CECL ONE YEAR CLOSER

Transcription

1 CECL ONE YEAR CLOSER Greg Clausen Partner Eide Bailly LLP Darrell Lingle Partner Eide Bailly LLP

2 CECL One Year Closer to Implementation Greg Clausen, CPA Partner Darrell Lingle, CPA Partner

3 Items Being Covered Effective Dates Regulatory Reviews of CECL Format Recommended Timeline Goals for Successful Implementation Choosing a Method Different Approaches to CECL Pros and Cons of Different Models Data Requirements Information to Start Gathering What to Do Now 3

4 Effective Dates PBE s (SEC Filers) Fiscal years beginning after December 15, 2019, including interim periods within those fiscal years Regulatory Reporting Effective Date March 31, 2020 Other PBE s (Non-Filers) Fiscal years beginning after December 15, 2020, including interim periods within those fiscal years Regulatory Reporting Effective Date March 31, 2021 Non-PBE s Fiscal years beginning after December 15, 2020, including interim periods beginning after December 15, 2021 Regulatory Reporting Effective Date December 31, 2021 Early Adoption Permitted Fiscal years beginning after 12/31/18 4

5 Regulatory Reviews of CECL Format Feedback Do not expect much direct feedback on methodology until after the implementation date Additional Guidance Do expect additional supervisory guidance and FAQ s as needed to aid in the successful implementation (likely through Financial Institution Letter s) Expectations A good faith effort to implement the new accounting standard in a sound and reasonable manner 5

6 Recommended Timeline to Implement 2016/ Select team Save data Review CECL methods Select a method Divide loans into risk pools Gather historical loan & economic data Populate with data Develop accounting policies Determine internal controls Fine tune Perform test runs Determine Impact to capital Up and running 6

7 Goals for Successful Implementation Loss estimate complies with GAAP and expectations for Bank s size and complexity Internal controls commensurate with a material significant estimate Sustainable, efficient process Level of effort to generate quantitative and qualitative disclosures is low 7

8 Basic CECL Elements Lifetime Historical Losses Adjust for Current Conditions Reasonable and Supportable Forecast = CECL 8

9 Selecting a Method Selecting a Method FASB does not require any one specific method to be used Modeling vs. Analytic Approach Scalability Consensus amongst regulators that calculation should be scalable to size and complexity of institution Common Approaches Vintage Analysis Migration Analysis Past Due Analysis Discounted Cash Flows Probability of Default/Loss Given Default 9

10 Purchasing a CECL Model Purchased Modeling Systems Several vendors available May use their own proprietary systems Can use multiple approaches (vintage, migration, discounted cash flow, etc.) Likely will need to use other modules of their product in order to link with the CECL module More computer driven and should have less manual input Should link with core system to capture data electronically Historical data will be limited to what is on the entity s computer system Modeling data usually based on national or regional data that may not be relevant to smaller entities Model will need to be validated by an independent party resulting in additional cost Should have more consistent results as generally data driven with little judgement involved Could be more costly than self developed format Should save on staff time not having to gather so much data 10

11 Various Approaches to CECL Vintage Analysis Use historical loss rates based on life of loan Need origination date by year Track losses by year of origination and risk pool Consider prepayments and renewals Determine economic data that will impact each risk pool Apply qualitative factors to adjust for current conditions Apply forecasted economic data to adjust for anticipated future conditions 11

12 Vintage Model Example 12

13 Why Vintage Approach Versus other Methods Pros Information is generally available Easier to understand and implement than other methods A little less data gathering than other methods Information can be used for those requiring the vintage footnote disclosure on audited financials involving PBEs Cons Still a lot of data gathering (but less than other methods) If a lot of different maturity terms for the same risk pool then more calculations needed (3, 5, 7, 10, 15 year terms each are a different vintage calculation-will try and use the term with the largest dollar volume and apply to rest) 13

14 Various Approaches to CECL Migration Analysis Divide loan pool by common risk pools Break down risk pool by loan risk rating using a historical point in time Follow the migration of loans over a period of time (4-7 years) including payments, charge-offs and changes in ratings Use that experience to apply to current common risk pools and calculate historical experience portion of reserve Link changes in related economic data by risk pool for the period of migration selected Apply qualitative and forecasted economic data to historical experience 14

15 Various Approaches to CECL Past Due Analysis Similar to migration analysis used to establish a historical loss experience The basis for determining historical loss experience is tied to the past due amounts by loan risk category for a point in time Breakout the loan balances by the normal 30, 60, 90 and over 120 days Track the loss experience over a period of time Link the time frame to relevant economic data for the period reviewed by loan risk pool Use the loss experience to apply to the current loan portfolio and then apply qualitative and forecasted factors to the historical experience 15

16 Various Approaches to CECL Probability of Default/Loss Given Default Approach The formula is EL= PD*LGD*EAD EL= Expected Loss PD= Probability of Default LGD= Loss Given Default Rate EAD= Exposure at Default PD is based on the institutions experience broken down by common risk pools LGD is generally based on current loan to value rates, it is the portion of the loan subject to loss exposure EAD is the loan balance at the time of calculation, generally it is the loan balance but the amount can be higher if other cost are involved 16

17 Various Approaches to CECL Example of PD/Loss Given Default Approach Borrower obtains a loan for $5 million (EAD). The borrower secures the loan with collateral worth $4 million after discounting. The probability of default (PD) is 18% based on the risk for these types of loans and historical losses. The loss given default (LGD) is ($5 million - $4 million)/ $5 million which equals 20%. The expected loss (EL) is calculated as 18% * 20% * $5 million = $180 thousand 17

18 Probability of Default/Loss Given Default-Pros and Cons Pros Used in valuing purchased loan portfolios More fined tuned as it considers loan to values Cons Probability of default (PD) needs to be calculated, usually based on historical experience Still need to link PD with historical economic data for context Need LTV data to determine the amount of loss exposure with periodic updating More difficult to create historical loss experience due to data requirements (LTV) 18

19 Brief Description of Various Approaches to CECL Discounted Cash Flow Expected cash flows are forecasted and discounted back using the original effective rate The difference between present value and the amortized cost basis is the allowance Subsequent changes in the allowance can be reported as an adjustment to the (ASC ): Allowance expense Interest income due to the passage of time Renewals, prepayments, extensions and modifications should not be considered when using this method 19

20 Discounted Cash Flow- Pros and Cons Pros Somewhat straight forward in conceptual terms Cons Requires documentation of cash flow expectations which maybe more difficult Still need to put in context of economic cycle both current and forecasted as that effects expected cash flow Larger institutions will need more computing power to run discounted cash flows calculations More complex spreadsheets needed Need to estimate recovery and charge off periods and document 20

21 Determining Risk Pools Collective (Pool) Evaluation Required for financial assets when similar risk characteristic(s) exists Individual Evaluation Required when a financial asset does not share risk characteristics with its other financial assets Examples of Shared Risk Characteristics Internal or external credit score Risk ratings or classification Financial asset type Collateral type Size Effective interest rate Term Geographical location Industry of borrower Vintage 21

22 Economic Trends Economic data/trends should match the risk pool being analyzed Make sure the data is relevant to your institution, much of the national data will not be relevant Movements in the economic data chosen should positively or negatively impact the potential loan losses for that pool Develop and maintain data sources to be used for updating Look for region, county, state or city economic data Cannot rely on one economic data point For example crop prices, look at local yields as well, yields is what has saved many of the farmers in the Midwest in

23 Current Qualitative Factors Qualitative factors are similar to current ASC 450 factors Numbers are rather subjective Increases or decrease should have some rational support through comments provided Can have negative numbers which should be supported Qualitative factors can be by loan category and entity specific Individual allocations are generally smaller amounts 23

24 Forecasted Data FASB requirement is reasonable and supportable Forecasted data relevant to smaller financial institutions will be difficult to obtain as little is available Much of the forecasted impact will have to be inferred based on historical and current trends Realistically any forecasting of data will be limited to months with most of it being 12 months Document conclusions through narrative 24

25 Information to Gather Now Current and Forecasted Data Identify data that correlate with losses incurred by loan risk pools No limitations to what can be used Use data that is relevant, somewhat easy to locate and don t use more data points than what you need. Be prepared to update to better data Build data base Ag loan data could include commodity prices; equip and land prices; crop yields; land rental rates; future contracts for commodities; long range weather forecast Commercial loans data might include local and national GDP; unemployment rates local, state or national; oil prices; manufacturing data; consumer sentiment 25

26 Potential Issues During Implementation Many banks will not have historical information going back further than 3 years, the basis for most any model will be driven by historical data Depending on the method chosen for CECL will drive the potential issues (each method has its own issues) Initial calculations will be time consuming Does the final outcome make sense Gathering economic data relative to the market area the bank operates in will be difficult Local forecasted data will be limited usually to one year, if available at all for smaller entities 26

27 Things to do Now Determine CECL team Save Historical Data Discuss with Core Provider Pick CECL Approach Determine Risk Pools Identify Economic Factors Create Historical Loss History Capture Other Relevant Data Develop Economic Trends Get educated on CECL and document meetings Determine if data maybe running off your IT system, many financial institutions have 3 year limits Discuss with your provider if they will be providing software updates to capture needed data Start looking at different CECL approaches and pick one (may be dependent on specific entity data available) Break loan portfolio down by common risk pools Think about the economic factors that impact losses by risk pool Start creating loan historical data by risk pool based on CECL format selected (loss history) Consider capturing other loan data such as FICO scores, loan to value, DSC and equity ratios depending on model Build economic data that links with historical loan data for economic trends (need historical and forecasted data) 27

28 Questions? This presentation is presented with the understanding that the information contained does not constitute legal, accounting or other professional advice. It is not intended to be responsive to any individual situation or concerns, as the contents of this presentation are intended for general informational purposes only. Viewers are urged not to act upon the information contained in this presentation without first consulting competent legal, accounting or other professional advice regarding implications of a particular factual situation. Questions and additional information can be submitted to your Eide Bailly representative, or to the presenter of this session. 28

29 Thank You! Greg Clausen, CPA Partner Darrell Lingle, CPA Partner