Interest Rate Risk Modeling: Static or Dynamic ... or both?
- Ryan Gilles
- Oct 7
- 8 min read
With all of the discussion of how dynamic interest rate risk models can help institutions forecast outcomes under different rate scenarios, running static models (that assume no growth or mix shifts) may seem outdated.
Yet regulators such as the OCC, FDIC, and NCUA continue to define static models as the baseline approach for best practice.
Why?
Perhaps because static models may serve as a powerful anchor: they strip away the noise of constant change and represent the “simple” relationship between a financial institution’s existing balance sheet and potential interest rate changes.
But is a “static” model really the best way to capture the inherent risk resident within a balance sheet?
In reality, no institution ever operates with a truly ‘static’ balance sheet. But by returning to this simplified baseline quarter after quarter, management may gain a clear point from which to test assumptions, pressure-check strategy, and help ensure that long-term decisions align with the institution’s evolving risk profile.
A clear alternative to “static” would be “dynamic” simulations. While these simulations may have more “moving parts” and appear more complicated, the reality is that they offer a potential look into the outcomes associated with anticipated changes that might occur as interest rates change.
So, what should a balance sheet manager focus on?
In short, the answer may be both!
This article will explore the pros and cons associated with both approaches and shed light on what really matters when evaluating the risk profile of your institution.
From the Editor
"There is a lot of clarity in hindsight." – Julia Hartz
Too bad bankers don’t get the benefit of hindsight when managing their businesses in real time. After all, wouldn’t it be nice to be able to get a “do over” for a specific strategy that did not materialize as expected.
While some may claim they know exactly how an environment will unfold, the reality is that the outcomes rarely match precisely what is expected.
There are lot of financial institutions that would like a “do over” in the wake of the COVID period. The rapid rise in inflation which led to significantly higher rates created margin pressures still being felt by many in the industry today.
But what if I told you, perhaps there is a better way to evaluate interest rate risk outcomes. And what if it’s not actually just one method, but the combination of disparate approaches.
In this month’s Bulletin, DCG Director Ryan Gilles identifies the utility of modeling both “static” and “dynamic” interest rate simulations. He outlines the pros and cons of each scenario and how your management team may interpret the results of both.
Admittedly, getting a management team to believe in the probability of a 500+bps increase in interest rates back in 2020/21 would have been a tough sell. (Particularly given the Federal Reserve chief himself claimed that inflation was just “transitory.”)
However, understanding all the potential pathways for earnings is paramount in ensuring your team understands the full range of outcomes in the future.
Vinny Clevenger, Managing Director
So, what makes an IRR model static?
Static modeling may be less about being fully realistic and more about providing a solid starting point in terms of:
Cash Flow Assumptions: All contractual amortization, maturities, and assumed prepayment cash flows are reinvested into like products at current market rates. This maintains the size of the balance sheet over time.
No Growth or Mix Shifts: The model excludes any projected changes in loan, investment, or deposit volumes and product mix.
Point-in-Time Snapshot: The analysis uses the balance sheet exactly as it exists on the as-of date, then applies interest rate shocks or ramps (up or down) to assess potential risk exposures.
Focus on Current Profile: The model isolates the interest rate sensitivity of the balance sheet in its current form, without incorporating assumptions about future business activity.
Potential Challenges of Static IRR Modeling
While static modeling remains the regulatory best-practice baseline, it carries two inherent limitations.
First, they may offer incomplete strategic insight. Static models do not capture how a bank’s risk profile could evolve under forecasted growth or mix shifts tied to future strategies and initiatives.
And second, static modeling does not capture customer behavior changes in different rate environments (e.g., migration of non-maturity deposits into CDs as rates rise), which can materially alter risk exposures.
These shortcomings became more evident over the past five years with the surge in low-cost deposits during the COVID-19 pandemic that eventually accounted for 26% of total NMDs by March 2022 (see my colleague Michael Hunker’s article The Evolution of COVID Surge Deposits: From Stimulus to Stability for some great insight and analysis on this very topic).
And shortly thereafter, the FOMC’s rapid 525bp rising rate cycle from 2022-2024 placed tremendous pressure on both deposit rates and balances in reaction.
According to Deposits360°® Cross-Institution data, the industry saw a 10% shift from non-maturity deposits into CDs beginning in March 2022, when the Fed started raising rates.
Source: Darling Consulting Group Deposits360°®
From an interest rate risk perspective, these two events had significant impacts on financial institutions’ balance sheets, affecting sensitivity measures and earnings projections in ways that static modeling could not have accurately captured or predicted.
So, the obvious question becomes, would dynamic modeling simulations have been able to better help financial institutions forecast the potential range of outcomes that could have occurred in the aftermath of those events?
Why Run Dynamic Models?
Dynamic interest rate risk models provide a comprehensive view of how a financial institution’s risk profile may evolve over time. By incorporating assumptions such as forecasted growth, product mix shifts, and residual funding cost changes (even under a stable rate environment), dynamic models offer insights beyond the static baseline.
These models can allow management to evaluate how strategic plans and future initiatives could interact with unexpected market conditions, helping leaders better anticipate the range of potential outcomes and make more informed decisions.
So, what makes an IRR model dynamic?
Balance Sheet Growth or Contraction: Incorporates growth or contraction in asset size based on budgets, new strategic initiatives, or even expanding a branch/market network.
Product Mix Shifts: Projects changes in the composition of assets (investment cashflow being used to fund loans rather than reinvested back into securities) or funding (migration from non-maturity deposits into CDs or assuming deposit outflows that will be offset with additional wholesale).
Non-Linear Pricing Assumptions: Assumes loan spreads will widen or tighten under different market rate conditions rather than keeping spreads consistent. Or deposit pricing betas that become progressively more sensitive to market rates as they rise.
Potential Challenges of Dynamic IRR Modeling
There is an illustration that we use for DCG's forecasted deposit study presentations that does a fantastic job of outlining what dynamic models are and are not…with the short answer being that they are not a crystal ball…
Source: Darling Consulting Group
While dynamic assumptions can add significant depth to interest rate risk modeling, institutions should always keep some important considerations in mind:
Data Support and Defensibility. All dynamic assumptions, whether growth projections, product mix changes, or rate behaviors, must be supported by reliable, defendable data. Overly optimistic or unrealistic assumptions can obscure risk and undermine the model’s credibility. Although assumptions matter in static modeling as well, their impact can be magnified when applied dynamically.
Operational Complexity. Dynamic models are inherently more complex than static models, requiring greater operational resources to execute effectively. The most significant demand lies in the data analysis needed to validate and support assumptions.
Limits of Predictive Power. Dynamic assumptions cannot anticipate unpredictable events such as black swan shocks (e.g., the Global Financial Crisis or COVID-19), government interventions, management changes, or the ultimate success of new business strategies. The most effective way to evaluate the potential impact of such events on the risk profile is through rigorous stress testing.
The Importance of Reliable Assumptions
Whether running static, dynamic, or both types of interest rate risk models, the value of the results ultimately depends on the quality of the assumptions. One of the sayings I always use with clients is “garbage in, garbage out”: if the assumptions or data you use for modeling are not strong and reliable, then you shouldn’t expect your results to be strong and reliable either. Every IRR model contains hundreds, if not thousands of assumptions, and while it is unrealistic to expect that every assumption be perfectly accurate, certain key drivers carry disproportionate weight in determining results.
Non-Maturity Deposit (NMD) Assumptions. Institution-specific pricing betas and average life/decay assumptions at the individual product level are essential. Given that NMDs often comprise 50 –70% of a balance sheet’s funding mix, even small changes in these assumptions can materially impact results. With today’s technology and data analytics, having robust institution-specific NMD assumptions should be non-negotiable.
Cash Flow Replacement Assumptions. Reinvestment assumptions should start with a review of recent loan production data (structures, terms, and rates/spreads) while deposit replacement assumptions should reflect current offering rates and, for term funding, the maturities experiencing the most activity. Relying solely on loan rate sheets often leads to inflated replacement rates, since actual pricing is shaped by factors such as relationship considerations, loan size, and borrower credit quality. Ensuring assumptions align with real market activity makes results far more reliable.
Prepayment Assumptions. Historical data across multiple rate environments show a clear pattern: prepayments accelerate when rates fall and slow when rates rise. Borrowers refinance when they can lower costs but hold onto loans when refinancing would increase costs. Reliable prepayment assumptions should reflect this rate sensitivity while also accounting for loan balance sizes and coupon ranges. Doing so improves the accuracy of cash flow projections and strengthens insights into how portfolio composition interacts with market environments, ultimately enhancing lending strategy and execution.
Static and Dynamic Modeling in Combination May Be the Best Approach
We believe that to gain a complete view of risk and earnings potential, and to maximize overall ALCO and risk management processes, institutions should consider employing both static and dynamic models.
After all, FDIC Interagency guidance on Interest Rate Risk Management from 2010 directly notes that institutions should use both approaches: “When performing dynamic simulations, institutions should also run static simulations to provide ALCO or senior management a complete and comparative description of the institution’s IRR exposure.”
Static models serve as the baseline or benchmark, providing a clear snapshot of the current risk profile. This consistency helps management make disciplined decisions aligned with established risk tolerances.
Dynamic models extend the analysis by showing how growth, product mix shifts, pricing changes, and balance sensitivities could affect future outcomes under different rate scenarios. This forward-looking perspective helps identify the full range of potential results, ensuring strategic plans are tailored to optimize earnings while guarding against unintended shifts in the risk profile.
Using static and dynamic modeling together may offer both clarity and perspective. A combined approach provides a foundation for informed decision-making today while helping ensure that you are prepared for tomorrow’s possibilities and unknowns.
For more insights from Darling Consulting Group, click here.
Ryan Gilles is a Director at Darling Consulting Group, where he assists community financial institutions throughout the U.S. to solidify and enhance balance sheet management. In this role, he works collaboratively with executive teams and ALCO committees to help develop and implement tailor-made strategies related to interest rate risk, liquidity, and capital while prudently managing risk to optimize earnings and satisfy regulatory compliance.
Ryan began his career at DCG in 2013 as a financial analyst and continues to work closely with the implementation and ongoing education of DCG’s decision-support tools (Deposits360°® and Loans360°®) as well as onboarding and building out new client ALCO models with the New Client Implementation team. He currently lives in South Boston with his wife and is a graduate of Assumption University with a B.A. in Accounting.
© 2025 Darling Consulting Group, Inc.
