How CECL Will Impact Your Credit Union & What You Can Do to - - PowerPoint PPT Presentation

1

How CECL Will Impact Your Credit Union & What You Can Do to Prepare For It: Part 2

Randy C Thompson, Ph.D. TCT Risk Solutions, LLC

Impact on ALLL Calculation

Baseline ALLL Calculation with Incurred Loss Application

Grade Balance Loss Factor ALLL Req. A+ $ 755,000 0.00% $ - A $ 649,000 0.00% $ - B $ 315,600 0.27% $ 852.12 C $ 126,000 0.00% $ - D $ 98,000 4.56% $ 4,468.80 E $ 57,000 6.54% $ 3,727.80 Missing $ 34,000 8.98% $ 3,053.20 $ 2,034,600 $ 12,101.92

Impact on ALLL Calculation

The first point is the impact on base line ALLL Calculation

• All loans must have a risk amount
• All grades must have an loss factor and ALLL

amount

• No zero loss factors

Impact on ALLL Calculation

• Same baseline Auto loan pool
• Expected Loss factors added

to all grades

Grade Balance Loss Factor ALLL Req. A+ $ 755,000 0.05% $ 377.50 A $ 649,000 0.15% $ 973.50 B $ 315,600 0.35% $ 1,104.60 C $ 126,000 0.79% $ 995.40 D $ 98,000 5.65% $ 5,537.00 E $ 57,000 9.87% $ 5,625.90 Missing $ 34,000 10.25% $ 3,485.00 $ 2,034,600 $ 18,098.90

Impact on ALLL Calculation

The initial impact is noticeable Baseline Incurred requirement $ 12,101.92 Baseline Expected Loss requirement $ 18,098.90 Increase requirement $ 5,996.98 Increase Percent 49.6%

Impact on ALLL Calculation

• Next financial impact to

consider relates to loan growth

• Lets assume this credit

union increases loans by $205,000

• No incurred losses in

the period

New Loans % of Growth A+ $ 50,000 24.39% A $ 50,000 24.39% B $ 50,000 24.39% C $ 25,000 12.20% D $ 10,000 4.88% E $ 10,000 4.88% Missing $ 10,000 4.88% $ 205,000 100.00%

Impact on ALLL Calculation

Steps 1. Increase each grade by indicated amount 2. Apply ALLL loss factor to each grade 3. Total new ALLL requirement

Impact on ALLL Calculation

New ALLL calculation for this pool

Grade Balance Loss Factor ALLL Req. A+ $ 805,000 0.00% $ - A $ 699,000 0.00% $ - B $ 365,600 0.27% $ 987.12 C $ 151,000 0.00% $ - D $ 108,000 4.56% $ 4,924.80 E $ 67,000 6.54% $ 4,381.80 Missing $ 44,000 8.98% $ 3,951.20 $ 2,239,600 $ 14,244.92

Impact on ALLL Calculation

Impact of loan growth with incurred loss model

Baseline Incurred requirement $ 12,101.92 Loan growth incurred requirement $ 14,244.92 Increase requirement $ 2,143.00 Increase Percent 17.7%

Impact on ALLL Calculation

Now calculate the impact with CECL

Grade Balance Loss Factor ALLL Req. A+ $ 805,000 0.05% $ 402.50 A $ 699,000 0.15% $ 1,048.50 B $ 365,600 0.35% $ 1,279.60 C $ 151,000 0.79% $ 1,192.90 D $ 108,000 5.65% $ 6,102.00 E $ 67,000 9.87% $ 6,612.90 Missing $ 44,000 10.25% $ 4,510.00 $ 2,239,600 $ 21,148.40

Impact on ALLL Calculation

Adjusting to CECL early minimizes the overall impact

Baseline Incurred requirement $ 12,101.92 Loan growth incurred requirement $ 21,148.40 Increase requirement $ 9,046.48 Increase Percent 74.8% Baseline CECL requirement $ 18,098.90 Loan growth CECL requirement $ 21,148.40 Increase requirement $ 3,049.50 Increase Percent 16.8%

Preparing for CECL

Build on

• Build on Credit

Migration Base

1

Identify

• Identify losses rates

2

Apply

• Apply to existing

methodology

3

Expected Credit Losses

The primary statistical tool for quantifying causality is Multiple Regression Regression is an analysis for estimating the causal relationship among variables Regression analysis is widely used for prediction and forecasting Focus is on the relationship between a dependent variable and

• ne or more

independent variables For our purposes we will consider that: Dependent variable = Expected Loan Losses Independent variable(s) = Those factors that are connected to (or predict) losses

13

Expected Credit Losses

Regression analysis identified the following factors as the primary predictors of impending loss Deterioration of credit score Advanced Delinquency Changing income or discretionary cash flow Changing economic conditions (relates to employment and cash flow) Loan to value (subject to existing risk change)

Expected Credit Losses

Formula for describing Expected Risk of Loss RT= (RC + GA) + (RIC + REC) Where: RT= Total Risk RC = Core Risk GA = Grade Adjustment RIC = Individual Risk Change REC = Economic Risk Change

15

Expected Credit Losses

Initial Loss Expectation based on:

• RC = Core Risk
• GA = Grade Adjustment

Should be addressed in Loan Pricing They provide a starting point for calculating expected credit loss at loan funding Implies that we are using a valid pricing model that quantifies loss risk in loan rates This formula repeats for each loan type addressing the CECL requirement regarding similar loan analysis

16

Expected Credit Losses

The remaining variables in the formula deal with changes 1. RIC = Individual Risk Change 2. REC = Economic Risk Change Movement of credit score, as indicated in credit migration provides the measurement of these variable.

17

Expected Credit Losses

A reminder of foundational considerations in CECL

• Similar loans
• Life of loan
• Probability of Loss
• Likelihood of loss or no loss
• Based on solid statistical analysis

18

Key Aspects of CECL Response

Initial Risk based

• n:
• Core Risk
• Grade Adjustment

These factors should be incorporated in Loan Pricing Establishes the starting point for any and all loans

19

How to Implement

• Step 1: Calculate historical loss model for “similar

loans”.

• Analyzed actual loss rates over an entire

economic cycle.

• Identified minimum loss and maximum loss
• Multiple distinct pools to match iterations of

loans

20

How to Implement

Step 1: Calculate historical loss model for “similar loans”.

• Sample Loan Pool – New Auto
• Isolate Credit Grade – A+
• Identify minimum loss factor – 0.001%
• Identify maximum loss factor – 0.091%

Repeat for each Credit Grade

21

How to Implement

• Step 2: Consider conditions that led to loss over life of loan.
• Conduct MANOVA to partition losses into sub-ranges for each grade
• Connect sub-ranges to economic condition
• Calculate mid-point for use in ALLL calculation

22

How to Implement

• Step 2: Consider conditions that led

to loss over life of loan.

• Sample Loan Pool – New Auto
• Range 0.001% to 0.091%
• Sub-ranges

23

1 2 3 4 5 6 0.001% 0.019% 0.037% 0.055% 0.073% 0.091%

How to Implement

• Step 3: Compare with present day conditions.
• Maintain a 3-5 year rolling average loan loss
• Connect actual loss to projected
• Consider all risk factors in ALLL calculation

24

• Step 3: Compare with present day conditions
• Sample Loan Pool – New Auto
• 5 Year Loss Ratio – 0.29%

How to Implement

25

1 2 3 4 5 6 0.001% 0.019% 0.037% 0.055% 0.073% 0.091%

How to Implement

• Step 4: Evaluate future conditions
• Conduct annual update of

population loss experience

• Update ranges and sub-ranges

for all pools

• Adjustments to historical loss

model can increase or decrease the overall allowance

26

Sample CECL Matrix

1 2 3 4 5 6 A+ 0.001% 0.019% 0.037% 0.055% 0.073% 0.091% A 0.003% 0.052% 0.102% 0.151% 0.201% 0.250% B 0.008% 0.144% 0.280% 0.416% 0.552% 0.688% C 0.021% 0.395% 0.769% 1.144% 1.518% 1.893% D 0.057% 2.002% 3.946% 5.891% 7.835% 9.780% E 0.157% 3.303% 6.448% 9.594% 12.739% 15.885% No Score 0.315% 5.977% 11.638% 17.300% 22.962% 28.624%

How to Implement

Adjustments to historical loss model can increase or decrease the

• verall allowance

Movement of credit scores may increase or decrease overall allowance

28

Starting point for sample loan

1 2 3 4 5 6 A+ 0.001% 0.019% 0.037% 0.055% 0.073% 0.091% A 0.003% 0.052% 0.102% 0.151% 0.201% 0.250% B 0.008% 0.144% 0.280% 0.416% 0.552% 0.688% C 0.021% 0.395% 0.769% 1.144% 1.518% 1.893% D 0.057% 2.002% 3.946% 5.891% 7.835% 9.780% E 0.157% 3.303% 6.448% 9.594% 12.739% 15.885% No Score 0.315% 5.977% 11.638% 17.300% 22.962% 28.624%

Credit grade drops to D

1 2 3 4 5 6 A+ 0.001% 0.019% 0.037% 0.055% 0.073% 0.091% A 0.003% 0.052% 0.102% 0.151% 0.201% 0.250% B 0.008% 0.144% 0.280% 0.416% 0.552% 0.688% C 0.021% 0.395% 0.769% 1.144% 1.518% 1.893% D 0.057% 2.002% 3.946% 5.891% 7.835% 9.780% E 0.157% 3.303% 6.448% 9.594% 12.739% 15.885% No Score 0.315% 5.977% 11.638% 17.300% 22.962% 28.624%

Expected loss increases to Range 3

1 2 3 4 5 6 A+ 0.001% 0.019% 0.037% 0.055% 0.073% 0.091% A 0.003% 0.052% 0.102% 0.151% 0.201% 0.250% B 0.008% 0.144% 0.280% 0.416% 0.552% 0.688% C 0.021% 0.395% 0.769% 1.144% 1.518% 1.893% D 0.057% 2.002% 3.946% 5.891% 7.835% 9.780% E 0.157% 3.303% 6.448% 9.594% 12.739% 15.885% No Score 0.315% 5.977% 11.638% 17.300% 22.962% 28.624%

Tracking the results of Down Migration

Balance Loss Factor ALLL $20,000 0.003% $1 $20,000 0.057% $11 $20,000 3.950% $790

Starting Point for New Loan

1 2 3 4 5 6 A+ 0.001% 0.019% 0.037% 0.055% 0.073% 0.091% A 0.003% 0.052% 0.102% 0.151% 0.201% 0.250% B 0.008% 0.144% 0.280% 0.416% 0.552% 0.688% C 0.021% 0.395% 0.769% 1.144% 1.518% 1.893% D 0.057% 2.002% 3.946% 5.891% 7.835% 9.780% E 0.157% 3.303% 6.448% 9.594% 12.739% 15.885% No Score 0.315% 5.977% 11.638% 17.300% 22.962% 28.624%

Economy improves to Grade 1

1 2 3 4 5 6 A+ 0.001% 0.019% 0.037% 0.055% 0.073% 0.091% A 0.003% 0.052% 0.102% 0.151% 0.201% 0.250% B 0.008% 0.144% 0.280% 0.416% 0.552% 0.688% C 0.021% 0.395% 0.769% 1.144% 1.518% 1.893% D 0.057% 2.002% 3.946% 5.891% 7.835% 9.780% E 0.157% 3.303% 6.448% 9.594% 12.739% 15.885% No Score 0.315% 5.977% 11.638% 17.300% 22.962% 28.624%

Loan grade improves to A

1 2 3 4 5 6 A+ 0.001% 0.019% 0.037% 0.055% 0.073% 0.091% A 0.003% 0.052% 0.102% 0.151% 0.201% 0.250% B 0.008% 0.144% 0.280% 0.416% 0.552% 0.688% C 0.021% 0.395% 0.769% 1.144% 1.518% 1.893% D 0.057% 2.002% 3.946% 5.891% 7.835% 9.780% E 0.157% 3.303% 6.448% 9.594% 12.739% 15.885% No Score 0.315% 5.977% 11.638% 17.300% 22.962% 28.624%

Tracking the results of Up Migration

Balance Loss Factor ALLL $20,000 3.950% $790 $20,000 0.057% $11 $20,000 0.003% $1

Applied Case Study

$1mil loans Starting FICO A+ New FICO D ALLL increases by $39,090 Balance Loss Factor ALLL $1,000,000 0.037% $370 $1,000,000 3.946% $39,460 Net Change $39,090

Applied Case Study

$ 1 mil loans Starting FICO E New FICO C ALLL decreases by $56,790

Balance Loss Factor ALLL $1,000,000 6.448% $64,480 $1,000,000 0.769% $7,690 Net Change $(56,790)

Applied Case Study

Effect of $1 mil in decreasing FICO scores Increases ALLL by $39,090

01

Effect of $1 mil in increasing FICO scores Decreases ALLL by $56,790

02

Net Effect of total score migration Decrease ALLL by $17,700*

03

*Note: This effect is post CECL adjustment

04

Sample CECL Matrix

1 2 3 4 5 6 A+ 0.001% 0.019% 0.037% 0.055% 0.073% 0.091% A 0.003% 0.052% 0.102% 0.151% 0.201% 0.250% B 0.008% 0.144% 0.280% 0.416% 0.552% 0.688% C 0.021% 0.395% 0.769% 1.144% 1.518% 1.893% D 0.057% 2.002% 3.946% 5.891% 7.835% 9.780% E 0.157% 3.303% 6.448% 9.594% 12.739% 15.885% No Score 0.315% 5.977% 11.638% 17.300% 22.962% 28.624%

How Can You Prepare for the Impact

Here are 3 simple steps you can take now to prepare for Expected Credit Loss

• 1. Implement a true

Risk Based methodology to allow for migration

• 2. Implement

procedures that support Credit Risk Management

• 3. Assure you have

statistically valid risk based pricing on loans

Support for CECL

“Credit Unions will need a third party provider to satisfy the requirements and implement CECL accurately.” Michael Richards, CP, Richard and Associates, CPAs

Summary

Choose the methodology that best fits your credit union

Employ sound statistical analyses to support your CECL implementation Understand the implications on your ALLL

Begin now to price for the impending implementation

Questions

• Randy C. Thompson, Ph.D.
• Email: rthompson@tctrisk.com
• Phone: (208) 939-8366
• Website: tctrisk.com