All of the Method None of the Madness

Transcription

1 CECL Simplified Current Expected Credit Losses ASU All of the Method None of the Madness The ProBank Austin 2017 Webinar Series November 9, :00 AM ESDT

2 Live Case Study - CECL I. Example Bank Historical Trends II. Loan Pool Segmentation Under CECL III. Example Loss Estimation Methods IV. Q-Factor Adjustment - Economy V. Determining the Allowance Amount 1

3 CECL Requirements 2 Expected loss Life of loan concept Advanced estimation methodologies Enhanced data collection Current conditions Reasonable and supportable forecasts Effective 2020 or 2021

4 Case Study Bank Historical Trends CECL Simplified All of the Method None of the Madness

5 2% 17% 12% 27% 81% 61% 3 Case Study Institution

6 Loan Composition over Time Retail 39% % 27% 1 4 FAM, $64,320 CRE, $91,563 AG PROD, $12,112 Retail 18% 2016 CONS IL, $4, FAM, $45,698 MF, $11,563 CRE, $101,919 17% 81% AG PROD, $6,552 CONS IL, $27,801 61% AG RE, $16,782 C&I, $26,183 C&D, $17,935 AG RE, $32,838 C&I, $47,343 C&D, $3,175 $236 M $274 M 4 Case Study Institution

7 5 Case Study Institution

8 Eleven Year Totals C/O s = $12.6 M +7.7% PFLL = $11.7 M 6.8% NCO s = $10.9 M 6 Case Study Institution

9 Loss Emergence Period Loan Origination Loss Event Bank Becomes Aware of Loss Charge Off or Reserved Upfront Recognition How Long is the Average Delay? CECL Requires 7

10 Average Annual PFLL 0.50% 8 Average of Most Recent 5 Years 0.17%

11 CECL Back-Test $4,000 $3,500 $3,533 $3,000 $2,500 $2,000 $1,500 $1,748 $1,617 $1,000 $898 $1,030 $835 $500 $444 $ & Before $27 $2 $1 $ Recognized Loss $0 $0 $309 $734 $1,891 $1,922 $2,925 $1,976 $678 $390 $852 Year Loss Incurred $1,748 $898 $1,030 $3,533 $835 $444 $1,617 $27 $2 $1 $ and After 9 Case Study Institution

12 C&D Net Charge-Offs $3,255, % C&I (INCL MUNI) $2,063, % OO CRE $2,118, % NOO CRE (INCL MF) CONS, OTHER 1 4 FAM RES RE $778,000 $582,000 $1,121, % 0.19% 0.40% Overall Average Charge Off Rate Last 11 years 0.43% HE / HELOC's CREDIT CARDS $535,000 $309, % 0.24% Total of 431 Loans C/O > $10 K each 145 Loans (92% of Total $ s) AG LOANS $116, % 13 Loans / Year $0 $500,000 $1,000,000 $1,500,000 $2,000,000 $2,500,000 $3,000,000 $3,500, Case Study Institution

13 Client s Current ALLL Process Pool Based Allocations Only 5 Pools Used 1. Commercial (all 6 Commercial products combined) 2. Agricultural (RE & AG Production Loans combined) 3. Residential RE (1 4 Fam. & HE combined) 4. Consumer (4 Product Types combined) 5. DDA Overdrafts 11 Case Study Institution

14 Client s 6/30/2017 Pool Based Allocations $913,000 Allocation for SM / Substd. $1,486,000 Specific Allocation (Impaired) $1,088,000 Total Reserve $3,487,000 Reserve as a % of Avg. Loans 1.24% 12 Case Study Institution

15 Loss Estimation Methods CECL Simplified All of the Method None of the Madness

16 Historical Loss Rate 13A Incurred Loss Basis Delayed / deferred recognition 3 5 year look back Guaranteed to be wrong and late in times of change

17 Historical Loss Rate To Make Historical Loss Rate Methodology CECL Compliant Extend time period from annual to life of loan Q Factors incorporate forward looking forecasts Correlate loss rate to macro economic trends 13B

18 Migration Analysis Requires extensive historical data Heavily reliant on consistent risk rating Best used if loss emergence period is short Most valid when used on large sample sizes 14

19 Millions $25.0 $20.0 Migration Analysis Results SM Substandard Doubtful Lowest Rate 0.00% 4.75% 16.05% Highest Rates 3.55% 17.34% 78.74% $15.0 $10.0 $5.0 $0.0 Q Q Q Q Q Q Q Q Q Q Q Q Q Special Mention $10.9 $11.5 $11.5 $11.4 $12.3 $14.0 $13.2 $13.1 $13.7 $11.6 $8.3 $10.3 $12.1 Substandard $10.4 $9.8 $9.8 $11.3 $12.2 $13.8 $14.8 $15.4 $16.4 $19.9 $19.2 $20.2 $19.5 Doubtful $0.1 $0.1 $0.1 $0.0 $0.1 $0.2 $0.1 $0.1 $0.8 $0.1 $0.1 $0.3 $ Case Study Institution

20 PD / LGD Possible Like Migration Defaults Analysis, requires large samples Delinquent > 90 Days Placed Specific on to Non each Accrual credit Charge off grade is processed A provision is established Credit Requires downgraded back testing, Loan regression terms are analysis restructured to Bankruptcy assure validity Other Requires loan covenants definition & breached tracking (DSC, of defaults TIE) 16

21 The ProBank Austin CECL Advisor 17

22 Vintage Analysis Life of loan Forward looking Connects nicely to economic Q Factors Minimum requirement 18

23 The ProBank Austin CECL Advisor 19

24 Discounted Cash Flows Single Purpose Apply to TDR s Not applicable to general loan portfolios Straight forward calculation 20

25 Loan Pool Segmentation Under CECL CECL Simplified All of the Method None of the Madness

26 Risk Characteristics Credit Score / Grade / Rating Term Loan to Value (LTV) Loan Type Industry Loan Size Vintage (year originated) Geographic Location Used to Determine Pool Segmentation See Detailed Analysis Of Loan Portfolio Assignment to Pools in Appendix A 21

27 Expansion of # of Pools Analyzed Based on Unique Risk Characteristics Commercial 1. CRE Loan Originated as C&D Loans* 2. C&I Loans Satisfactory 3. C&I Loans Monitored 4. C&I Loans Substandard 5. OO CRE 1 Cont lterm 7 yrs. or less 6. OO CRE 2 Cont lterm 8 17 yrs. 7. OO CRE 3 Cont lterm > 17 yrs. 8. NOO CRE 1 Term 7 yrs. or less 9. NOO CRE 2 Term 8 17 yrs. 10. NOO CRE 3 Term > 17 yrs. 11. Multifamily 12. Ag RE 13. Ag Other 14. Muni Retail 1. Credit Card Loans* 2. HE Revolving 3. HE Lines Family Hi Credit Quality Family Moderate Quality Family Low Credit Quality 7. Consumer Other 21 Pools vs 5 Pools, currently * Higher risk pools highlighted in RED 22

28 Example Loss Estimation Methods Determining the Allowance Amount CECL Simplified All of the Method None of the Madness

29 The ProBank Austin CECL Advisor 23

30 The ProBank Austin CECL Advisor 24

31 The ProBank Austin CECL Advisor 25

32 Determining the Allowance Amount $5,000 Estimated Range of Future Losses Reserve Required $4,000 $3,487 $3,662 $3,931 $3,000 $2,841 $2,000 $1,000 $0 PD / LGD Historical Incurred Loss Vintage Migration 26

33 Determining the Allowance Amount Assessment of the Methods Method Strength / Weakness Migration Credit Rating Process Quality PD / LGD Vintage Values of CRE (LGD) Best Representation of Expected Losses Conclusion Did not use Did not use Method Selected 27

34 Qualitative Factor Adjustments CECL Simplified All of the Method None of the Madness

35 Qualitative Factors Changes in Credit Policies Underwriting standards C/O, Recovery rates Economic Conditions Macro (National) Micro (Local / State) Nature of Loan Portfolio Volume of New Loans Made Lending Capabilities Experience Depth Loan Attributes Delinquencies Classified balances # of TDR s Quality of Loan Review Level of Board Oversight Concentrations Level of Competition Legal & Regulatory conditions Strength of Collateral 28

36 The ProBank Austin CECL Advisor 29

37 The ProBank Austin CECL Advisor 30

38 CECL Data Requirements General Recommendations Build a single CECL Data Warehouse Integrate Data from Multiple Sources G/L, Loan Applications, Credit, External Sources Maintain History ~= Contractual Life of Each Loan Pool Know everything there is to know about any loans that were charged off 31

39 CECL Data Requirements Specific Recommendations Analyze Specific Data Needs of Each Loss Methodology, e.g.: Migration Accurate & Timely Credit Grades PD / LGD Default Definition Identifiers Appraisals of Collateral Vintage Origination, Renewal Dates, Contractual Terms 32

40 Some Final Thoughts CECL s Purpose / Effect? Force Banks to Recognize Loan Losses Earlier Over Longer Time Periods i.e. a Business Cycle Net Charge Offs = Reported Provision for Loan Losses This has been true historically It holds for banks of all sizes One time effect? 33

41 Some Final Thoughts Adjusting for current economic conditions implementation of CECL may indicate lower Reported Losses Obtaining the ability to forecast losses accurately will require a significant investment of time and resources (data, systems, etc.), and result in learning & improvement over time 34

42 ProBank Austin Services Available CECL Support Consulting (Education, Design, Implementation) Outsourced Implementation Support Data Collection System Design Pool Segmentation Supporting Analysis & Reporting CECL Data Warehouse Development & Storage ALCO Client Other Client 35

43 Thank You! For attending today s presentation! Please feel free to contact us at anytime we d be pleased to discuss with you your Institutions unique circumstances, or answer any specific questions you may have. 36

44 Presenters Jeff Morris (419) Managing Director ProBank Austin Financial Management Division Jeff joined Austin Associates (now ProBank Austin) in He is responsible for profitability analysis services and works with clients who utilize the firm's innovative software products, including its proprietary organization/product/customer profitability software and its loan and deposit pricing software, and provides consulting services to ALM clients. Jeff earned his CPA in 1982 in Ohio, has a BBA in Accounting from the University of Toledo, and an MBA from Northern Illinois University. Jeff has over 35 years of experience in various financial management roles. Vince Van Nevel (502) Managing Director ProBank Austin General Consulting Division vannevel@probank.com Vince specializes in safety and soundness issues such as policy development, risk management, funds management, loan review, and credit administration, as well as troubled bank assistance, strategic planning, and litigation support. Prior to joining ProBank Austin, Mr. Van Nevel was an operations officer in a community bank, and spent over ten years with the Comptroller of the Currency as a field examiner and later as a team leader in the problem bank division of the OCC s Central District Office. Mr. Van Nevel is a graduate of the University of Wisconsin Madison, and holds a Master s Degree from J.L. Kellogg Graduate School of Management at Northwestern University. 37

45 Appendix A Loan Pool Segmentation Under CECL CECL Simplified All of the Method None of the Madness

46 Construction & Development Loans (High Risk) Risk Characteristics NCO Rate = 1.41% Represent 9% of loan balances Responsible for 26% of historical C/O s Currently holding 32 Loans (22 Relationships) Average loan Size = $690,000 Five loans make up 68% of portfolio Concentrations in Hotels & Healthcare Considerable repeat business

47 Construction & Development Loans Relevant Charge Off History ( ) 1 Large C/O Golf Course Housing Development $1.5 M (w/ No Recoveries) 9 Relationships All Home Builders Originated between of 9 were AR Median Net C/O = $199 K Recovery Rate is 54%

48 Construction & Development Loans Relevant Charge Off History ( ) 1 Large C/O Golf Course Housing Development $1.5 M (w/ No Recoveries) 9 Relationships All Home Builders Originated between of 9 were AR Median Net C/O = $199 K Recovery Rate is 54%

49 Construction & Development Loans $14.0 $12.0 $10.0 $8.0 $6.0 $4.0 $13.2 Balances by Credit Grade (Millions) $7.4 $2.0 $0.0 $0.1 Satisfactory Monitored Watch

50 Commercial & Industrial Loans (High Risk) Risk Characteristics NCO Rate = 0.50% Represent 16% of loan balances Responsible for 19% of historical C/O s 269 Loans Active $48 M in balances 135 Open w/ no balance ($18 M committed) Nine Loans > $1 M ($18 M) Median Loan = $52 K 56% are AR One customer w/ $5.25 M outstanding No other (industry) concentrations

51 Commercial & Industrial Loans (High Risk) Charge Off History ( ) NCO Rate = 0.50% 23 Loans w/ significant C/O s (>$10 K ea.) 1 Loan w/ C/O of $1.5 M (61% of total) (Data Technology Company) Median C/O = $14 K Recovery Rate is 12%

52 Commercial & Industrial Loans (High Risk)

53 Commercial & Industrial Loans (High Risk) Risk Characteristics Typical Contractual Term is either 12 or 24 months (median term is 15 months) Loans are renewed if no new credit concerns Pooling Decision Migration Analysis Maintain 3 separate pools of C&I loans based on credit grade: Satisfactory Monitored Substandard

54 Owner Occupied CRE Loans (Low Risk*) Risk Characteristics NCO Rate = 0.30% Represent 25% of loan balances Responsible for 19% of historical C/O s 278 Loans w/ $92 M in Balances Median loan is $120 K 20 loans > $1 M ($46 M in balances) One customer w/ $10.4 M 76% are AR * But still large dollars at risk due to largest portfolio in the bank

55 Owner Occupied CRE Loans (Low Risk*) Charge Off History ( ) NCO Rate = 0.30% 21 Loans w/ significant C/O s (>$10 K ea.) Largest C/O = $380 K (18% of total) (Health & Beauty Salon) Median C/O = $43 K Recovery Rate is 4% * But still large dollars at risk due to largest portfolio in the bank

56 Owner Occupied CRE Loans (Low Risk*) Balances by Credit Grade (Millions) $40.0 $35.0 $34.9 $30.0 $27.0 $25.0 $20.0 $15.0 $10.0 $5.0 $1.5 $2.5 $0.0 Satisfactory Monitored Watch Substandard * But still large dollars at risk due to largest portfolio in the bank

57 Owner Occupied CRE Loans (Low Risk*) * But still large dollars at risk due to largest portfolio in the bank

58 Owner Occupied CRE Loans (Low Risk*) Risk Characteristics Typical Re pricing Term is 5 Years Amortization terms 15, 20, 25 Years Bank has no ability to call loan, except for default, prior to full amortization period Pooling Decision PD / LGD Maintain 3 separate pools based on contractual maturity: 7 years or less 8 17 years 18 years or longer * But still large dollars at risk due to largest portfolio in the bank

59 Non Owner Occupied CRE Loans (Low Risk) Risk Characteristics Represent 12% of loan balances Responsible for 10% of historical C/O s 113 Currently Active Loans 6 Loans > $1 M ($11 M Total) Median Loan size = $78 K, all others Average Contractual Term is Years Average Credit Grade is NCO Rate = 0.40%

60 Non Owner Occupied CRE Loans (High Risk) Charge Off History ( ) NCO Rate = 0.50% 14 Loans w/ significant C/O s (>$10 K ea.) Largest C/O = $314 K (28% of total) (Assisted Living Center) Median C/O = $29 K Recovery Rate is 7%

61 Non Owner Occupied CRE Loans (Low Risk*) * But still large dollars at risk due to largest portfolio in the bank

62 Non Owner Occupied CRE Loans (Low Risk*) * But still large dollars at risk due to largest portfolio in the bank

63 Non Owner Occupied CRE Loans (Low Risk*) Risk Characteristics Typical Re pricing Term is 5 Years Amortization terms 15, 20, 25 Years Bank has no ability to call loan, except for default, prior to full amortization period Pooling Decision PD / LGD Maintain 3 separate pools based on contractual maturity: 7 years or less 8 17 years 18 years or longer * But still large dollars at risk due to largest portfolio in the bank