Credit Risk Analysis Project - Calculating LGD

0.0(0)
studied byStudied by 0 people
0.0(0)
full-widthCall with Kai
learnLearn
examPractice Test
spaced repetitionSpaced Repetition
heart puzzleMatch
flashcardsFlashcards
GameKnowt Play
Card Sorting

1/10

encourage image

There's no tags or description

Looks like no tags are added yet.

Study Analytics
Name
Mastery
Learn
Test
Matching
Spaced
Call with Kai

No study sessions yet.

11 Terms

1
New cards

Loss Given Default

Amount of money the bank loses when the borrower defaults

2
New cards

LGD Forwards calculation

LGD = [OAD(money owed at default) - Recovery(amount company chooses to pay back to the bank) + Costs(third party for collection, accountant, etc.)] /OAD (forwards calculation)

3
New cards

What is LGD classified according to/based on?

Assets: real estate, cash, government guarantees - collaterals are given at the time the money is borrowed

4
New cards

Order of Quality of collateral type

Cash, Gov Guarantees, real assets, fixed assets (Fixed assets are bad because when the companies default, fixed assets depreciate or are gone based on TVM)

5
New cards

Exposure at default(EAD)

estimated amount of money from company owed to the bank at default

6
New cards

What is recovery determined by

daily outstanding change(change in money owed)

7
New cards

Used and How?

used python to calculate recovery by taking daily outstanding balance and subtracting the second days outstanding balance from the first to get cash flow, then calculate present value of cashflow and use recovery in LGD formula

8
New cards

Write off

If the collection department decides(based on estimation) that the company will not return the money owed

9
New cards

Project Challenge/Objective

Challenge: Backwards calculation for loss given default does not equal the forwards calculation for loss given default.

Objective: Develop a python application that derives transaction level cash flows and computes the realized loss given default, back calculates loss given default from write-offs, compares those values, then identifies the error in calculation that causes backwards and forwards calculation of LGD to be different.

10
New cards

Common reasons for inequality in forwards and backwards calculation

Data-quality gap: Orphan balance - last recorded outstanding balance is not equal to zero like it is supposed to if no more money is owed.

Solution: cross-track different systems to look for orphan balance(ex. business comments, specifically account manager comments)

Product Linkage:

For example, if you have a credit card and a term loan, during collection process, credit card is suspended because obviously the bank will not continue allowing companies to accumulate credit card debt if they already defaulted. If you owe 1M in credit card debt and 10M in term loan debt, your new term loan debt is 11M. When paying back term loan at default(10M), all the recovery(11M) is paid to the term loan debt which would make LGD < 0 because it looks like the bank is being overpaid by 1M. However, we forget that 1M is actually credit card debt, so LGD for credit card is 100%. Because credit card is not properly linked to the term loan, meaning you can’t see it in the system, this results in a product linkage error.

Solution: Contact business department to properly link the 2 products into the system, redistribute the recovery into each product,

11
New cards

Impact of project

Identified issue or orphan balance/product linkage error and fixed those issues in the system to ensure all debts are properly recorded and paid.