Farin & Associates, Inc. Farin Foresight Software Certification as of November 30, 2017

Transcription

1 Farin & Associates, Inc. Farin Foresight Software Certification as of November 30, 2017 by Alpha-Numeric Consulting, LLC December 20, 2017

2 Introduction Financial institutions recognize the need for accurate equity and earnings simulations as part of their comprehensive strategic planning; additionally, there is a regulatory and business requirement to identify exposure, monitor risk, and measure compliance with established parameters. The accurate capturing of the contractual and expected behavior of the current and future balance sheet positions starts with precise record level data, code that applies relevant and defensible assumptions within an industry accepted theoretical and conceptual framework, and calculations that result in meaningful and timely output. Farin & Associates, Inc. (Farin) has engaged the independent and expert services of Alpha-Numeric Consulting, LLC to perform extensive testing of Farin Foresight to ensure that the software is able to meet the Asset-Liability Management (ALM) and Interest Rate Risk (IRR) modeling needs for a variety of balance sheet compositions, mathematically calculate simple and complex formulae, accurately and consistently apply default and custom assumptions to multiple sets of data, as well as respecting the parameters of the categories characteristics. This supplemental third party review satisfies OCC Bulletin , and other model guidance with respect to the software (code) component of an institution s overall model validation and verification, and was performed in addition to the internal analysis conducted by the vendor s staff. Scope All testing was performed in version of Farin Foresight. Access to the software was via the internet in a unique and password protected location. The database was rolled forward from the previous version to test that process. To realistically test the EVE and NII IRR results in a practical application, categories from a representative Independence Expertise Results 2

3 $14.23 billion institution with a moderately complex balance sheet were utilized. To remove any possible bias, the database contained no institution or customer information. Internal testing including regression analysis, incremental code changes, sample databases, real world application, bug reporting, enhancement requests, version to version comparison and other quality control procedures are part of the on-going responsibilities of the programming department at Farin. This external independent testing is supplemental to the internal Quality Assurance and is not intended as a replacement form of Q/A in any way. To fully comply with Bulletin OCC and the more recent FDIC Advisory in Interest Rate Risk (January 6, 2010) Model Governance requirements, financial institutions must also have an independent review of their policies and procedures, data (input and imported), assumptions, and output. Model Description Farin Foresight is designed to allow modeling of interest rate risk, earnings and equity projections, forecasting, budgeting, alternative funding strategies, and any number of what-if predictive and risk analysis simulations. Current Farin clients represent a variety of balance sheet compositions, asset sizes, complexities, charters, and geographical locations. Data, Assumptions, Theory, Code and Reporting To assess the mathematical capabilities of the model, multiple scenarios were created to enable the use of a myriad of data, assumptions, behavioral characteristics and interest rate scenarios. Independence Expertise Results 3

4 Six unique scenarios were the focus for certification purposes, representing a Base Case (rates remaining at current levels), and instantaneously and sustained +/-200 bp shocks. Categories were chosen for review to not only represent those on a typical financial institution s balance sheet, but also to represent a broad cross section of attributes for interest sensitive and non interest sensitive assets and liabilities. Testing Matrix The Testing Matrix on the following page illustrates the categories and resulting scenarios utilized as part of the process of verifying the model math. Independence Expertise Results 4

5 EVE NII Category Balance Yield/Cost % Assets/Liab Dn200 BC Up200 Dn200 BC Up200 Fed Funds Sold - CMOs 5,025, % 35.33% MBS - 30 Yr Fixed 1,100, % 7.73% Agency - 10 Yr Fixed 500, % 3.51% Callable Agency 100, % 0.70% Agency Step-up Bonds 500, % 3.51% Domestic CDs - 1 Yr Fixed 50, % 0.35% Municipal 10 Yr Fixed (Taxable) 25, % 0.18% Municipal HTM - 10 Yr Fixed 25, % 0.18% FHLB Stock 800, % 5.62% Trust Preferred 50, % 0.35% FASB 115 Adj - Total Investments 8,175, % 57.47% Comml - 10 Yr Fixed 500, % 3.51% Comml RE - 10 Yr Fixed - 3 Yr Balloon 500, % 3.51% Comml - 10 Yr Variable 500, % 3.51% Commercial Loans 1,500, % 10.54% REM - 30 Yr Fixed 500, % 3.51% REM - 30 Yr 1/1 ARM 1,000, % 7.03% REM - 30 Yr 3/1 ARM 1,000, % 7.03% Mortgage Loans 2,500, % 17.57% Personal - 5 Yr Fixed 1,000, % 7.03% Credit Cards 100, % 0.70% Consumer Loans 1,100, % 7.73% Total Loans 5,100, % 35.85% Reserve for Loan Loss (50,000) -0.35% Total Net Loans 5,050, % 35.50% Earnings Assets 13,275, % Cash 250, % Fixed Assets 750, % Total NIB Assets 1,000, % TOTAL ASSETS 14,225, % % DDA - Personal 1,000, % 7.03% Demand Deposits 1,000, % 7.03% NOW - Personal 1,000, % 7.03% NOW Accounts 1,000, % 7.03% MM 2,000, % 14.06% MMDAs 2,000, % 14.06% Regular Savings 1,000, % 7.03% Savings Accounts 1,000, % 7.03% CD - 6 Mo 100, % 0.70% CD - Variable 275, % 1.93% Regular CDs 375, % 2.64% Total Certificates 375, % 2.64% Total Deposits 5,375, % 37.79% Fed Funds Purchased - Repo - 6 Mo 100, % 0.70% Repos 100, % 0.70% Convertible - 5 Yr Annl 1,000, % 7.03% Convertible Advances 1,000, % 7.03% Amortizing - 5 Yr 3,000, % 21.09% Amortizing Advances 3,000, % 21.09% Bullet - 5 Yr 3,000, % 21.09% Bullet Advances 3,000, % 21.09% FHLB Borrowings 7,000, % 49.21% Total Borrowings 7,100, % 49.91% Total Interest Bearing Liabilities 12,475, % 87.70% Other Liabilities 950, % FASB 115 Tax Payable - Total NIB Liabilities 950, % TOTAL LIABILITIES 13,425, % 94.38% Common Stock - Surplus - Retained Earnings 800, % Net Unreal G/L on AFS - TOTAL EQUITY 800, % TOTAL LIABILITIES & EQUITY 14,225, % Independence Expertise Results 5

6 Data Data can be imported and/or input at the record level or summary level for all balance sheet categories. Balances and rates (yields and costs) are available for beginning, new, ending, average and summary periods, and flow by definition into subtotal and total categories. Maturity and repricing cash flow is available for all periods. Manual input allows for easy modification and automated data feeds reduce the likelihood of entry error. Web-based updates are also an integral part of the model population. Assumptions The flexibility of the model allows for a great amount of detail to be applied in the definitions of the assumptions for the existing portfolio, new and replacements to the portfolio and new/replacement runoff. Multiple databases can be established and assumptions can be copied from one to another. Assumptions that will alter the current and future positions are correctly applied and results are calculated immediately for review and analysis. The math for the following assumptions was reviewed and no errors were found: - Category type, subtype, special purpose classifications - Historical data: amortized cost, market value, average balance, ending rate, coupon rate, par, pledged amount, premium/discount - Current data: amount, par, premium/discount, type, at rate, coupon, payment fees, collateral spread %, servicing %, haircut % <1 year, haircut % > 1 year, loan to value %, NAICS code, relationship, non maturing balance, maturity term, payment information, balloon data, repricing, calls, CMO detail, specified cash flows, market value Independence Expertise Results 6

7 - Interest Sensitivity: payment frequency, accrual method, first and subsequent repricing terms, - Call options - Rate restrictions: floors, caps, (anniversary, lifetime, absolute) - Pricing: method, pricing driver, repricing driver, offsets, lags, - Cash flow: amortizing/non-amortizing, timing - Original and replacement maturity term, non-maturity designation, servicing percentage, timing - Prepayments, decay, early withdrawal percentage - Growth rates, - Sales, - Market Value: driver method, driver, offsets, cost adjustment, overrides - Present value calculation method: discounted cash flow, book value, zero, account overrides, formula, instrument overrides - Taxes: Federal, State, Local Theory Industry best practices for simulation models require that the model be able to use beginning balances that tie to the institution s application and subsidiary ledgers, apply unique behavioral assumptions, and create results for a variety of interest rate and economic environments from which management can measure and monitor risks and make applicable corrections and enhancements to strategic planning. Farin Foresight has the fundamental conceptual ability to meet these standards. Independence Expertise Results 7

8 Code and Calculations In addition to the internal testing applied to the software code, the math was tested externally to determine if the model is correctly presenting the beginning, interim and final results. As part of a comprehensive certification, model math was proven out by independent calculations and compared to actual performance. When evaluated within the context of expected behaviors, results were reviewed to determine if they were correct in terms of increases/decreases, volatility, asymmetric/symmetric variances, and adherence to defined parameters. After extensive testing, the following hypotheses were proven: - Math proven for one type of category is presumed to work for all subsets of the category assuming similar characteristic definition. For example, if the model math is correct for the 1-5 month CD category, the math is correct for all short term, fixed rate, bullet time deposit categories. - Categories and assumptions copied from one database to another for the same institution are presumed to work in an identical fashion. - Results produced in one database will be reproduced in another database if defined in the same manner. Reporting The model contains a variety of reports and reporting formats. Within the on-line Help, there is a documentation of internal assumptions and formulas, which ties back to the actual mathematics and results. Standard and custom reports, in multiple formats, also reconcile to the data within the model and to the final results without exception. Independence Expertise Results 8

9 Summary Statement This certifies that, as of November 30, 2017, the Foresight model, version , has the conceptual ability to apply industry best practices to a financial institution s balance sheet, produce EVE and NII IRR results, represent current and future balance sheet positions, and is mathematically accurate in a variety of most likely, rising, and declining rate scenarios. Alpha-Numeric Consulting, LLC Hoschton, GA Independence Expertise Results 9