Two lenses on the same exposure
Net interest income, or NII, measures interest earned less interest paid over a period. An interest-rate simulation estimates how that income might change under alternative rate paths. Economic value of equity, or EVE, instead considers the present value of asset cash flows less liability cash flows, with relevant off-balance-sheet positions. It is a modeled economic measure, not the bank’s accounting equity or market capitalization.
Interagency interest-rate-risk guidance discusses both earnings and economic-value perspectives. The OCC’s July 2004 embedded-options guidance similarly emphasizes short-term earnings and longer-term value. Those are longstanding analytical principles, not a claim that one standardized stress test predicts every bank’s outcome. The two measures can disagree because they cover different horizons and react differently to cash-flow timing.
Why higher earnings can accompany lower value
A bank may earn more immediately when floating-rate loans reset upward while deposit rates rise slowly. Yet its fixed-rate assets can lose economic value as market discount rates increase. The near-term earnings gain and longer-term value decline can both be real within the model. Neither result invalidates the other.
The size of the effects depends on the balance sheet. Fixed-rate debt can offset some asset-value sensitivity; stable low-cost deposits can have economic value that depends heavily on behavioral assumptions. A conclusion based only on loan repricing or only on securities marks misses those interactions. The useful question is which exposures drive each result and whether the assumed funding behavior remains plausible.
A hypothetical comparison
Assume a bank has $100 million of floating-rate assets that reprice immediately after a one-percentage-point market-rate increase. If its related funding cost initially rises only 0.3 percentage point, the simple annualized NII benefit is $700,000 before other changes. This is an illustrative repricing calculation, not a forecast.
Separately, assume $100 million of fixed-rate assets has an approximate duration of four years and the related $100 million of liabilities has duration of one year. A small one-percentage-point parallel rate rise implies roughly a $4 million asset-value decline and a $1 million liability-value decline, or a $3 million reduction in net economic value before convexity and other effects. The two examples isolate mechanisms; a full bank model must combine all positions without treating them as independent balance sheets.
Assumptions can dominate the answer
measures how much deposit pricing responds to market rates; decay assumptions address how balances leave over time. Both can change under competitive pressure. Treating nonmaturity deposits as permanently cheap and stable can make the modeled economic position look stronger than the business reality. Conversely, assuming every deposit immediately reprices and leaves can overstate risk.
Loan prepayments create another sensitivity. When rates fall, borrowers may refinance attractive high-coupon loans, shortening the bank’s income stream. When rates rise, prepayments can slow and extend low-coupon exposure. Rate floors, caps and customer choices can also make the response nonlinear. A single parallel shock is useful but insufficient when the institution has material basis or yield-curve exposure.
Static and dynamic balance sheets
A static simulation generally isolates the existing portfolio under specified runoff or replacement assumptions. A dynamic simulation adds anticipated business growth, pricing and management actions. Both can be useful, but they should not be presented as the same experiment. A favorable growth plan can conceal weakness in the existing book if the reader is not told what new business the model assumes.
Recommended reporting shows the contribution from the current portfolio separately from planned management actions. Test whether those actions would remain feasible under the scenario. A plan to replace deposits cheaply may be inconsistent with an environment in which customers are moving to higher-yielding alternatives. A hedge assumed to be available at a convenient price may become more expensive when it is needed.
Controls, costs and interpretation
The bank should reconcile model inputs to financial records, review key behavioral assumptions and back-test forecast income against realized results. Differences should be explained by volume, pricing, timing and model error. Limit reporting should show both the magnitude of exposure and the assumptions driving it, rather than only a green or red status.
Hedging can reduce one exposure while creating another cost or dependency. A swap may alter rate sensitivity but introduce collateral, counterparty or accounting considerations. Extending funding maturity can protect and reduce earnings flexibility. EVE and NII are therefore decision inputs for balancing tradeoffs, not instructions to eliminate every modeled fluctuation regardless of cost.
Evidence that would change the view
Confidence should increase when deposit and prepayment assumptions remain consistent with observed behavior across different rate environments and when independent review can reproduce the results. It should weaken when the favorable case relies on untested customer persistence, unexplained overrides or growth assumptions that have repeatedly missed.
For readers comparing banks, a quoted NII sensitivity needs its horizon, rate scenario and balance-sheet assumptions. An EVE percentage needs its denominator and treatment of deposits. Without those definitions, apparently similar figures can describe different risks. The practical conclusion is to use both lenses and trace their disagreement back to cash flows, rather than selecting whichever measure tells the more comfortable story.