protected] only.

Proprietary content. ©Great
Sharing Learning.
or publishing theAll RightsinReserved.
contents part or fullUnauthorized use
is liable for legal or distribution
action. prohibited.
 Discriminant Function
• India Z Score Model
• Z (Public) = 1.06 + 0.01PBITINT – 3.12TDTA + 0.48QR + 4.61NCATA
[email protected] • PBITINT = Profits Before Interest & Taxes / Total Interest
7OPR3VT6YI
• TDTA = Total Borrowings / Total Assets
• QR = Current Assets – Inventories / Current Liabilities & Provisions
• NCATA = Profit After Tax & Depreciation / Total Assets

This file is meant for personal use by [email protected] only.

Proprietary content. ©Great
Sharing Learning.
or publishing theAll RightsinReserved.
contents part or fullUnauthorized use
is liable for legal or distribution
action. prohibited.
 • Challenges of Discriminant Function
• Zone of Indifference (Indeterminate)
• Sensitivity to Industry
• Dealing with New Companies
[email protected]• Loss of Predictive Power across Time – use of Penalty Functions
7OPR3VT6YI

• Assumption of Multivariate Normality of Variables

This file is meant for personal use by [email protected] only.

Proprietary content. ©Great
Sharing Learning.
or publishing theAll RightsinReserved.
contents part or fullUnauthorized use
is liable for legal or distribution
action. prohibited.
 Steps to Build a Good Function
• Selection of the ‘Default’ Sample
• Creation of the Non-Default Sample
• Choice of Industry, Firm Size & Time Period

• Appropriate Treatment of the Data

[email protected]
7OPR3VT6YI

• Choose the predictors carefully

• Choice of a Good Out of Sample Validation Data – Stress Test the Data

This file is meant for personal use by [email protected] only.

Proprietary content. ©Great
Sharing Learning.
or publishing theAll RightsinReserved.
contents part or fullUnauthorized use
is liable for legal or distribution
action. prohibited.
 Logistic function – a better fit!
So, we need a function that stays within the bounds of 0 & 1 and
represents the data in a much better manner

Sigmoid or Logistic Function

[email protected]
7OPR3VT6YI

1
_______
Y= 1 + e-z

Z= + X
i i

This file is meant for personal use by [email protected] only.

Proprietary content. ©Great
Sharing Learning.
or publishing theAll RightsinReserved.
contents part or fullUnauthorized use
is liable for legal or distribution
action. prohibited.
 CASE-2
• Prediction of probability of default for various companies given certain
information such as Total Assets, Net Worth, Cash Profit, Profit After Tax etc.
• The data is given in CompanyDefault.csv file.
• There are 1384 observations listed and you are asked to build a probability model
[email protected]
7OPR3VT6YI
to deploy and predict the risk of default for companies.

This file is meant for personal use by [email protected] only.

Proprietary content. ©Great
Sharing Learning.
or publishing theAll RightsinReserved.
contents part or fullUnauthorized use
is liable for legal or distribution
action. prohibited.
 Steps for Case-Study-2

• Selection of the ‘Default’ Sample - Assumptions that if net worth <=5,

defaulted companies.
• Creation of the Non-Default Sample
• Choice of Industry, Firm Size & Time Period

• Appropriate Treatment of the Data

[email protected]
7OPR3VT6YI

• Check –Min, Max, Mean,Std Dev, Various percentiles to treat missing values
and outliers
• Create relevant ratios as new features is very essential

• Import the csv version in Python and run logistic regression

• Choose the predictors carefully

• Choice of a Good Out of Sample Validation Data – Stress Test the Data
This file is meant for personal use by [email protected] only.
Proprietary content. ©Great
Sharing Learning.
or publishing theAll RightsinReserved.
contents part or fullUnauthorized use
is liable for legal or distribution
action. prohibited.
 Process of Building a Smart Default Model
• Define Default
• Do a data quality check and remove outliers
• Build new features such as various ratio-based variables
• Divide the data into development and validation datasets
[email protected]
7OPR3VT6YI • Develop Logistic function
• Develop Classification table
• Test performance on out-of-sample data

This file is meant for personal use by [email protected] only.

Proprietary content. ©Great
Sharing Learning.
or publishing theAll RightsinReserved.
contents part or fullUnauthorized use
is liable for legal or distribution
action. prohibited.
 Define Default
• Step 1 : Define Default in your dataset
• All the company with negative Net worth should be defined as default
• If Net worth <= 5 , then default = 1, else default = 0 ( Add a new column in your dataset called
Default which takes a value of 0/1)
[email protected]• What is the default rate in your dataset ?
7OPR3VT6YI

This file is meant for personal use by [email protected] only.

Proprietary content. ©Great
Sharing Learning.
or publishing theAll RightsinReserved.
contents part or fullUnauthorized use
is liable for legal or distribution
action. prohibited.
 Quality Check
• Step 2 – Quality Check of the data
• Have a look at the data and remove the outliers
• These outliers will distort your equation if not removed
• After you are satisfied with the data, randomly divide it into development dataset (65%) and
[email protected] validation dataset(35%)
7OPR3VT6YI
• Check the default rates of the development and validation dataset
• since it is randomly split, the default rates should be similar
• You will now work on development dataset to develop the model equation

This file is meant for personal use by [email protected] only.

Proprietary content. ©Great
Sharing Learning.
or publishing theAll RightsinReserved.
contents part or fullUnauthorized use
is liable for legal or distribution
action. prohibited.
 Identification of Predictors
• Step 3 – Identify the predictors
• Look at the variables which give a good discrimination between the defaulters and non-
defaulters in the dataset
• E.g. If you want to use variable X in your model, you should look at the mean value of variable X for
non-defaulters , and the mean value for defaulters. If there is a good discrimination between the
[email protected]
7OPR3VT6YI two mean values, then the variable X should be used in your model

• Ratio variables MIGHT be more appropriate for the model equation!

This file is meant for personal use by [email protected] only.

Proprietary content. ©Great
Sharing Learning.
or publishing theAll RightsinReserved.
contents part or fullUnauthorized use
is liable for legal or distribution
action. prohibited.
 Develop Logistic Equation
• Step 4 : Develop Logistic Equation
• Read the datafile in Python
• Run Logistic Regression
• Dependent variable is the 0/1 Default Column
[email protected]• Independent variables are the predictors that you have chosen
7OPR3VT6YI

• Remove the variables that are insignificant

• Fine tune your final model by looking at the sign of the predictors
• Look at the efficiency rate of the model prediction

This file is meant for personal use by [email protected] only.

Proprietary content. ©Great
Sharing Learning.
or publishing theAll RightsinReserved.
contents part or fullUnauthorized use
is liable for legal or distribution
action. prohibited.
 Out of Sample Validation
• Step 5 : Your model is now ready
• Take the validation dataset
• Score each row with your equation – get the score for everyone
• Count the number of defaulters in your dataset
[email protected]
7OPR3VT6YI
• Lets assume that there are 30 defaulters in your dataset
• Based on your score, take the top 30 companies
• Look at the default rate of these top 30 companies – This gives you the efficiency of your model
• Higher the efficiency, better is the model

This file is meant for personal use by [email protected] only.

Proprietary content. ©Great
Sharing Learning.
or publishing theAll RightsinReserved.
contents part or fullUnauthorized use
is liable for legal or distribution
action. prohibited.
 You Now have your Own Rating Company!
o Arrange the data in Descending Order
o Bottom Ten Percent – Rate them A
o Top Ten Percent Rate – Rate them D
o Next Thirty Percent – Rate Them C
[email protected]
7OPR3VT6YI

o Next Forty Percent – Rate them B

This file is meant for personal use by [email protected] only.

Proprietary content. ©Great
Sharing Learning.
or publishing theAll RightsinReserved.
contents part or fullUnauthorized use
is liable for legal or distribution
action. prohibited.
[email protected]
7OPR3VT6YI
Examples of Risks in next few slides

This file is meant for personal use by [email protected] only.

Proprietary content. ©Great
Sharing Learning.
or publishing theAll RightsinReserved.
contents part or fullUnauthorized use
is liable for legal or distribution
action. prohibited.
 Chanda Kochhar – Reputational Risk
• In December 2008, ICICI Bank MD Chanda Kochhar’s husband Deepak Kochhar and the Videocon
group promoter Venugopal Dhoot float a clean energy firm NuPower Renewables Pvt Ltd (NRPL).
Deepak Kochhar and his relatives owned one half of the company. The remainder was owned by
Dhoot, his family and his associates, according to a media report.
• In 2012, a consortium of over 20 banks and financial institutions lend Rs 40,000 crore to the
[email protected]
7OPR3VT6YI Videocon group (Videocon Industries Ltd and 12 of its subsidiaries/ associates as co-obligors). The
State Bank of India (SBI) was the lead banker in that consortium. ICICI Bank was part of the
consortium and lent about Rs. 3250 crore, which is less than 10 percent of the total consortium
facility, The Indian Express reported.
• Some six months after the sanction of the abovementioned loan, Deepak Kocchar reportedly
assumes majority control of NRPL.

This file is meant for personal use by [email protected] only.

Proprietary content. ©Great
Sharing Learning.
or publishing theAll RightsinReserved.
contents part or fullUnauthorized use
is liable for legal or distribution
action. prohibited.
 Kingfisher – Credit Risk
According to sources in the CBI, Kingfisher applied for a corporate loan of Rs. 750 crore to IDBI on
October 1, 2009. Even as this application was pending, Mr Mallya met with the then IDBI chairman
Yogesh Agarwal on October 6, seeking a short term loan of Rs. 150 crore. On the very next day,
Kingfisher officially filed for this loan and it was sanctioned immediately. Again on November 4,
Kingfisher applied for another short term loan to IDBI, which was, once again, granted the same day.
[email protected]
7OPR3VT6YI
Kingfisher Chief Financial Officer Raghu Nathan wrote to IDBI mentioning a meeting between Mr Mallya
and Mr Aggarwal, which led to this loan being sanctioned the same day. The credit committee of IDBI
was to meet to consider Kingfisher's corporate loan application later that month. For that Kingfisher had
to pledge its unencumbered shares, a mandatory condition to get the loan. But the CBI says on
November 10, Mr Mallya mailed Kingfisher CFO Raghu Nathan telling him there was no need to pledge
the shares because he had already discussed this with the IDBI CMD and that was the 'accepted
position'. The IDBI credit committee had a meeting on this on the 24th of the month and Mr. Aggarwal
allegedly approved the minutes, accepting Mr Mallya's condition. CBI sources also claim that 263 crores,
which was a part of the loan, has been diverted out of the country and used to pay up previous loans of
Kingfisher
This file is meant for personal use by [email protected] only.
Proprietary content. ©Great
Sharing Learning.
or publishing theAll RightsinReserved.
contents part or fullUnauthorized use
is liable for legal or distribution
action. prohibited.
Apply Data Science at your workplace to gain some instant benefits:

• Get noticed by your management with your outstanding analysis backed by data science.
[email protected]
7OPR3VT6YI

• Create an impact in your organization by taking up small projects/initiatives to solve critical issues
using data science.

• Network with members from the data science vertical of your organization and seek opportunities
to contribute in small projects.

• Share your success stories with us and the world to position yourself as a subject matter expert in
data science.
This file is meant for personal use by [email protected] only.
Proprietary content. ©Great
Sharing Learning.
or publishing theAll RightsinReserved.
contents part or fullUnauthorized use
is liable for legal or distribution
action. prohibited.
[email protected]
7OPR3VT6YI

ANY QUESTIONS
This file is meant for personal use by [email protected] only.
Proprietary content. ©Great
Sharing Learning.
or publishing theAll RightsinReserved.
contents part or fullUnauthorized use
is liable for legal or distribution
action.
Proprietary content. ©Great Learning. All Rights Reserved. Unauthorized use or distribution prohibited.
prohibited.
[email protected]
7OPR3VT6YI

HAPPY LEARNING
This file is meant for personal use by [email protected] only.
Proprietary content. ©Great
Sharing Learning.
or publishing theAll RightsinReserved.
contents part or fullUnauthorized use
is liable for legal or distribution
action.
Proprietary content. ©Great Learning. All Rights Reserved. Unauthorized use or distribution prohibited.
prohibited.