Managing the Simulation Model

By Dennis Zimmerman Jr.

While decision making at the community bank level relies on management’s experience, reason and careful deliberation, analysis derived from the use of financial models play an important role in proper decisioning. Models turn information into analysis, which then is used to make better decisions. While financial institutions use models in a wide range of applications, today’s focus is on the asset/liability simulation model — the tool that best measures the possible adverse effects on earnings and equity precipitated by movements in interest rates. Recent shifts in the Treasury yield curve have elevated the importance of an effective asset/ liability management program. As such, management should consider, not just as best practices but as regulatory expectations, incorporating the following items into this year’s deliverables:

  • Key assumption review — With primary focus on non-maturity deposit and prepayment assumptions. Institutions should strive to use assumptions that their profiles and activities and generally avoid reliance on industry estimates or model defaults.
  • Sensitivity testing — Sensitivity analysis tests the model’s parameters without relating those changes to an underlying event or real-world outcome. It should focus on the most influential assumptions to identify key variables whose volatility may significantly affect interest rate risk exposure.
  • Stress testing — Stress analysis uses the model to predict a possible future outcome given an event or series of events. It should consist of scenarios that are severe yet plausible considering the existing level of rates and the interest rate cycle. Stress scenarios could include, but are not limited to, the following:
  • Instantaneous and significant changes in the level of rates;
  • Substantial changes in rates over time;
  • Changes in the relationship between key rates;
  • Changes in the slope and shape of the yield curve; and
  • Key assumption/sensitivity testing with primary focus on prepayment speeds and NMD decay and lag/betas.

Commonly performed stress tests include a combination of scenarios chosen from a matrix of sensitivity tests and rate scenarios, including the lengthening/shortening of decay rates, the shifting in lag/beta assumptions, the increase/decrease of prepayment speeds and the introduction of a credit event (deterioration in credit).

  • Back testing — Back testing analysis allows management to compare model results to actual behavior to form an opinion as to whether the simulation model is predicting reasonable future earnings and equity balances. Back testing is an integral part of a comprehensive asset/liability management and IRR review process. When performed correctly, back-testing analysis can either confirm that the data and the assumptions are correctly forecasting present values and earnings or highlight areas that require attention. Back testing is never one and done. It should be continuous and ongoing.
  • Statement of certification — If you haven’t already, ask your ALM provider for the most current copy of the model’s Statement of Certification confirming that a qualified third party has performed an independent, external certification verifying that the model’s underlying methodologies and mathematics are accurate, in line with industry standards and consistent with the vendor documentation.
  • Validation — This is a biggie as it can be costly, particularly for smaller institutions. However, using non-validated models to manage risks to the institution is potentially an unsafe and unsound practice — using regulatory speak. Even when the risk is not particularly material, the reliance on non-validated models can be a costly business practice. The assessment of the costs and benefits of model validations is subjective and context-driven and is the responsibility of bank management. To promote a sound process, regulators (the Office of the Comptroller of the Currency) provided a summary list of expectations, and suggested that formal policies and bank practices address the following:
  • Decision makers should understand the meaning and limitation of a model’s results. Where the models are too abstract for non-specialists to understand the underlying theory, the bank must have a model reporting system in place that transforms the model’s outputs into useful decision-making information without disguising the model’s inevitable limitations.
  • The bank should demonstrate a reasonable effort to audit the information inputs to the model. Input errors should be addressed in a timely manner.
  • The seniority of the management overseeing the modeling process should be commensurate with the materiality of the risk from the line of business in process.
  • To the extent feasible, model audits must be independent from model construction.
  • Responsibilities for the various elements of the model-audit process must be clearly defined.
  • Modeling software should be subject to change control procedures, so that developers and users do not have the ability to change code without review and approval by an independent party.

This laundry list of deliverables may seem exhaustive. As such, seek guidance from your ALM team. They’ll have the skillset necessary to break down the list into manageable tasks.

  • Sign Up

  • Categories

  • Archive

Software: Kryptronic eCommerce, Copyright 1999-2019 Kryptronic, Inc. Exec Time: 0.067 Seconds Memory Usage: 3.807922 Megabytes