CECL Initial and Subsequent Measurement

Transcription

1 CECL Initial and Subsequent Measurement About the Presenter Neekis Hammond, CPA Advisory Services

2 Implementation Timelines. 1. SEC Filing Institutions.

3 Implementation Timelines. 2. Non-SEC Filing Public Business Entities.

4 Implementation Timelines. INTERIM 3. All Other Entities + Not-For-Profit Institutions.

5 Agenda. Segmentation Life-of-Loan Methodology Forecasting Q&A

6 Segmentation. AGENDA Segmentation Life-of-Loan Methodology Forecasting Q&A Proper Segmentation ( ) Segmentations or pools should have similar risk characteristics. These pools should be as granular as possible while maintaining statistical significance. Management will need to evaluate pools on an ongoing basis to ensure that the underlying assets continue to exhibit similar risk behavior.

7 Segmentation. AGENDA Segmentation Life-of-Loan Methodology Forecasting Q&A Proper Segmentation ( ) Segmentations or pools should have similar risk characteristics. These pools should be as granular Granular as possible while maintaining statistical significance. Management will need to evaluate pools on an ongoing basis to ensure that the underlying assets continue to exhibit similar risk behavior.

8 Segmentation. AGENDA Segmentation Life-of-Loan Methodology Forecasting Q&A Proper Segmentation ( ) Segmentations or pools should have similar risk characteristics. These pools should be as granular Granular as possible while maintaining statistical Statistical significance. Significance Management will need to evaluate pools on an ongoing basis to ensure that the underlying assets continue to exhibit similar risk behavior.

9 Segmentation. AGENDA Segmentation Life-of-Loan Methodology Forecasting Q&A Proper Segmentation ( ) Segmentations or pools should have similar risk characteristics. These pools should be as granular Granular as possible while maintaining statistical Statistical significance. Significance Management will need to evaluate pools on an ongoing basis to ensure that the underlying assets continue to exhibit similar risk behavior. Continue to exhibit similar risk behavior

10 Segmentation. 10% 8% 6% 4% 2% 0%

11 Segmentation. 10% 8% 6% Loss Rate 4% 2% 0%

12 Segmentation. 10% 8% 6% Loss Rate 4% 2% 0% Risk Rating

13 Segmentation. 10% 8% 6% Loss Rate 4% 2% 0% Risk Rating

14 Segmentation. 10% 8% 6% Loss Rate 4% 2% 0% Risk Rating

15 Segmentation Example. 1-4 Family First Lien Loan Count Loan Balance Loss Rate Estimated Losses Total 2, ,000, % 6,472,500 Pass 1, ,000, % 4,712,500 Special Mention 30 2,000, % 10,000 Substandard 340 5,000, % 1,750,000 Doubtful % - Loss % -

16 Segmentation Example. 1-4 Family First Lien Loan Count Loan Balance Loss Rate Estimated Losses Total 2, ,000, % 6,665,000 Pass 1, ,000, % 4,712,500 Special Mention 30 2,000, % 10,000 Substandard 340 5,000, % 1,750,000 Doubtful % - Loss % -

17 Segmentation Example. 1-4 Family First Lien Loan Count Loan Balance Loss Rate Estimated Losses Total 2, ,000, % 3,658, % , % - 3 1, ,000, % 1,260, ,500, % 398, ,000, % 2,000, ,500, % 7, ,500, % 1,225, % % -

18 Segmentation Example. 1-4 Family First Lien Loan Count Loan Balance Loss Rate Estimated Losses Total 2, ,000, % 3,658, % , % - 3 1, ,000, % 1,260, ,500, % 398, ,000, % 2,000, ,500, % 7, ,500, % 1,225, % % -

19 Segmentation Example. 1-4 Family First Lien Loan Count Loan Balance Loss Rate Estimated Losses Total 2, ,000, % 3,658, % , % - 3 1, ,000, % 1,260, ,500, % 398, ,000, % 2,000, ,500, % 7, ,500, % 1,225, % % -

20 Segmentation Example. 1-4 Family First Lien Loan Count Loan Balance Loss Rate Estimated Losses Total 2, ,000, % 3,658, % , % - 3 1, ,000, % 1,260, ,500, % 398, ,000, % 2,000, ,500, % 7, ,500, % 1,225, % % -

21 Segmentation Example. 1-4 Family First Lien Loan Count Loan Balance Loss Rate Estimated Losses Total 2, ,000, % 6,472, Family First Lien Loan Count Loan Balance Loss Rate Estimated Losses Total 2, ,000, % 3,658,750

22 Segmentation Example. 1-4 Family First Lien Loan Count Loan Balance Loss Rate Estimated Losses Total 2, ,000, % 6,472, Family First Lien Loan Count Loan Balance Loss Rate Estimated Losses Total 2, ,000, % 3,658,750 Risk Level Sub-segmentation» Pass cohort too consolidated.» Watchlist loans did not share risk characteristics with other Pass rated credits.

23 Segmentation Example. 1-4 Family First Lien Loan Count Loan Balance Loss Rate Estimated Losses Total 2, ,000, % 6,472, Family First Lien Loan Count Loan Balance Loss Rate Estimated Losses Total 2, ,000, % 3,658,750 Risk Rating Sub-segmentation» Pass cohort less consolidated.» Watchlist loans no longer impacting higher quality Pass rated credits.» Risk rating sub-segmentation still requires evaluation.

24 Agenda. Segmentation Life-of-Loan Methodology Forecasting Q&A

25 Life-of-Loan. AGENDA Segmentation Life-of-Loan Methodology Forecasting Q&A Determining the Expected Life of Each Segment ( ) Entities are required to estimate expected credit losses over the contractual term of the financial asset(s). Prepayments will need to be considered as a separate input or embedded in the credit loss experience. Expected life is a critical component of all methodologies used to determine loss experience. Prepayment and/or mortality rates will provide for increased flexibility and defensibility.

26 Life-of-Loan. AGENDA Segmentation Life-of-Loan Methodology Forecasting Q&A Determining the Expected Life of Each Segment ( ) Entities are required to estimate expected credit losses over the contractual contractual term of the financial term asset(s). Prepayments will need to be considered as a separate input or embedded in the credit loss experience. Expected life is a critical component of all methodologies used to determine loss experience. Prepayment and/or mortality rates will provide for increased flexibility and defensibility.

27 Life-of-Loan. AGENDA Segmentation Life-of-Loan Methodology Forecasting Q&A Determining the Expected Life of Each Segment ( ) Entities are required to estimate expected credit losses over the contractual contractual term of the financial term asset(s). Prepayments will need to be considered as a separate input or embedded in the credit loss experience. Expected life is a critical component of all methodologies used to determine loss experience. Prepayments Prepayment and/or mortality rates will provide for increased flexibility and defensibility.

28 Life-of-Loan Calculation. Product Type Life of Pool Commercial RE 4.90 Commercial/Ag 1.95 Consumer - Auto 2.58 Farm RE 5.27 HELOC 5.54 RE Construction 3.31 RE Mortgage 6.27 Grand Total 3.63

29 Life-of-Loan Calculation. Product Type Attrition Active Commercial RE 417 7,514 Commercial/Ag 1,095 6,673 Consumer - Auto 1,222 10,572 Farm RE 76 1,483 HELOC ,753 RE Construction RE Mortgage ,599 Grand Total 3,971 51,311

30 Life-of-Loan Calculation. Product Type Attrition Active Attrition Commercial RE 417 7,514 6% Commercial/Ag 1,095 6,673 16% Consumer - Auto 1,222 10,572 12% Farm RE 76 1,483 5% HELOC ,753 5% RE Construction % RE Mortgage ,599 4% Grand Total 3,971 51,311 8%

31 Life-of-Loan Calculation. Product Type Attrition Active Attrition Annual Rate Commercial RE 417 7,514 6% 20% Commercial/Ag 1,095 6,673 16% 51% Consumer - Auto 1,222 10,572 12% 39% Farm RE 76 1,483 5% 19% HELOC ,753 5% 18% RE Construction % 30% RE Mortgage ,599 4% 16% Grand Total 3,971 51,311 8% 28%

32 Life-of-Loan Calculation. Product Type Commercial RE Commercial/Ag Consumer - Auto Farm RE HELOC RE Construction RE Mortgage Grand Total Annual Rate Life of Pool 20% % % % % % % % 3.63

33 Life-of-Loan Impact. Commercial RE Loan Count Loan Balance Loss Rate Estimated Losses Total 1, ,500, % 6,752,500 Pass ,000, % 5,820,000 Special Mention 25 8,500, % 212,500 Substandard 150 6,000, % 720,000

34 Life-of-Loan Impact. Commercial RE Loan Count Loan Balance Loss Rate Estimated Losses Total 1, ,500, % 6,752,500 Pass ,000, % 5,820,000 Special Mention 25 8,500, % 212,500 Substandard 150 6,000, % 720,000 Commercial RE Loan Count Loan Balance Loss Rate Estimated Losses Total 1, ,500, % 4,115,950 Pass ,000, % 3,395,000 Special Mention 25 8,500, % 90,950 Substandard 150 6,000, % 630,000

35 Life-of-Loan Impact. Commercial RE Loan Count Loan Balance Loss Rate Estimated Losses Total 1, ,500, % 6,752,500 Pass ,000, % 5,820,000 Special Mention 25 8,500, % 212,500 Substandard 150 6,000, % 720,000 Commercial RE WAM Loan Count Loan Balance Loss Rate Estimated Losses WAL Total 1, ,500, % 4,115,950 Pass ,000, % 3,395,000 Special Mention 25 8,500, % 90,950 Substandard 150 6,000, % 630,000

36 Life-of-Loan Impact. Commercial RE Loan Count Loan Balance Loss Rate Estimated Losses Total 1, ,500, % 6,752,500 Pass ,000, % 5,820,000 Prepayments will need to be considered as a separate input or embedded in the credit loss experience. (ASC ) Special Mention 25 8,500, % 212,500 Substandard 150 6,000, % 720,000 Commercial RE Loan Count Loan Balance Loss Rate Estimated Losses Total 1, ,500, % 4,115,950 Pass ,000, % 3,395,000 Special Mention 25 8,500, % 90,950 Substandard 150 6,000, % 630,000

37 Life-of-Loan Impact. Commercial RE Loan Count Loan Balance Loss Rate Estimated Losses Total 1, ,500, % 6,752,500 Pass ,000, % 5,820,000 Calculating, documenting, and properly utilizing prepayment data is a critical component of the new standard. Special Mention 25 8,500, % 212,500 Substandard 150 6,000, % 720,000 Commercial RE Loan Count Loan Balance Loss Rate Estimated Losses Total 1, ,500, % 4,115,950 Pass ,000, % 3,395,000 Special Mention 25 8,500, % 90,950 Substandard 150 6,000, % 630,000

38 Agenda. Segmentation Life-of-Loan Methodology Forecasting Q&A

39 Methodology. AGENDA Segmentation Life-of-Loan Methodology Forecasting Q&A Calculating Historical Loss Experience ( , 7-8) Understanding and documenting how each segment or pool reacts using various methodologies will yield more meaningful reserve levels and provide management, auditors and regulators with confidence that proper due diligence was performed. Some methodologies contain inherent limitations that limit utilization for some loan pools.

40 Methodology. AGENDA Segmentation Life-of-Loan Methodology Forecasting Q&A Calculating Historical Loss Experience ( , 7-8) Understanding and documenting how each segment or pool reacts using various methodologies will yield more meaningful reserve levels and provide management, auditors and regulators with confidence that proper due diligence was performed. Some methodologies contain inherent limitations that limit utilization for some loan pools.

41 Methodology. AGENDA Segmentation Life-of-Loan Methodology Forecasting Q&A Calculating Historical Loss Experience ( , 7-8) Understanding and documenting how each segment or pool reacts using various methodologies will yield more meaningful reserve levels and provide management, auditors and regulators with confidence that proper due diligence was performed. Some methodologies contain inherent limitations Inherent limitations that limit utilization for some loan pools.

42 Methodology. Consumer - Auto Loan Count Loan Balance Loss Rate Estimated Losses Total 2,250 9,850, % 195,375 Pass 2,000 9,500, % 171,000 Special Mention , % 14,375 Substandard , % 10,000

43 Methodology. Consumer - Auto Loan Count Loan Balance Loss Rate Estimated Losses Total 2,250 9,850, % 195,375 Pass 2,000 9,500, % 171,000 Special Mention , % 14,375 Substandard , % 10,000 Consumer - Auto Loan Count Loan Balance Loss Rate Estimated Losses Total (Vintage) 2,250 9,850, % 118,002

44 Methodology. Consumer - Auto Loan Count Loan Balance Loss Rate Estimated Losses Total 2,250 9,850, % 195,375 Pass 2,000 9,500, % 171,000 Special Mention , % 14,375 Substandard , % 10,000 Consumer - Auto Loan Count Loan Balance Loss Rate Estimated Losses Total (Vintage) 2,250 9,850, % 118,002

45 Vintage Analysis. Origination Year Total Loss Remaining Loss 12/31/ % 1.10% 0.60% 0.25% 0.01% 0.03% 2.39% 0.00% 12/31/ % 0.85% 1.25% 0.40% 0.06% 0.03% 2.84% 0.03% 12/31/ % 0.95% 0.75% 0.06% 0.04% 0.03% 3.08% 0.07% 12/31/ % 0.65% 0.40% 0.24% 0.04% 0.03% 2.10% 0.30% 12/31/ % 0.60% 0.75% 0.24% 0.04% 0.03% 1.85% 1.05% 12/31/ % 0.83% 0.75% 0.24% 0.04% 0.03% 2.18% 1.88% Average 0.53% 0.83% 0.75% 0.24% 0.04% 0.03% 2.41%

46 Vintage Analysis. Loss Rate Year Subsequent to Origination Origination Year Total Loss Remaining Loss 12/31/ % 1.10% 0.60% 0.25% 0.01% 0.03% 2.39% 0.00% 12/31/ % 0.85% 1.25% 0.40% 0.06% 0.03% 2.84% 0.03% 12/31/ % 0.95% 0.75% 0.06% 0.04% 0.03% 3.08% 0.07% 12/31/ % 0.65% 0.40% 0.24% 0.04% 0.03% 2.10% 0.30% 12/31/ % 0.60% 0.75% 0.24% 0.04% 0.03% 1.85% 1.05% 12/31/ % 0.83% 0.75% 0.24% 0.04% 0.03% 2.18% 1.88% Average 0.53% 0.83% 0.75% 0.24% 0.04% 0.03% 2.41%

47 Vintage Analysis. Loss Rate Year Subsequent to Origination Origination Year Total Loss Remaining Loss 12/31/ % 1.10% 0.60% 0.25% 0.01% 0.03% 2.39% 0.00% 12/31/ % 0.85% 1.25% 0.40% 0.06% 0.03% 2.84% 0.03% 12/31/ % 0.95% 0.75% 0.06% 0.04% 0.03% 3.08% 0.07% Remaining loss estimate 12/31/ % 0.65% 0.40% 0.24% 0.04% 0.03% 2.10% 0.30% 12/31/ % 0.60% 0.75% 0.24% 0.04% 0.03% 1.85% 1.05% 12/31/ % 0.83% 0.75% 0.24% 0.04% 0.03% 2.18% 1.88% Average 0.53% 0.83% 0.75% 0.24% 0.04% 0.03% 2.41%

48 Vintage Analysis. Loss Rate Year Subsequent to Origination Origination Year Total Loss Remaining Loss 12/31/ % 1.10% 0.60% 0.25% 0.01% 0.03% 2.39% 0.00% 12/31/ % 0.85% 1.25% 0.40% 0.06% 0.03% 2.84% 0.03% 12/31/ % 0.95% 0.75% 0.06% 0.04% 0.03% 3.08% 0.07% 12/31/ % 0.65% 0.40% 0.24% 0.04% 0.03% 2.10% 0.30% 12/31/ % 0.60% 0.75% 0.24% 0.04% 0.03% 1.85% 1.05% Loss curve 12/31/ % 0.83% 0.75% 0.24% 0.04% 0.03% 2.18% 1.88% Average 0.53% 0.83% 0.75% 0.24% 0.04% 0.03% 2.41%

49 Vintage Analysis. Loss Rate Application Origination Year Remaining Balance Loss Rate Remaining Balance 12/31/ , % - 12/31/ , % 30 12/31/ , % /31/2013 1,500, % 4,525 12/31/2014 2,600, % 27,343 12/31/2015 4,550, % 85,616 Total 9,850, % 118,002

50 Vintage Analysis. Loss Rate Application Origination Year Remaining Balance Loss Rate Remaining Balance 12/31/ , % - 12/31/ , % 30 12/31/ , % /31/2013 1,500, % 4,525 12/31/2014 2,600, % 27,343 12/31/2015 4,550, % 85,616 Total 9,850, % 118,002

51 Methodology. Consumer - Auto Loan Count Loan Balance Loss Rate Estimated Losses Total 2,250 9,850, % 195,375 Pass 2,000 9,500, % 171,000 Special Mention , % 14,375 Substandard , % 10,000 Consumer - Auto Loan Count Loan Balance Loss Rate Estimated Losses Total (Vintage) 2,250 9,850, % 105,395

52 Methodology. Consumer - Auto Loan Count Loan Balance Loss Rate Estimated Losses Total 2,250 9,850, % 195,375 Understanding Pass and 2,000 documenting 9,500,000 how 1.80% each segment 171,000 or pool Special reacts Mentionusing various 150 methodologies 250, % will yield 14,375 more Substandard meaningful reserve 100 levels. 100, % 10,000 Consumer - Auto Loan Count Loan Balance Loss Rate Estimated Losses Total (Vintage) 2,250 9,850, % 118,002

53 Methodology. 3% 2% 2% 1% 1% 0% C & I Consumer CRE Mortgage Migration PD & LGD Vintage Cumulative

54 Methodology. Loss Rate 3% 2% 2% 1% 1% 0% C & I Consumer CRE Mortgage Migration PD & LGD Vintage Cumulative

55 Methodology. Loss Rate 3% 2% 2% 1% 1% 0% C & I Consumer CRE Mortgage Migration PD & LGD Vintage Cumulative Methodology

56 Methodology - Considerations. DISCOUNTED CASH FLOW

57 Methodology - Considerations. DISCOUNTED CASH FLOW + Flexible loss application and forecasting (loss curve)

58 Methodology - Considerations. DISCOUNTED CASH FLOW + Flexible loss application and forecasting (loss curve) + Less historical data required

59 Methodology - Considerations. DISCOUNTED CASH FLOW + Flexible loss application and forecasting (loss curve) + Less historical data required - Technical and data-heavy

60 Methodology - Considerations. MIGRATION, CUMULATIVE, AND PD/LGD

61 Methodology - Considerations. MIGRATION, CUMULATIVE, AND PD/LGD + Valuable portfolio risk management insight

62 Methodology - Considerations. MIGRATION, CUMULATIVE, AND PD/LGD + Valuable portfolio risk management insight + Valuable loan/product pricing insight

63 Methodology - Considerations. MIGRATION, CUMULATIVE, AND PD/LGD + Valuable portfolio risk management insight + Valuable loan/product pricing insight - Historical data required

64 Methodology - Considerations. VINTAGE

65 Methodology - Considerations. VINTAGE + Loss curve and declining balance is inherent in the calculation

66 Methodology - Considerations. VINTAGE + Loss curve and declining balance is inherent in the calculation + Valuable time specific and strategic insights

67 Methodology - Considerations. VINTAGE + Loss curve and declining balance is inherent in the calculation + Valuable time specific and strategic insights - Revolving loans and CRE renewals create heavy logistical and relevancy concerns

68 Agenda. Segmentation Life-of-Loan Methodology Forecasting Q&A

69 Forecasting. AGENDA Segmentation Life-of-Loan Methodology Forecasting Q&A Reasonable and Supportable Forecasts ( ) The adjustments to historical loss information may be qualitative in nature and should reflect changes related to relevant data (such as changes in unemployment rates, property values, commodity values, delinquency, or other factors that are associated with credit losses on the financial asset or in the group of financial assets).

70 Forecasting. AGENDA Segmentation Life-of-Loan Methodology Forecasting Q&A Reasonable and Supportable Forecasts ( ) The adjustments to historical loss information may be qualitative in nature and should reflect changes related to relevant data (such as changes in unemployment rates, property values, commodity values, delinquency, or other factors that are associated with credit losses on the financial asset or in the group of financial assets). should reflect changes to related relevant data

71 Forecasting. AGENDA Segmentation Life-of-Loan Methodology Forecasting Q&A Reasonable and Supportable Forecasts ( ) The adjustments to historical loss information may be qualitative in nature and should reflect changes related to relevant data (such as changes in unemployment rates, property values, commodity values, delinquency, or other factors that are associated with credit losses on the financial asset or in the group of financial assets). should reflect changes to related relevant data unemployment

72 Forecasting. AGENDA Segmentation Life-of-Loan Methodology Forecasting Q&A Reasonable and Supportable Forecasts ( ) The adjustments to historical loss information may be qualitative in nature and should reflect changes related to relevant data (such as changes in unemployment rates, property values, commodity values, delinquency, or other factors other that factors are associated associated with credit losses with on the losses financial asset or in the group of financial assets). should reflect changes to related relevant data unemployment

73 Forecasting. AGENDA Segmentation Life-of-Loan Methodology Forecasting Q&A Reasonable and Supportable Forecasts ( ) An entity shall not rely solely on past events to estimate expected credit losses. When an entity uses historical loss information, it shall consider the need to adjust historical information to reflect the extent to which management expects current conditions and reasonable and supportable forecasts to differ from the conditions that existed for the period over which historical information was evaluated.

74 Forecasting. AGENDA Segmentation Life-of-Loan Methodology Forecasting Q&A Reasonable and Supportable Forecasts ( ) An entity shall not rely solely on past events to estimate expected credit losses. When an entity uses historical loss information, it shall consider the need to adjust historical information to adjust reflect the historical extent which information management expects current conditions and reasonable and supportable forecasts to differ from the conditions that existed for the period over which historical information was evaluated.

75 Forecasting. AGENDA Segmentation Life-of-Loan Methodology Forecasting Q&A Reasonable and Supportable Forecasts ( ) An entity shall not rely solely on past events to estimate expected credit losses. When an entity uses historical loss information, it shall consider the need to adjust historical information to adjust reflect the historical extent which information management expects current conditions and reasonable and supportable forecasts to differ from from the conditions the conditions that existed for that the period over which historical information was evaluated. existed

76 Forecasting - Application. Commercial RE Loan Count Loan Balance Loss Rate Estimated Losses Total 1, ,500, % 6,752,500 Pass ,000, % 5,820,000 Special Mention 25 8,500, % 212,500 Substandard 150 6,000, % 720,000 Commercial RE Loan Count Loan Balance Loss Rate Estimated Losses Total 1, ,500, % 4,115,950 Pass ,000, % 3,395,000 Special Mention 25 8,500, % 90,950 Substandard 150 6,000, % 630,000

77 Forecasting - Application. Include Static Date Balance Charge-offs Recoveries Loss Rate Yes 12/31/ ,000,000 3,000, , % Yes 3/31/ ,000,000 2,750, , % Yes 6/30/ ,000,000 3,500, , % Yes 9/30/ ,000,000 2,700, , % Yes 12/31/ ,000,000 2,300, , % Yes 3/31/ ,000,000 1,850, , % Yes 6/30/ ,000,000 1,850, , % Yes 9/30/ ,000,000 1,700, , % Yes 12/31/ ,000,000 1,400,000 55, %

78 Forecasting - Application. Include Static Date Balance Charge-offs Recoveries Loss Rate Yes 12/31/ ,000,000 3,000, , % Yes 3/31/ ,000,000 2,750, , % Yes 6/30/ ,000,000 3,500, , % Yes 9/30/ ,000,000 2,700, , % Yes 12/31/ ,000,000 2,300, , % Yes 3/31/ ,000,000 1,850, , % Yes 6/30/ ,000,000 1,850, , % Yes 9/30/ ,000,000 1,700, , % Yes 12/31/ ,000,000 1,400,000 55, %

79 Forecasting - Application. Include Static Date Balance Charge-offs Recoveries Loss Rate Yes 12/31/ ,000,000 3,000, , % Yes 3/31/ ,000,000 2,750, , % Yes 6/30/ ,000,000 3,500, , % Yes 9/30/ ,000,000 2,700, , % Yes 12/31/ ,000,000 2,300, , % Yes 3/31/ ,000,000 1,850, , % Yes 6/30/ ,000,000 1,850, , % Yes 9/30/ ,000,000 1,700, , % Yes 12/31/ ,000,000 1,400,000 55, %

80 Forecasting - Application. Include Static Date Balance Charge-offs Recoveries Loss Rate No 12/31/ ,000,000 3,000, , % Unemployment > 8% (exceeds current forecast) No 3/31/ ,000,000 2,750, , % No 6/30/ ,000,000 3,500, , % Yes 9/30/ ,000,000 2,700, , % Yes 12/31/ ,000,000 2,300, , % Yes 3/31/ ,000,000 1,850, , % Yes 6/30/ ,000,000 1,850, , % Yes 9/30/ ,000,000 1,700, , % Yes 12/31/ ,000,000 1,400,000 55, %

81 Forecasting - Application. Commercial RE Loan Count Loan Balance Loss Rate Estimated Losses Total 1, ,500, % 6,752,500 Example calculation No prepayments No forecasting

82 Forecasting - Application. Commercial RE Loan Count Loan Balance Loss Rate Estimated Losses Total 1, ,500, % 6,752,500 Commercial RE Loan Count Loan Balance Loss Rate Estimated Losses Total 1, ,500, % 4,115,950 Example calculation No forecasting

83 Forecasting - Application. Commercial RE Loan Count Loan Balance Loss Rate Estimated Losses Total 1, ,500, % 6,752,500 Commercial RE Loan Count Loan Balance Loss Rate Estimated Losses Total 1, ,500, % 4,115,950 Commercial RE Loan Count Loan Balance Loss Rate Estimated Losses Total 1, ,500, % 2,747,250 Example calculation Prepayments - Forecasting

84 Forecasting - Application. Commercial RE Loan Count Loan Balance Loss Rate Estimated Losses Total 1, ,500, % 6,752,500 Commercial RE Loan Count Loan Balance Loss Rate Estimated Losses Total 1, ,500, % 4,115,950 Commercial RE Loan Count Loan Balance Loss Rate Estimated Losses Total 1, ,500, % 2,747,250 Example calculation Prepayments - Forecasting

85 Forecasting - Regression. 5.00% Baseline Scenario 4.50% 4.00% 3.50% 3.00% 2.50% 2.00% 1.50% 1.00% 0.50% 0.00%

86 Forecasting - Regression. 5.00% 4.50% Baseline Scenario Actual C/O Experience 4.00% 3.50% 3.00% 2.50% 2.00% 1.50% 1.00% 0.50% 0.00%

87 Forecasting - Regression. 5.00% 4.50% Baseline Scenario Actual C/O Experience 4.00% 3.50% 3.00% 2.50% 2.00% W Avg. - Singular Regression 1.50% 1.00% 0.50% 0.00%

88 Forecasting - Regression. 5.00% 4.50% Baseline Scenario Actual C/O Experience 4.00% 3.50% Multi Regression 3.00% 2.50% 2.00% W Avg. - Singular Regression 1.50% 1.00% 0.50% 0.00%

89 Forecasting - Regression. 5.00% Adverse Scenario 4.50% 4.00% 3.50% 3.00% 2.50% 2.00% 1.50% 1.00% 0.50% 0.00%

90 0.00% 0.50% 1.00% 1.50% 2.00% 2.50% 3.00% 3.50% 4.00% 4.50% 5.00% Severely Adverse Scenario Forecasting - Regression.

91 Forecasting - Regression. External Factor (National Data) R Square Slope Lead (months) Real GDP growth 19.14% -0.24% 12 Nominal GDP growth 13.73% -0.17% 6 Real disposable income growth 2.89% -0.06% 6 Nominal disposable income growth 5.71% -0.08% 6 Unemployment rate 94.92% 0.78% 3 CPI inflation rate 3.28% -0.10% 6 3-month Treasury rate 41.29% -0.54% 3 5-year Treasury yield 36.22% -0.66% 6 10-year Treasury yield 18.55% -0.59% 6 BBB corporate yield 0.31% 0.07% 6 Mortgage rate 27.30% -0.74% 3 Prime rate 40.02% -0.51% 3 Dow Jones Total Stock Market Index (Level) 3.61% 0.00% 3 House Price Index (Level) 65.62% -0.06% 3 Commercial Real Estate Price Index (Level) 19.02% -0.02% 3 Market Volatility Index (Level) 40.78% 0.07% 12

92 Forecasting - Regression. External Factor (National Data) R Square Slope Lead (months) Real GDP growth 19.14% -0.24% 12 Nominal GDP growth 13.73% -0.17% 6 Real disposable income growth 2.89% -0.06% 6 Nominal disposable income growth 5.71% -0.08% 6 Unemployment rate 94.92% 0.78% 3 CPI inflation rate 3.28% -0.10% 6 3-month Treasury rate 41.29% -0.54% 3 5-year Treasury yield 36.22% -0.66% 6 10-year Treasury yield 18.55% -0.59% 6 BBB corporate yield 0.31% 0.07% 6 Mortgage rate 27.30% -0.74% 3 Prime rate 40.02% -0.51% 3 Dow Jones Total Stock Market Index (Level) 3.61% 0.00% 3 House Price Index (Level) 65.62% -0.06% 3 Commercial Real Estate Price Index (Level) 19.02% -0.02% 3 Market Volatility Index (Level) 40.78% 0.07% 12

93 Forecasting - Regression. External Factor (National Data) R Square Slope Lead (months) Real GDP growth 19.14% -0.24% 12 Nominal GDP growth 13.73% -0.17% 6 Real disposable income growth 2.89% -0.06% 6 Nominal disposable income growth 5.71% -0.08% 6 Unemployment rate 94.92% 0.78% 3 CPI inflation rate 3.28% -0.10% 6 3-month Treasury rate 41.29% -0.54% 3 5-year Treasury yield 36.22% -0.66% 6 10-year Treasury yield 18.55% -0.59% 6 BBB corporate yield 0.31% 0.07% 6 Mortgage rate 27.30% -0.74% 3 Prime rate 40.02% -0.51% 3 Dow Jones Total Stock Market Index (Level) 3.61% 0.00% 3 House Price Index (Level) 65.62% -0.06% 3 Commercial Real Estate Price Index (Level) 19.02% -0.02% 3 Market Volatility Index (Level) 40.78% 0.07% 12

94 Forecasting - Regression. External Factor (National Data) R Square Slope Lead (months) Real GDP growth 19.14% -0.24% 12 Regression Utilization Coefficients Nominal GDP growth 13.73% -0.17% 6 Real Intercept dispo-sable income growth 2.89% -0.06% -6.62% 6 Nominal dispo-sable income growth 5.71% -0.08% 6 Un-employment rate 0.93% Un-employ-ment rate 94.92% 0.78% 3 CPI 5-year inflation Treasury rate yield 3.28% -0.10% 0.01% 6 3-month Treasury rate 41.29% -0.54% 3 House Price Index (Level) 0.01% 5-year Treasury yield 36.22% -0.66% 6 Market Volatility Index (Level) 0.00% 10-year Treasury yield 18.55% -0.59% 6 BBB corporate yield 0.31% 0.07% 6 Mortgage rate 27.30% -0.74% 3 Prime rate 40.02% -0.51% 3 Dow Jones Total Stock Market Index (Level) 3.61% 0.00% 3 House Price Index (Level) 65.62% -0.06% 3 Com-mercial Real Estate Price Index (Level) 19.02% -0.02% 3 Market Volatility Index (Level) 40.78% 0.07% 12

95 Forecasting - Regression. External Factor (National Data) R Square Slope Lead (months) Real GDP growth 19.14% -0.24% 12 Regression Utilization How much Coefficients of the change in the Intercept dependent variable -6.62% can be explained Un-employment rate by each respective 0.93% independent 5-year Treasury yield (external) factor 0.01% House Price Index (Level) 0.01% Market Volatility Index (Level) 0.00% Nominal GDP growth 13.73% -0.17% 6 Real dispo-sable income growth 2.89% -0.06% 6 Nominal dispo-sable income growth 5.71% -0.08% 6 Un-employ-ment rate 94.92% 0.78% 3 CPI inflation rate 3.28% -0.10% 6 3-month Treasury rate 41.29% -0.54% 3 5-year Treasury yield 36.22% -0.66% 6 10-year Treasury yield 18.55% -0.59% 6 BBB corporate yield 0.31% 0.07% 6 Mortgage rate 27.30% -0.74% 3 Prime rate 40.02% -0.51% 3 Dow Jones Total Stock Market Index (Level) 3.61% 0.00% 3 House Price Index (Level) 65.62% -0.06% 3 Commercial Real Estate Price Index (Level) 19.02% -0.02% 3 Market Volatility Index (Level) 40.78% 0.07% 12

96 Forecasting - Regression. External Factor (National Data) R Square Slope Lead (months) Regression Utilization The impact Coefficients on the dependent Intercept variable given a -6.62% one unit change in Un-employment rate the independent 0.93% (external) factor 5-year Treasury yield 0.01% House Price Index (Level) 0.01% Market Volatility Index (Level) 0.00% Real GDP growth 19.14% -0.24% 12 Nominal GDP growth 13.73% -0.17% 6 Real dispo-sable income growth 2.89% -0.06% 6 Nominal dispo-sable income growth 5.71% -0.08% 6 Un-employ-ment rate 94.92% 0.78% 3 CPI inflation rate 3.28% -0.10% 6 3-month Treasury rate 41.29% -0.54% 3 5-year Treasury yield 36.22% -0.66% 6 10-year Treasury yield 18.55% -0.59% 6 BBB corporate yield 0.31% 0.07% 6 Mortgage rate 27.30% -0.74% 3 Prime rate 40.02% -0.51% 3 Dow Jones Total Stock Market Index (Level) 3.61% 0.00% 3 House Price Index (Level) 65.62% -0.06% 3 Commercial Real Estate Price Index (Level) 19.02% -0.02% 3 Market Volatility Index (Level) 40.78% 0.07% 12

97 Forecasting - Regression. External Factor (National Data) R Square Slope Lead (months) Real GDP growth 19.14% -0.24% 12 Regression Utilization Coefficients Intercept -6.62% Un-employment rate 0.93% 5-year Treasury yield 0.01% House Price Index (Level) 0.01% Market Volatility Index (Level) 0.00% Nominal GDP growth 13.73% -0.17% 6 Real dispo-sable income growth 2.89% -0.06% 6 Nominal dispo-sable income growth 5.71% -0.08% 6 Un-employ-ment rate 94.92% 0.78% 3 CPI inflation rate 3.28% -0.10% 6 3-month Treasury rate 41.29% -0.54% 3 5-year Treasury yield 36.22% -0.66% 6 10-year Treasury yield 18.55% -0.59% 6 BBB corporate yield 0.31% 0.07% 6 Mortgage rate 27.30% -0.74% 3 Estimated C/O = Intercept + (Coefficient * Underlying Variable) Prime rate 40.02% -0.51% 3 Dow Jones Total Stock Market Index (Level) 3.61% 0.00% 3 House Price Index (Level) 65.62% -0.06% 3 Commercial Real Estate Price Index (Level) 19.02% -0.02% 3 Market Volatility Index (Level) 40.78% 0.07% 12

98 Agenda. Segmentation Life-of-Loan Methodology Forecasting Q&A