Unravelling the Guidelines in Preparation for CECL (ASU 2016-13) 11/29/2016 1
Today s Agenda Introductions CECL Overview Impact on the Institution i Calculation Methodologies Data Requirements Disclosure Requirements What Should You Do? 2
Ardmore Fintellix Solutions OnApproach Partners Peter Cherpack, EVP Ardmore Banking Advisors, CEO Ardmore Fintellix, LLC Peter.Cherpack@ArdmoreFintellix.com Sunny Malhi, VP Fintellix Solutions, COO Ardmore Fintellix LLC Sunny.Malhi@ArdmoreFintellix.com In 2015 Ardmore Banking Advisors partnered with Fintellix Solutions - An overseas firm successfully providing Financial compliance technology solutions for10 years in Europe and Asia Together Ardmore and Fintellix created a secure, hosted credit portfolio management solution ( CPM ) to support US smaller financial institution s needs in credit reporting, stress testing and ALLL automation Ardmore s 25 years of working with smaller financial institutions and their credit data along with Fintellix s global tested technology create a uniquely valuable solution for US community banks, offering advanced features and local hands-on implementation Ardmore s local team supports all of the institution's data management, data structure mappings and any required data access needs 3
Issues with Current Incurred Loss Model 600.00 350 3.50 cies in $m Delinquen 500.0 400.0 300.0 200.0 3.00 2.50 2.00 1.50 1.00 Charge off rate (%) 100.0 0.50 0.0 2006/01 2007/01 2008/01 2009/01 2010/01 2011/01 2012/01 2013/01 2014/01 0.00 Delinquencies On All Loans And Leases Scenario 1 - Dec 31 2009 ALLL Reporting: Applicable loss rates calculated using loss rates between 2005 and 2008 Substantially low charge-offs resulted in low loss rates Regulator s expectation was to include higher ALLL for existing market conditions Many banks opted to have lower ALLL to show stable earnings Issue: Delayed Recognition of losses Charge-Off Rate On All Loans ALLL/Loans Scenario 2 - Dec 31 2011 ALLL Reporting: Applicable loss rates calculated using loss rates between 2008 and 2010 Substantially high charge-offs resulted in high loss rates Regulator s expectation was to include stable ALLL for existing market conditions Banks reported high ALLL as earnings had stabilized and later resorted to negative provisioning 4
CECL Overview A principle based rule Current Expected Credit Loss The new CECL rules are pro-active, and attempt to more adequately tie reserves to actual loss cycles CECL is based around the concept of projecting expected losses for the life of each loan (LOL) based on loan risk characteristics and supportable projections Projections include a reasonably supportable economic projection of the future (period undefined) The reasonably supportable projections need to be correlated to past loss patterns of performance for specific asset types compared to national and regional economic metrics Life of Loan and CECL projections may not be the same thing (necessarily)
What s in scope, what s not Inclusions Exclusions Financial Receivables Financial Assets at Fair Value thru Net Held To Maturity Debt Income Loan Commitments, Stand By L/C, Available for Sale Debt (separate model) Guarantees Insurance Policy Loans Lease Receivables Loans made to participants by defined Reinsurance Receivables contribution employee benefit plans Receivables on Repurchase and Securities Loans and receivables between entities Lending Agreements under common control Definition: Amortized Cost Basis The amortized cost basis is the amount at which a financing receivable or investment is originated i or acquired, adjusted d for applicable accrued interest, t accretion or amortization of premium, discount, and net deferred fees or costs, collection of cash, writeoffs, foreign exchange, and fair value hedge accounting adjustments
CECL Update Time Line Others Fiscal years beginning after Dec. 15, 2020 & interim periods within fiscal years beginning after Dec. 15, 2021 Non-SEC filers Fiscal years beginning after Dec. 15, 2020 SEC filers Fiscal years beginning after Dec. 15, 2019 Early adoption is permitted for fiscal years beginning after Dec. 15, 2018 No carve-outs; guidelines applicable to all organizations holding financial assets 7
Potential Financial Impact Of CECL 0.6 0.5 0.4 Loss Curve Large increase in reserves possible depending on portfolio mix One time hit to an institution s capital Lending and acquisition strategy may be modified 0.3 0.2 0.1 0 Q1 Q2 Q3 Q4 Q5 Q6 Q7 Q8 Q9 Q10 Q11 Q12 Q13 Q14 Q15 Q16 Q17 Q18 Q19 Q19 Q20 Q21 Currently applicable 12 Month Loss rate (point in time) CECL Model expects area under the loss curve to be applied But there is confusion in the industry: This is not a tweak for anybody or any bank Steven Merriett, Chief accountant at the Federal Reserve System If you have a fairly straightforward approach today, even using Excel spreadsheets there is an expectation that you should be able to continue to use that type of an approach Robert Storch, Chief Accountant Division of Risk Management Supervision, FDIC 8
Challenging institutions to consider relevant available information Includes current conditions, past events and reasonable supportable forecasts Includes internal, external or combined information May not require a search for what s not reasonably available May find... internal information is sufficient Typical historical credit loss data provides a basis for loss information by pool but should be adjusted for Current asset specific risk characteristics Term/Vintage Economic conditions 9
Challenging institutions to consider how to adjust historical credit loss information From 326-20-55-4: Due to the Borrower s: Financial condition, credit rating, credit score, asset quality, or business prospects and ability to make scheduled interest or principal payments Due to the Specific Asset Type s: Remaining payment terms, remaining time to maturity, the timing and extent of prepayments, nature and volume, volume and severity of past dues, and of adversely classified or rated, and the value of underlying collateral (if not using practical expedient) 10
Challenging institutions to consider how to adjust historical credit loss information Due to the Institution's: Lending policies and procedures, including changes in lending strategies, underwriting standards, collection, writeoff, and recovery practices, as well as knowledge of the borrower s operations and standing in the community, quality of the credit review system and the experience, ability, and depth of the entity s management, lending and relevant staff Due to Environmental Factors: Changes in regulatory, legal, or technological environment, and expected changes in the general market condition of geographical area or the industry, and in international, national, regional, and local economic and business conditions and developments, including the condition and expected condition of various market segments. 11
And, When Forecasts are not Supportable Revert to historical credit loss data for life of loan (term) - No adjustment for economic conditions - No reversion technique the amount an entity expects to collect If the Institution doesn t have historical lifetime data: - They can use historical credit loss experience of financial assets with similar risk characteristics as a basis for an entity s assessment of expected credit losses - Historical loss information can be internal or external historical loss information (or a combination of both) - An entity shall consider differences in current asset specific risk characteristics: - Differences in underwriting standard/portfolio mix/asset term within a pool - Q factors (again)? External data? 12
Calculation Methodologies 13
General Principle for all Non-Cash Flow Methodologies Identify the risk Historical Data (Loss Reasonable and Revert to Mean for criteria for pooling rates / Probability of Supportable unavailable future Defaults) Forecast 14
Calculation Methodologies A Judgment by Asset Class Current Methodologies Historical Loss Rate Migration Analysis Cohort Analysis Vintage Analysis Probability of Default Discounted Cash Flow Calculate estimate based on past experience Apply Historical loss rates calculated based on look back period Apply loss rates for the life of the loan Apply adjustment based on qualitative and environmental factors Risk ratings or credit grading of assets required Migration of assets from one classification to another tracked over life of loan Apply adjustment based on qualitative and environmental factors Classify all loans into different cohorts based on relevant risk characteristics of the loan Analyze historical asset performance Develop a cumulative loss curve for cohort Future year loss estimation based on current and expected future economic factors Measure impairment based on the age of the loan Historical asset performance analyzed Develop a cumulative loss curve for each financial asset Future year loss estimation based on current and expected future economic factors Losses = PD * EAD * LGD Continuous Back testing required PD/LGD from Basel systems only have a limited outlook of one year instead of life of loan Newer models would have to be created Based on the present value of expected cash flows Effective Interest rate for discounting Loss curves might be the building block to arrive at time of cash flows Probability weighing of cash flows not allowed 15
Possible Historical Loss Rate Calculation Scenario 16
Vintage Analysis Scenario Year of Loss Experience in Years Following Origination Origination Year 1 Year 2 Year 3 Year 4 Total Expected 20X1 50 120 140 30 340-20X2 40 120 140 30 340-20X3 40 110 150 30 330-20X4 60 110 150 40 360-20X5 50 130 170 50 400-20X6 70 150 180 60 460 60 20X7 80 140 190 70 480 260 20X8 70 150 200 80 500 430 20X9 70 160 200 80 510 510 Originations of 4 year amortizing loans secured by collateral that provides a relatively consistent range of loan-to to-collateral-value ratios at origination Track these loan on the basis of calendar year of origination and accumulated losses on year on year basis for the contractual term of the loan Based on the analysis of historical loss patterns and the current conditions and reasonable and supportable forecasts, the entity determines the expected losses of each year of remaining maturity for each origination year 17
Extrapolating from the Basics Things to consider in real life estimation: Contractual Maturity: Same pool of loans could have varying contractual maturity Assumes a short term amortizing loan with loss data available. If vintage analysis is applied on longer term loans, mean reversion to be applied for years/quarters of unavailable loss rates Prepayments: Assumes prepayments built into loss rate, but practically prepayments would reduce the contractual term of loan for analysis Assumptions: Forecasts losses/loss rates are based on judgement (economic conditions i impacting the pool). Banks might be required to use quantitative methods to forecast losses for the remaining maturity of each origination period 18
Data Requirements 19
Internal Data That May be Needed Origination Data Historical Loss Data* Transactional Data Origination Date Interest Rate Maturity Date/Term Risk Rating Repricing Date Product/Loan Type Days Past Due Loan Purpose Borrower Industry CRE Property Types Accrual/Non Accrual Credit Score Borrower Location Borrower Financial Ratios* Collateral Details* LTV/NOI/DSCR* Charge Off Dates Charge Off Amounts Recovery Dates Recovery Amounts Partial Charge Off Dates Partial Charge Off Amounts Rating Changes Post Partial Charge Offs Charge Off Events * This data often not in core systems Risk Rating Changes Prepayments Interest Rate Changes Payment Dates 20
External Data That May be Needed National Level Regional Level Projections Real GDP growth / Nominal GDP growth National Level Projections Housing Price Index available under DFAST Per Capita Income Forecast state level market Disposable Income factors using historical Unemployment Rates series Interest Rates (Treasury) Housing Price Index Commercial Real Estate Price Index Management judgement to Workers statistics (within a particular industry) predict the economic Hourly Wages of employees in industry factors for state level Consumer Price Indexes Peer Data (ALLL, Charge offs, Delinquencies etc.) 21
Core System and Data Management Issues Questions to Consider Is your core in-house or hosted? How much account and transaction history do you keep on line? How can this data be accessed? Where does your data go when it s no longer available on-line? Do you have a data warehouse? What account and transaction data is maintained How much of this data has been captured and for what period of time? What tools/resources are required to access this data? (Interface?) Where is charge offs and recovery data kept? Excel Spread Sheets? If a portfolio has been acquired, what historical data has been brought over for those loans? Assuming our data is accessible and available, what s the integrity of our data that would be used as the basis for creating pools? What are the estimated costs we must budget for to get the data? 22
Disclosures 23
Increased Disclosure Requirements Detailed disclosure guidance provided for: 1. Credit Quality Information 2. Allowance for credit losses 3. Past due status & non-accrual status deterioration o 4. Purchased financial a assets with credit 5. Collateral-dependent financial assets 6. Off-balance-sheet credit exposure Enhanced disclosures related to credit quality and underwriting standards of an organization s portfolio Description i of entity s accounting policies i & methodology to estimate ALLL Discussion of factors influencing management s estimates of expected losses including past events, current conditions & the forecasts Credit-quality indicators for all classes of financing receivables (assets) disaggregated by vintage year must be disclosed for five reporting periods (years) Some transition relief for some disclosures: Non-SEC filers may phase-in vintage disclosure of credit quality indicators by presenting only three most recent origination years 24
Increased Disclosures Disaggregation of credit quality indicators by vintage: - More than five annual reporting periods not required - Prior to fifth annual reporting period shown in aggregate - Must reconcile purchase price and par value of purchased credit deteriorated (PCD) assets - Roll forward of the allowance for credit losses 25
Past Due Disclosures 26
How Should You Start?
2020 Is Around The Corner Data Capital Considerations Audit & Examinations If 5 years of data is needed for CECL, institutions need loan and borrower data dating back to 2015 Most core systems don t store date for more than 13 months Prepayments, recoveries & charge-off related data typically available in unreliable excel sheets Use of peer data allowed but could increase reserves further There is a possible 20%-40% increase in reserves on transitioning to CECL resulting in a one-time hit to capital. Institutions need an early view of the amount of their capital hit, otherwise they will be flying blind when they do their capital planning Institutions will need to run parallel CECL calculations for a period of time prior to 2020 validating capital planning estimates. Recently auditors & examiners have increased their demands related to documentation on Q & E adjustments CECL is inherently more subjective involving multiple assumptions related to life of loan estimates, forecasting etc. Expectation of booking losses at the time of origination i now brings underwriting process under heightened scrutiny 28
Suggested Path To CECL Compliance For Financial Instituions Credit Data Assessment Automation of Current ALLL Early 2017 CECL Awareness Mid 2017 CECL Parallel Runs Automation Solution Alternatives (Budget) CECL Fine Tuning & Capital Impact Early 2018 Data Remediation 2019 2020 CECL Solution Working in Production Data Inventory, Gap Analysis Data Clean up and Augmentation Step One: Automate Current ALLL Step Two: Run Parallel with CECL Step Three: Adjust & Plan 29
Proactive Credit Data Management Road Map Long Term Goal State Stress Testing Create Data Once, Use Many Ways ALLL Calculation Reporting & Analytics Assessment & Cleanup Program One Version of the Truth Secure, Historical Data Mart - CECL Justifications/Support - Stress Testing Data/Support - Board Reporting - Management Reporting - Regulatory Reporting - Conduct Inventory of Data and Sources - Establish Data Gaps and Goals - Establish Coding & Guidance Standards Typical Current State - Minimal Governance - Focus on Getting Loans Booked - Much Follow up & Scrubbing 30
How Can Ardmore Fintellix Help?
Rapid Gap Data Assessment CECL and Stress Testing Readiness Assessment Service Is your portfolio data ready to support regulatory needs for CECL, Stress Testing & Concentration Management? How do you know? Is there enough time to prepare? Ardmore Fintellix s Credit Data Experts Can Help You Answer These Questions Quickly and Affordably A proven expert third-party Credit Data Assessment Service, R-GAP is a comprehensive review of your institution's current Credit/Risk Portfolio Data Management processes and tools. Gap Analysis of your current credit data management vs. industry best practices Review of the credit data collection processes and status of credit data accessibility & data integrity Custom Action Plan to address data quality and governance issues Ardmore s Credit Data Assessment service was extremely valuable to Brookline Bancorp as it helped us quickly identify strategic and tactical actions we could take to be better prepared for CECL and stress testing initiatives. The process was quick, virtually painless, and the results were actionable. Robert Rose, Chief Credit Officer, Brookline Bancorp 32
ArdmoreFintellix s Credit Portfolio Management Suite BORROWER S FINANCES ANALYZED (SPREAD) LOAN DEAL WRITTEN UP & APPROVED PIPELINE MANAGEMENT ALLL RESERVE CALCULATION (CECL READY) PORTFOLIO ANALYTICS & REPORTING CREDIT PORTFOLIO MANAGEMENT UNDERWRITING, ORIGINATION, DOC PREP CLOSING DOCS CREATED BY BANK PREP FOR DATA ENTRY DATA ENTRY (BOOKED) IN ACCOUNTING SYSTEM TRANSACTIONAL TRANSACTION PROCESSING REPORTING TO OPERATIONS CAPITAL STRESS TESTING PAYMENT PROCESSING & STATEMENTS TO BORROWER PORTFOLIO STRESS TESTING COLLECTIONS & RECOVERY 33
Solution Architecture Pre-built, automated data Extract and Load Credit Data Sources ArdmoreFintellix ALLL Data Warehouse Financial Institution Specific Data Marts Reports and Dashboards Applications Reporting/CECL/Stress Testing Configure Reclassify Analyze Compute 34
A Workflow Application Approach Manager View Analyst View Manager confirms CECL number after collating CECL data from various analysts. Post closure, the CECL confirmed numbers can be used for further analysis Closure The Analyst calculates CECL for allocated FAS 5 segments or FAS 114 accounts. Sends the CECL calculation sheets to Manager Admin Review Analyst Review Computation The Analyst can move accounts between FAS 5 and FAS 114 or change asset classifications.. The changes are sent for approval to Manager At this stage the CECL Manager views and makes any changes he wants on the system distribution of FAS 5 and FAS114 or Risk Classifications. Once he freezes the set-up, the segments/accounts move to respective accountable users (analysts) 35
36
37
Thank You, Questions? For more information: petercherpack@ardmorefintellix peter.cherpack@ardmorefintellix.comcom sunny.malhi@ardmorefintellix.com 38