Uploaded by vg09082002

PI

advertisement
Credit Risk
model
What is credit
risk
Credit risk arises when a
corporate or individual borrower
fails to meet their debt
obligations. It is the probability
that the lender will not receive
the principal and interest
payments of a debt required to
service the debt extended to a
borrower.
Different measure of
credit risk
• Traditional credit models – credit rating
&
credit
scoring
–
strengths
n
weaknesses
• Probability density function of credit
losses (discussion on VaR)
• Parameter specifications – eg. Loss given
default, prob of default etc.
• Structural models
• Reduced form models
Model predictions
Logistic Regression for Defaults
Gradient Boosted Trees Using XGBoost
With the loan data fully prepared, we
Decision trees are another standard credit
will discuss the logistic regression
model which is a standard in risk
risk model. We will go beyond decision trees
modeling. We will understand the
by using the trendy XGBoost package in
components of this model as well as
Python to create gradient boosted trees.
how to score its performance. Once
After developing sophisticated models, we
we've created predictions, we can
explore
the
financial
utilizing this model.
impact
of
will stress test their performance and
discuss column selection in unbalanced
data.
Advantages of Credit Risk Management:
• Ability to measure and predict the risks of any single application.
• Allows banks planning strategies ahead to avoid a negative
outcome.
• Using various credit scoring models, it’s possible to figure out the
best ones for the business and determine the level of risk while
lending.
DisAdvantages of Credit Risk Management:
• Ability to measure and predict the risks of any single application.
• Allows banks planning strategies ahead to avoid a negative outcome.
• Using various credit scoring models, it’s possible to figure out the best ones for the business and determine the level of risk while
lending. Risk predictions do not guarantee low percentages of bad loans; the approach isn’t scientific, so the results might be judged
in several ways.
• Outdated systems can overlook various factors and thus make incorrect predictions about certain borrowers.
• Financial losses due to the failure of a credit risk model.
• A long period of time between a loan application, its approval, and issuance.
• Credit scoring models may provide completely different scoring results, complicating the lending process.
• The cost and work of the majority of credit scoring models are questionable.
Challenges To Successful Credit
Risk model
02
04
Inefficient Data Management
Limited Group-Wide Risk Modeling
Infrastructure
Credit risk management solutions require the ability to
securely store, categorize and search data based on a
variety of criteria.
Sometimes it’s not enough to examine the risk qualities
posed by a single entity—a broad, comprehensive view
of all risk measures as seen from above is key to
understanding the risk posed by a new borrower to the
group
Lacking Risk Tools
AccessibleLess-than-intuitive
Reporting and Visualization
Identifying portfolio concentrations or re-grade portfolios is
essential to ensure you’re seeing the big picture. A
comprehensive risk assessment scorecard should be able
to quickly and clearly identify strengths and weaknesses
associated with a loan.
Cista can be used across multiple devices such
as mobile gadgets and laptop computers.Implementing
smart data modeling for decisioning and leveraging
alternative data sources throughout your credit
application processing system is a holistic approach to
risk analytics.
Optimization in Credit
Risk Modeling
Mathematical optimization is a class of methods
that are often used together with predictive models.
In credit risk modeling, while the predictive models
estimate the probability of default or prepayment,
optimization methods can be used to find the best
mix of credit products in a portfolio to keep risk low
while achieving a high expected return.
Conclusion
● Credit risk modeling is a technique used by lenders to determine the level of
credit risk associated with extending credit to a borrower.
● Credit risk analysis models can be based on either financial statement
analysis, default probability, or machine learning.
● High levels of credit risk can impact the lender negatively by increasing
collection costs and disrupting the consistency of cash flows.
Thank you
Related documents
Download