In the third and final component of this series of articles, we will discuss some regulatory and better practice approaches to A/L model validation and ALM policies and procedures. In the previous articles, we discussed the following components:
- Input
- Processing
- Assumptions
- Output
The final components for discussion will be model validation and ALM policies and procedures.
Model Validation
Model validation usually involves an independent review of all of the modeling processes (including ALM policies) that we have previously noted. To that end, managers must focus on the independence and competence of the entity responsible for performing the validation. Most community banks utilize an external party to perform this process, as a thorough model validation is time consuming and requires a specialized skill set.
As we mentioned, regulators are pressing institutions to step up their A/L management in terms of market risk modeling, ALM management, and board awareness. Your primary regulatory authority, on a periodic or yearly basis depending on how your institution is chartered, requires model validation. A thorough validation usually follows the procedures detailed in OCC Bulletin 2000-16. In addition, recent guidance has been provided by the FDIC titled "Model Governance," and information can also be found in OTS TB-13a. All of these issuances emphasize the importance of independent validation and competence of the reviewer. However, none of the issuances seem to reflect or emphasize the strategic aspects of having a "second look" at your modeling practice. This is exactly what an institution's board and management should be considering when weighing the cost versus inherent value of a model validation.
Oftentimes, we hear clients question why a validation is necessary, especially if they outsource their ALM reporting services. However, if you look at the review from a different perspective-for instance, if you do outsource your reporting service-wouldn't you want an independent third party to verify the accuracy, granularity, and robustness of your institution's ALM modeling process?
When performing model validations, we frequently find problems with data entry, EVE assumptions, and general cash flow assumptions (liquidity). We also identify "better practice" approaches, ways to quantify versus qualify assumptions, and point out various benchmarks that improve the accuracy of reported A/L exposures. This is where the strategic benefits of a model review can enhance your modeling capabilities (knowing there are certain limitations to all models) and improve your ALM risk management. As an example, if management uses the model to project your annual budget goals, wouldn't it be valuable to understand that the model is, in fact, set up properly and simulating your risk exposures accurately based on your growth targets for the next fiscal year, especially for exchange-listed institutions that manage to an earnings-per-share target?
Your A/L model is a powerful tool that can help pay for itself, and then some, if used strategically!
ALM Policies
Finally, we discuss the cornerstone of the ALM risk management framework - policies and risk limits.
Policies are the ultimate responsibility of directors!
As stated previously for boards, bankers, and ALM practitioners, it is difficult, if not impossible, to understand the proprietary engines behind most modeling systems. However, developing a solid control function around the use of these risk management tools is not.
ALM policies provide the control function that is vital to all of the components we have discussed throughout this series of articles. The board should use their ALCO policy framework to establish a clear line of responsibility and expectations.
While many times executive management and outside consultants will write policies for an institution, we often notice that bank boards act more like a "rubber stamp," if you will, in terms of approving these ALCO-related policies. Generally, this is caused by a lack of knowledge in terms of subject matter such as economic capital or the EVE measurement.
This is exactly why board members should consider furthering their understanding of these topics through increased education and awareness. Why? Because within these policies are risk limits that are used to manage your sensitivity to market risk and serve as your risk controls. Typical risk limits are established for earnings at risk, EVE, and liquidity. This is where education and a well-defined set of policies can actually improve your ability to manage the bank.
We encourage board members and executive management to take charge of educating themselves and taking an active approach to ALM management, especially in terms of modeling risk.