revanthnemani/lgd-engine

LGD Engine

LGD Engine

• LGD Engine
  • Add Renviron file
  • API Documentation
  • LGD models and methodologies
    • Introduction
    • Corporate and SME LGD model
    • Retail (excluding Credit Cards) LGD model
    • Credit Cards LGD model
    • Banks and Sovereigns LGD model
    • Conclusion

Add Renviron file

env="dev"

dev.host="0.0.0.0"
dev.port=3005
dev.secret="Secret"

test.host="0.0.0.0"
test.port=3006
test.secret="Secret"

prod.host="0.0.0.0"
prod.port=3010
prod.secret="Secret"

API Documentation

https://documenter.getpostman.com/view/9006400/SzzobbXD?version=latest

LGD models and methodologies

Introduction

Loss given default (LGD) is the amount of money we as a bank lose when a borrower defaults on a loan, depicted as a percentage of total exposure at the time of default.

The simple formula for LGD is as follows:

$$LGD = 1 - Recovery\ Rate$$

So the goal is to arrive at the recovery rates for each contract we have.

In case we have LGD lower than 10%, we have taken 10% as the floor rate as prescribed by BASEL II paragraph 266. Even though this is only for retail exposures secured by residential properties, we have applied this floor rate across the portfolio when our LGD estimates fall below 10% as a prudential step.

So at alizz islamic bank our LGD formula is as follows in most cases:

$$LGD = max(1 - Recovery\ Rate, 10%)$$

Corporate and SME LGD model

Most of our Corporate and SME customers have collaterals in place which could be used to recover our exposures as a last resort. Due to lack of enough past recovery data all models involving techniques like regression would fail in this segment at our bank. So we have taken a logical and expert judgement route to arrive at recovery rate for each contract in our portfolio. Care has been taken to keep it simple and mathematical at every chance.

We have observed that it would take on average three (3) years to recover our exposures after default of an account incase they are not cured. So we are using a discounting period of 3 years to arrive at the discounted useable collateral value (after haircut).

At alizz islamic bank we recognize the following collaterals for the purposes of recovery/LGD. Their corresponding $1-Haircut%$ are also listed in the table below:

Recognised Collateral Types	1-Haircut	Description
Building	70%	Building
Land	80%	Land
Deposit	100%	Cash Deposit
MotorVehicle	50%	Motor Vehicle
PersonalGuarantees	70%	Personal Guarantees on exposure
CorporateGuarantee	50%	Corporate Guarantees on exposure
LocalGovtGuarantee	80%	Local Government Guarantees on exposure
GeneralPlantMachinery	50%	General Plant and machinery
QuotedShares	70%	Listed Shares
NotQuotedShares	50%	Non Listed Shares

After applying haircuts we arrive at a customer level useable collateral values.

Eg: A customer has pledged 2 collaterals. One is Land with a valuation of OMR 1,000,000 and the second is a Building with a valuation of OMR 500,000 for all loans he has taken from alizz islamic bank. so the useable collateral value is as follows:

$$Useable\ collateral\ value = \ (OMR\ \ 1,000,000\ *\ 80%) + (OMR\ \ 500,000\ *\ 70%)$$

$$Useable\ collateral\ value = OMR\ \ 1,150,000$$

Once we have the customer level useable collateral value, we apportion it to all the customer's contracts on a pro rata basis based on their Exposure at Default (EAD).

We then arrive at the recovery amount. Recovery is the minimum of useable collateral value and Exposure at Default (EAD).

$$Recovery\ Amount = min(EAD, useable\ collateral\ value)$$

This recovery amount is then discounted for 3 years using the contract's Effective Interest Rate.

$$Discounted\ Recovery = \frac{Recovery\ Amount}{1+(\frac{EIR}{100})^{3}}$$

Our Recovery Rate is the percentage of discounted recovery on exposure at default:

$$Recovery\ Rate = \frac{Discounted\ Recovery}{EAD} *100%$$

And finally LGD:

$$LGD = max(1 - Recovery\ Rate, 10%)$$

Retail (excluding Credit Cards) LGD model

In retail we have fifty (50) products in place and for recovery amount we take a percentage of the Exposure at Default (EAD) based on the following table. This table has been made using expert judgement and industry best practices. Where industry best practices are not relevant or meeting our prudential needs, we have tweaked the useable collateral percentages to meet our prudential needs.

Retail Products	Product Code	Useable Collateral %
CL02-CL Goods Murabaha Finance- Retail	CL02	10%
CL03-CL Property Murabaha Finance-Retail	CL03	90%
CL41-CL DM Finance - Retail	CL41	10%
R101-Auto Murabaha Finance-Retail OC	R101	75%
R102-Goods Murabaha Financing -Retail-OC	R102	10%
R103-Property Murabaha Finance-Retail-OC	R103	90%
R111-Auto Murabaha Finance-Staff-OC	R111	75%
R112-Goods Murabaha Finance-Staff-OC	R112	10%
R114-Property Murabaha Finance - Staff	R114	90%
R201-Ijarah Finance - Retail -OC	R201	90%
R202-Forward Ijarah Finance -Retail-OC	R202	90%
R203-Service Ijarah Finance- Retail	R203	90%
R204-Ijarah Finance FXD_PRIN-Retail-OC	R204	90%
R211-Ijarah Finance - Staff -OC	R211	90%
R212-Forward Ijarah Finance -Staff-OC	R212	90%
R213-Service Ijarah Finance - Staff	R213	90%
R301-IjarahFinance H_Taveover-Retail-OC	R301	90%
R302-Ijarah Finance P_Takeover-Retail-OC	R302	90%
R311-Ijarah Finance H_Takeover Staff-OC	R311	90%
R312-Ijarah Finance P_Takeover-Staff-OC	R312	90%
R401-DM Property Finance-Retail	R401	90%
R404-DM Land Finance -Retail	R404	90%
R411-DM Property Finance - Staff	R411	90%
R412-DM Personal Takeover - Staff	R412	10%
R414-DM Land Finance -Staff	R414	90%
R420-DM Under Construction -Retail	R420	70%
R421-DM Under Construction -Staff	R421	70%
RF01-Ijarah Finance - Retail	RF01	90%
RF02-Forward Ijarah Finance - Retail	RF02	90%
RF11-Ijarah Finance - Staff	RF11	90%
RF12-Forward Ijarah Finance - Staff	RF12	90%
RF15-Ijarah Finance H_Takeover -Staff	RF15	90%
RF16-Ijarah Finance P_Takeover- Staff	RF16	90%
RM01-Auto Murabaha Finance Mora-Retail	RM01	75%
RM03-GoodsMurabaha Finance Mora-Retail	RM03	10%
RM21-Ijarah Finance _Mora- Staff	RM21	90%
RT01-Auto Murabaha Finance - Retail	RT01	75%
RT02-Goods Murabaha Finance - Retail	RT02	10%
RT03-Property Murabaha Finance-Retail	RT03	90%
RT11-Auto Murabaha Finance-Staff	RT11	75%
RT12-Goods Murabaha Finance- Staff	RT12	10%
RT13-Property Murabaha Finance -Staff	RT13	90%
CL01-CL Auto Finance - Retail	CL01	75%
CL21-CL Ijarah Finance - Retail	CL21	90%
CL42-DM Under Construction -CL	CL42	70%
CL43-DM Land Finance -CL	CL43	90%
RF05-IjarahFinance H_Takeover-Retail	RF05	90%
RF06-Ijarah Finance P_Takeover -Retail	RF06	90%
CL22-CL Forward Ijarah Finance -Retail	CL22	90%
R402-DM Personal Takeover -Retail	R402	10%

Eg: A retail customer has an Auto Murabaha, the recovery amount will be 75% of EAD.

$$Recovery\ Amount = EAD * Useable\ Collateral %$$

This recovery amount is then discounted for 3 years using the contract's Effective Interest Rate.

$$Discounted\ Recovery = \frac{Recovery\ Amount}{1+(\frac{EIR}{100})^{3}}$$

Our Recovery Rate is the percentage of discounted recovery on exposure at default:

$$Recovery\ Rate = \frac{Discounted\ Recovery}{EAD} *100%$$

And finally LGD:

$$LGD = max(1 - Recovery\ Rate, 10%)$$

Credit Cards LGD model

For credit cards we are using BASEL prescribed forty five percent (45%) LGD.

$$LGD = 45%$$

Banks and Sovereigns LGD model

Usually Banks and Sovereigns have low PDs (Probability of Defaults) and LGD (Loss Given Defaults), so we are using our floor rate LGD ten percent (10%).

$$LGD = 10%$$

Conclusion

To sum up:

Segment	LGD Formula
Corporate	$max(1 - Recovery\ Rate, 10%)$
SME	$max(1 - Recovery\ Rate, 10%)$
Retail	$max(1 - Recovery\ Rate, 10%)$
Credit Card	$45%$
Banks	$10%$
Sovereigns	$10%$

alizz islamic bank©2020. All rights reserved