Dynamic IRRBB: Beyond Static Balance Sheet Assumptions

The standard IRRBB framework assumes a static balance sheet — maturing contracts run off, no new business is originated, and the bank's size shrinks over time. This simplification makes measurement tractable and comparable across banks. But it also means the IRRBB number describes a bank that doesn't exist — a bank that stops doing business the moment the rate shock hits.

The Static Assumption

Under the static balance sheet assumption:

• Maturing assets are not replaced. A $100M loan that matures in 6 months simply disappears from the balance sheet. The bank doesn't originate a new loan to replace it.
• Maturing liabilities are not renewed. A $200M CD that matures is repaid and not rolled over. Funding shrinks.
• NMD deposits follow their behavioral decay schedule. Deposits gradually run off according to the core/volatile model, but no new deposits arrive.
• Prepayments are reinvested at the risk-free rate (or simply run off, depending on the convention).

The result: the bank's balance sheet shrinks over time, with asset and liability volumes declining as instruments mature or run off. By year 10, a significant portion of the original balance sheet may have disappeared.

Why Static Is the Default

Despite its obvious unrealism, the static assumption is the regulatory default — and for good reasons:

The static approach answers the question: "Given the bank's current exposures, how much would it lose if rates moved and it did nothing?" This is a clean, well-defined question with a clear answer.

What Static Misses

The problem is that banks don't "do nothing." They originate loans, take deposits, adjust pricing, and manage their balance sheet actively. The static assumption misses several important dynamics:

1. New production changes the risk profile. A bank that originates $500M of mortgages per quarter will have a very different balance sheet in 12 months than the static runoff suggests. If new mortgages are originated at current market rates (not the old coupon), the repricing profile changes — potentially reducing or amplifying the ΔNII impact.

2. Reinvestment at market rates matters for ΔNII. When a 5% loan matures in a 4% rate environment, the static model shows the income disappearing. A dynamic model shows it being replaced by a new loan at 4% — a loss relative to the old coupon, but not a complete loss of income. The ΔNII difference can be substantial.

3. Deposit volume responds to rates. In a rising rate environment, deposit balances may grow (higher yields attract savings) or shrink (customers move to money market funds). The static assumption ignores this, holding deposit balances constant except for behavioral decay.

4. Pricing decisions are endogenous. When rates rise, banks typically widen loan spreads (partially offsetting the funding cost increase) and adjust deposit rates with a lag (the beta effect, but as a deliberate pricing strategy, not just a statistical regularity). Static models miss this management response.

The Dynamic Approach

A dynamic IRRBB framework introduces assumptions about how the balance sheet evolves over the projection horizon:

The key insight: dynamic assumptions primarily affect ΔNII (which depends on what happens over the next 12 months, where new production matters most) more than ΔEVE (which depends on the present value of all future cashflows, where new production adds both assets and offsetting liabilities).

Deposit Migration

One of the most important dynamic effects is deposit migration — the movement of balances between deposit segments in response to rate changes.

Migration modeling requires assumptions about:

• Trigger thresholds: At what rate level do customers begin migrating? (Typically when the spread between market rates and deposit rates exceeds a threshold — say, 200bp.)
• Migration speed: How quickly do balances shift? (Gradual over months, or sudden?)
• Volume effects: Does total deposit volume change, or just the mix between segments?
• External competition: Do depositors leave the bank entirely for money market funds or Treasury bills?

The SVB episode in March 2023 was a stark reminder that deposit migration can be much faster and larger than historical models suggested — driven by social media, mobile banking, and the unprecedented gap between deposit rates and Treasury yields.

Static vs Dynamic: An Example

Governance Challenges

Dynamic IRRBB introduces a new class of governance questions that don't arise in the static framework:

1. Who owns the assumptions? New production volume, pricing spreads, and deposit migration parameters are business decisions, not risk model parameters. Should the business lines provide them? The CFO's planning team? Or should risk management set conservative defaults?

2. How are assumptions challenged? A business line that provides optimistic new production volumes will produce a favorable dynamic ΔNII. Independent challenge — by risk, audit, or an external validator — is essential but difficult when the assumptions are inherently forward-looking.

3. Stress consistency. If the rate scenario assumes +300bp of tightening, the new production assumptions should reflect a slowdown in loan demand (higher rates reduce borrowing). An internally inconsistent scenario — extreme rate stress with business-as-usual production — is worse than a static approach.

4. Regulatory skepticism. Regulators are cautious about dynamic approaches because they can be used to make IRRBB numbers look better. A bank that shows a small dynamic ΔNII under stress may simply have optimistic production assumptions rather than genuinely lower risk. This is why most regulatory frameworks use static balance sheet for the outlier test and reserve dynamic analysis for internal management purposes.

The Bigger Picture

The static vs dynamic question is not about which is "correct" — both are useful abstractions that answer different questions:

• Static IRRBB answers: "What is the bank's inherent rate sensitivity given its current positions?" — a clean, comparable, conservative measure.
• Dynamic IRRBB answers: "What would actually happen to earnings and value if rates moved and the bank continued operating?" — a more realistic but assumption-dependent measure.

The dynamic framework also connects IRRBB to broader strategic planning: pricing decisions, growth targets, funding strategy, and hedging all interact in a dynamic world. A bank that understands these interactions can make better trade-offs — accepting some ΔNII volatility in exchange for franchise growth, or sacrificing some near-term earnings to build a more resilient long-term position. These are the decisions that separate competent risk management from genuine strategic advantage.