The IRR Framework: NII vs EVE

Why This Matters

Interest rate risk (IRR) is the exposure that changes in interest rates will reduce your earnings or economic value. But here's the catch: a rate move that kills earnings in the next quarter might improve your economic value over a 10-year horizon, and vice versa. The fundamental insight of modern ALM is that you need to track both perspectives — Net Interest Income (NII) sensitivity in the near term and Economic Value of Equity (EVE) sensitivity over longer horizons.

This distinction isn't academic. It's why a bank that looks "protected" on NII might still blow up on earnings surprises, and why a Treasury team chasing EVE optimization can accidentally destroy shareholder value. You need to manage both, understand the tension between them, and make deliberate trade-offs.

The Two Frameworks at a Glance

Net Interest Income (NII): The Earnings Lens

NII is straightforward: it's the difference between the interest you collect on loans and securities and the interest you pay on deposits and borrowings. IRR on the NII side means answering: "If rates move tomorrow, how much does my earnings for the next 12 months change?"

NII sensitivity captures the repricing mismatches in your balance sheet. If you have USD 100B of floating-rate loans repricing in Q2 and USD 80B of floating-rate deposits repricing in Q3, you've got an earnings exposure. When rates rise, your cost of funds will catch up to your loan yields — just late enough to squeeze NII.

The NII framework is intuitive because earnings matter. When your bank misses NII guidance, your stock gets hit. Regulators care about it. Board members understand it. But it's also myopic: it ignores value creation (and destruction) beyond the horizon you're measuring.

Economic Value of Equity (EVE): The Value Lens

EVE is the present value of all future cash flows from the balance sheet, minus the present value of all funding costs. It's essentially the "economic" book value of equity — what the balance sheet is really worth if we mark everything to fair value and run it off.

IRR on the EVE side means: "If rates move tomorrow and stay there, what happens to the discounted value of all my future cash flows?" This captures longer-dated exposures, optionality embedded in deposits and mortgages, and the reinvestment risk when your portfolio matures.

EVE is harder to explain to a board, but it's more economically complete. A rate shock that extends deposit duration and crushes your EVE is a real economic problem — you've just become a longer-dated bond at a time when bond valuations compressed.

The Fundamental Tension

Here's where it gets interesting. The same balance sheet position can have opposite implications for NII vs EVE:

Example: You fund a 10-year fixed-rate loan book with floating-rate deposits. Your NII is protected: deposit rates haven't moved yet, so NII is high. But your EVE is vulnerable: if rates rise today, your fixed-rate loans are locked in at yesterday's rates, while the discount rate used to value future cash flows goes up. EVE falls sharply.

Conversely: You lock in floating-rate deposits at a 3% beta to SOFR for 5 years. Your NII is suddenly exposed (costs rise immediately with rates), but your EVE is protected because you've fixed your cost of funds.

How Practitioners Use Both

Senior ALM teams maintain a two-axis risk framework:

1. NII sensitivity: Typically measured as the change in NII under a ±100bp parallel rate shock over the next 12 months. Regulatory guidance (Federal Reserve, OCC) suggests a decline of more than 10-12% of net interest margin is concerning. Most large banks aim to keep NII sensitivity within ±5% under a 100bp shock.

2. EVE sensitivity: Measured as the percentage change in the present value of equity under the same rate shocks. A ±10% swing in EVE under a 100bp shock is typical; some banks are tighter (±5%), others accept broader ranges depending on strategy.

The trade-off is real: reducing NII sensitivity often means shortening the funding side, which increases EVE sensitivity. Reducing EVE sensitivity often means extending long-term borrowings, which increases funding costs and pressures NII.

Regulatory and Strategic Context

Regulators focus on both. The Federal Reserve's stress testing framework (CCAR/DFAST) includes NII projections under various rate paths. But the Fundamental Review of the Trading Book (FRTB) and banking book guidance increasingly emphasize EVE-like metrics: what's the true economic loss if rates move?

For strategy, the frameworks answer different questions:

• NII management drives earnings guidance, executive compensation, and quarterly results.


• EVE management drives long-term shareholder value, protects against tail risks, and often constrains leverage and business mix.

A bank that optimizes NII at the expense of EVE might have strong earnings for 2–3 years, then get hammered when the rate cycle turns. A bank that's obsessed with EVE protection might miss earnings opportunities and underperform peers.

What You'll Learn in This Module

In the sections ahead, we'll dive deep into:

• How gap analysis (the original IRR tool) captures repricing mismatches for NII


• How NII simulation models actually work and why behavioral assumptions matter so much


• The economics of EVE and why the discount rate and reinvestment assumptions can swing your results


• The practical mechanics: key rate duration, basis risk, convexity, and optionality


• How to build, validate, and use these models to make real portfolio decisions


• How to read peer disclosures and benchmark your risk profile

By the end, you'll understand why a 50bp rate move is sometimes meaningless and sometimes catastrophic — it depends on which lens you're looking through, and whether your model actually captures what will really happen to your balance sheet.

The IRR Framework: NII vs EVE — Deep Dive

Part 1: The Economics of NII Sensitivity

Repricing Mechanics

At its core, NII sensitivity is about repricing gaps — the timing mismatches between when your assets reprice and when your liabilities reprice.

Consider a stylized bank balance sheet:

| Item | Amount | Maturity/Repricing | Rate |
|------|--------|--------------------|---------|
| Mortgages (30-year fixed) | $50B | 30 years | 6.5% |
| Auto loans (5-year ARM, resets annually) | $10B | 1 year | 8.0% |
| C&I loans (floating, SOFR + 250bps) | $20B | Quarterly | SOFR + 250bps |
| Securities | $15B | Mixed (2–10 year avg) | 4.5% avg |
| Customer deposits (NOW, ~30bp beta) | $60B | Daily repricing | 4.5% (current) |
| Wholesale funding | $20B | 3 months | SOFR + 90bps |
| Equity | $10B | — | — |

Under a +100bp shock:

• First year impact: The mortgage portfolio (50B @ 6.5%) doesn't reprice — you keep earning 6.5%. The C&I and auto loans reprice at +100bp, adding $300M of revenue. Deposits may not fully reprice (the 30bp beta is conservative; competitive pressure might push it to 50bp), so you lose maybe $200–300M on the deposit side. The net in year 1: roughly break-even to slightly positive on NII.

• Why? You have a repricing asset gap: assets repricing slowly (mortgages dominate) relative to liabilities repricing fast (deposits tied to market rates).

This is the positive carry effect: when you have a long asset duration and short liability duration, and rates are rising, you initially gain NII because your funding costs haven't caught up yet.

But here's the catch: that positive effect is temporary. In year 2, if rates stay elevated, deposit repricing might accelerate further, and now you're fighting a headwind. In year 3, as mortgages age and new originations come on board at higher rates, that effect compounds.

Duration of NII

A practical insight: NII sensitivity is a function of the weighted-average repricing duration of assets minus liabilities.

If your assets reprice (on average) in 1.5 years and your liabilities reprice in 0.5 years, your "NII repricing gap" is 1 year. This means you have roughly 1 year of exposure to parallel rate shocks.

Regulators and most ALM teams measure NII sensitivity over a 12-month horizon because this is the furthest you can reasonably project deposit behavior and borrowing costs. Beyond 12 months, assumptions become too speculative.

Behavioral Betas and the Non-Linear Problem

Here's where theory meets messy reality: deposit betas are not constant.

A 30bp beta to SOFR works fine when rates are moving slowly and competition for deposits is moderate. But in a sharp +200bp shock, your customers get angry. They move money to money-market funds. Competitors raise rates aggressively. Your actual beta might become 80–100bp even though your "average" beta is 30bp.

Conversely, in a -100bp shock, you might drop customer rates by only 10bp (you can't go below zero), so your beta becomes near-zero on the downside.

This convexity of NII is real and matters hugely for risk measurement. A symmetric ±100bp shock doesn't actually stress your balance sheet symmetrically. You need to test both directions and measure the asymmetry. Most banks find they're more vulnerable to rate rises than rate falls (because deposits don't reprice downward as much).

Why NII Matters

1. It drives earnings: NII is typically 50–80% of a bank's operating revenue. A 5% swing in NII is a material earnings miss.
2. It constrains strategy: If you're NII-constrained (most community and mid-size banks are), you can't make long-term funding decisions without hitting your NII targets. You can't fund a 10-year portfolio with short-term deposits without squeezing earnings.
3. It's visible: Wall Street tracks NII expectations closely. Missing NII guidance is expensive in terms of stock multiple.

Part 2: The Economics of EVE Sensitivity

Present Value and the Discount Rate

EVE is calculated as:

EVE = PV(All future cash inflows) - PV(All future cash outflows)

The discount rate is typically the current risk-free rate (Treasury curve) plus a spread that reflects the bank's cost of equity and funding.

Suppose your balance sheet generates expected net cash flows like this:

| Year | Cash Flow | Discount Factor (3% rate) | PV at 3% | Discount Factor (4% rate) | PV at 4% |
|------|-----------|---------------------------|----------|---------------------------|----------|
| 1 | $200M | 0.971 | $194.1M | 0.962 | $192.3M |
| 2 | $250M | 0.943 | $235.7M | 0.925 | $231.2M |
| 3 | $300M | 0.915 | $274.5M | 0.889 | $266.7M |
| 4–10 | $2,500M (total) | 5.797 | $14,492.5M | 5.242 | $13,105.0M |
| Total EVE | — | — | $15,196.8M | — | $13,795.2M |

Notice: A 100bp increase in the discount rate (from 3% to 4%) reduces EVE by $1,401.6M, or about 9.2%.

This is the interest rate sensitivity of EVE. When rates rise:
1. Your assets' future cash flows are worth less in present-value terms (discount rate effect).
2. Your liabilities' future cash flows are also worth less (but the net effect depends on the duration mismatch).
3. If assets have longer duration than liabilities, EVE falls. If liabilities have longer duration, EVE rises.

Duration Dominance

For most banks, assets have significantly longer duration than liabilities:

• Mortgage books: 5–8 year duration


• Securities: 3–6 year duration


• Loans: 1–4 year duration (on average, shorter if floating rate)


• Deposits: 0.5–1.5 year duration (some demand deposits have near-zero duration)


• Wholesale funding: 0.3–2 year duration

The result: a typical bank's EVE is negatively convex to rates. When rates rise, EVE falls sharply. When rates fall, EVE rises, but usually not as much (because deposit values are bounded by the zero lower bound — you can't pay negative rates).

Real Example: The 2022 Regional Bank Crisis

Consider Silicon Valley Bank (SVB) in early 2022:

• Balance sheet: ~$120B in assets, heavily weighted to securities (mostly Treasuries and MBS) with an average duration of ~6 years


• Funding: ~$90B in deposits, mostly from tech startups and venture portfolios, with ultra-low rate sensitivity


• EVE position: In Q4 2021, with 10-year Treasuries at ~1.5%, SVB's EVE was optimized. But there was massive duration risk.

When the Fed began raising rates:

• By March 2023, 10-year Treasuries hit 3.5%


• SVB's securities portfolio, marked to market, had lost ~$15–16B in value (roughly 13% of equity)


• EVE collapsed. The unrealized losses became realized when depositors panicked and the bank had to sell securities at a loss to fund withdrawals.

This is a textbook EVE problem: you had positive NII (rates rising, floating deposits haven't repriced as much as loan/securities yields), but massively negative EVE (your long-duration assets marked to current rates were worth far less).

Regulators missed it partly because they focus on NII, not EVE. The Fed's own models (CCAR) capture some of it, but stress testing is backward-looking; it didn't anticipate a 200bp rate shock combined with a deposit flight.

Why EVE Matters

1. It captures long-term value: EVE reflects the true economic cost of today's decisions. Locking in low-duration funding today protects tomorrow's shareholders.
2. It's stable: Unlike NII, which is forecast-dependent, EVE is based on current market prices and reasonable assumptions. It's hard to game.
3. Regulators increasingly care: The Federal Reserve's banking book guidance, CCAR market risk metrics, and Basel III endgame proposals all move toward marking more of the balance sheet to fair value. EVE-like thinking is the future.

Part 3: The Trade-Off in Practice

Why You Can't Optimize Both Simultaneously

Suppose you want to reduce NII sensitivity. Your options:
1. Shorten liability duration: Issue short-term CDs, reduce deposit retention targets, rely more on wholesale funding. But now you're refinancing constantly, and you're exposed to funding shocks. Also, in a falling-rate environment, your funding costs might lock in high, hurting NII.
2. Lengthen asset duration: Shift from floating-rate C&I loans to fixed-rate mortgages or longer-dated securities. But now you're locking in current yields, which might be below-market in a rising-rate scenario. And your EVE becomes very sensitive to rates.

Conversely, if you want to reduce EVE sensitivity (which a prudent board should care about), you might:
1. Issue long-term, fixed-rate debt: Lock in funding costs. But this is expensive and pressures NII.
2. Shorten the asset book: Shift to floating-rate, shorter-duration loans. But now NII becomes sensitive.

A Real Decision: Fixed vs Floating Mortgages

In 2020–2021, mortgage rates fell to 2.5–3.0%. Banks could originate 30-year fixed mortgages at 2.75% or push floating-rate mortgages at SOFR + 2.75% (SOFR was ~0% then, so all-in ~2.75%).

The NII argument: Originate floating. NII stays high as rates rise. The bank's equity funding costs are mostly deposit-based, which reprices slower than SOFR. You pocket the spread.

The EVE argument: Originate fixed. You lock in the current low rate environment. Future balance sheets will have higher rates and lower durations. Your current EVE is protected.

Most large banks did both — some fixed, some floating. Smaller, NII-focused banks leaned floating. Community banks, concerned about deposit competition and less capital, leaned fixed.

When rates rose 300bp+ in 2022–2023, the floating-rate originators won on NII (earnings were strong). But the fixed-rate banks won on EVE (they weren't as long-dated and vulnerable to the rate spike).

Of course, in 2024, as rates stabilized and mortgage originations fell, the floating-rate origination strategy became less valuable (lower volumes). The trade-off was real and path-dependent.

Part 4: Measurement and Governance

Standard Shock Scenarios

Most banks measure both NII and EVE sensitivity under these standard shocks:

1. Parallel shift, +100bp: Entire yield curve shifts up 100bp
2. Parallel shift, -100bp: Entire curve shifts down 100bp
3. Flattening: Long end up, short end down (or less up) — tests duration positioning
4. Steepening: Short end up, long end down (or less up) — tests short/long positioning
5. Non-parallel shocks: Key rate durations at specific maturity buckets

For NII, most boards set a tolerance: "NII should not decline more than 10% under a ±100bp parallel shock." Some tighter banks say 5%.

For EVE, the tolerance is often wider: "EVE should not decline more than 10–15% under a ±100bp parallel shock," reflecting the longer-term nature of the metric.

Governance Red Lines

When a bank's NII sensitivity exceeds tolerance:

• Immediate action: Reduce rate sensitivity. Shift deposit pricing, reduce floating-rate loan origination, lengthen liability duration.


• Medium-term action: Rebalance the balance sheet over 12–24 months. Originate more fixed-rate, shorter-duration products. Reduce longer-dated securities.

When EVE sensitivity exceeds tolerance:

• Immediate action: Reduce duration mismatch. Reduce long-dated securities. Issue longer-term debt to match long-term assets.


• Medium-term action: Shift loan mix to higher-floating exposure (which reduces asset duration). Reduce mortgage share of balance sheet.

Both actions are expensive and involve trade-offs. A board that micromanages both will paralyze the bank. A board that ignores both will eventually have a crisis.

Part 5: The Choice Framework

How do you decide which framework to prioritize? Here's a practical framework:

If You're a Community Bank:

• NII dominates. You have less sophisticated investors, less capital, and limited access to capital markets. Earnings stability is your lifeblood. You should manage NII tightly (±5% sensitivity) and tolerate slightly more EVE risk.
• Strategy: Slightly longer liability duration (3–5 year CDs), slightly shorter assets (more floating, more C&I). You trade some EVE risk for NII certainty.

If You're a Large, Well-Capitalized Bank:

• Both matter, but EVE protection is increasingly important. Sophisticated investors and regulators care about economic value. You have capital to absorb shocks but not indefinitely.
• Strategy: Manage NII within ±5–10%, but focus heavily on EVE (<±10% under 100bp shock). Issue longer-term debt (3–5 year average maturity). Shift to more floating-rate assets.
• Why: You can tolerate higher NII sensitivity if EVE is protected. You have the capital and market access to issue long-term funding.

If You're in a Specific Rate Environment:

• Rising rate world: NII sensitivity is less of a concern (you earn the positive carry). EVE becomes the bigger risk. You want to protect EVE by reducing asset duration and extending liability duration.
• Falling rate world: EVE is less of a concern (rates can't fall much further; you're bounded by the zero lower bound). NII sensitivity becomes critical. You want to protect NII by shifting to more floating assets and extending liabilities.
• Flat world: Both matter equally. Balance sheet positioning is less profitable, so capital allocation becomes more important.

Part 6: Connecting to the Broader Modules

In the modules ahead, we'll explore the mechanics of measuring both NII and EVE:

• Gap analysis (Module 32) is the simple, original way to measure NII sensitivity
• NII simulation (Module 33) is the sophisticated version, projecting actual earnings under multiple scenarios
• Scenario design (Module 34) teaches you how to build realistic rate paths
• EVE (Module 36) dives deep into the discount rate, reinvestment assumptions, and why EVE can diverge from NII
• Key rate duration (Module 37) captures non-parallel curve risk for both NII and EVE
• Duration and convexity (Module 39) explains why symmetric shocks don't produce symmetric results
• Model infrastructure (Module 41) shows you how real banks build these systems
• Model validation (Module 42) teaches you how to sanity-check your outputs

By the end of this track, you'll be able to build a full IRR framework, understand the trade-offs, and make defensible choices about how to position your balance sheet.