Several key IRR modeling practices were thrust into the spotlight by the most recent IRR advisory. Many of the more popular practices have made their way to the examiner’s checklist, including the concept of a static versus dynamic forecast. Here’s an excerpt from the advisory:
In general, simulation models can be either static or dynamic. Static simulation models are based on current exposures and assume a constant balance sheet with no new growth. In contrast, dynamic simulation models rely on detailed assumptions regarding changes in existing business lines, new business, and changes in management and customer behavior.
Most people read the above passage and assume that static means “no-growth” and dynamic means “growth”.
That’s almost right. It is generally accepted that static means “no-growth” or a constant balance sheet. But dynamic doesn’t necessarily mean “growth”. It really means “changes” in the balance sheet. One change could certainly be growth, but it could also mean a decrease in portfolio size either by letting the portfolio run-off or perhaps (in the case of bonds) selling certain investments.
It might also mean changing the nature of a portfolio, not by a change in volume per se, but maybe by changing a portfolio’s duration. For example, suppose a bank has $10 million in brokered deposits maturing next quarter. They could model, in a dynamic forecast, a similar replacement volume of $10 million but target a different maturity term, say three years.
Many banks use their budget or strategic plan in their IRR modeling. Typically these types of forecasts include new loan, deposit, and even equity growth. These are considered dynamic forecasts. The results of a dynamic forecast depend on the number of changes, like the amount of growth, or the nature of new pricing. These results are possibly driven by key assumptions. It is these “key assumptions” that the advisory mentions directly:
…dynamic simulation is highly dependent on key variables and assumptions that are extremely difficult to project with accuracy over an extended period. Furthermore, model assumptions can potentially hide certain key underlying risk exposures.
To minimize the impact such key assumptions might have they suggest a bank also analyze their IRR using a static forecast.
Another important thing to recognize is that when you’re talking about static versus dynamic you’re talking about the earnings simulation. It doesn’t apply to the EVE (economic value of equity) simulation. It is the forecast that is static or dynamic, referring to the changes in balance sheet volumes, mix, or duration (or all three). The behavior of future dollars (i.e. balances not on the balance sheet today) will not impact the EVE simulation which is run as of a point in time. This is a common misconception. I often hear examiners and auditors ask for, “a model rerun that shows a static earnings and EVE simulation”. This request for a static EVE simulation doesn’t make sense (much the same way that a “A 2-year or 24-month EVE simulation just doesn’t make sense”.)
(This post is part of a series which provides a basic overview and discussion of interest rate risk stress-testing.)
(follow-up post: The cover-up)