Banks measure interest rate risk in the banking book in two ways, and the two often disagree. Net interest income (NII) sensitivity, sometimes called earnings at risk, projects how the bank’s net interest income over the next 12 or 24 months would change if rates moved. Economic value of equity (EVE) sensitivity measures how the present value of all the bank’s assets minus all its liabilities would change if rates moved today, taking every cash flow to its final maturity into account. NII answers “what happens to earnings over the next year or two?” EVE answers “what happens to the value of the balance sheet as a whole?” A bank that funds long fixed-rate loans and securities with deposits whose rates lag the market can see NII rise when rates go up while its EVE falls sharply, which is exactly the position Silicon Valley Bank was in during 2022. Supervisors expect banks to measure, limit and report both, and internal audit’s job is to test whether the measures are built on assumptions the bank can defend.
This guide sets the two measures side by side, explains how each is calculated, why they diverge, which rate scenarios regulators expect, how deposit assumptions decide both results, how limits are set, what the Basel, EU and US frameworks require, what Silicon Valley Bank’s failure showed, and how internal audit tests both, with a worked example from a regional bank. It builds on the site’s interest rate risk overview and its guide to IRR, IRRBB and structural interest rate risk; the deposit assumptions are covered in depth in the guide to deposit beta modeling.
In this guide
- EVE and NII side by side
- How NII sensitivity is measured
- How EVE sensitivity is measured
- Why the two measures disagree
- The rate scenarios regulators expect
- Deposit assumptions: where both measures are won and lost
- Limits and risk appetite
- What the Basel, EU and US frameworks require
- Silicon Valley Bank: what happens when one measure is explained away
- How internal audit tests both measures
- Reading the ALCO report: ten questions
- Worked example: Lakeshore Bancorp’s EVE limit and the deposit assumption
- Common mistakes
- Questions about EVE and NII
- Related guides
EVE and NII side by side
| Aspect | NII sensitivity (earnings at risk) | EVE sensitivity (economic value) |
|---|---|---|
| The question | How would net interest income over the next 12 or 24 months change if rates moved? | How would the present value of assets minus liabilities change if rates moved now? |
| Horizon | Short: usually 12 months, often extended to 24 | Long: every cash flow to maturity |
| Balance sheet | Constant or dynamic, with assumptions about new business and reinvestment | Run-off: existing positions only, no new business |
| What drives it | Repricing: how quickly asset yields and funding costs adjust | Value: how the present value of fixed cash flows changes with discount rates |
| What it captures well | Near-term earnings pressure; the effect of deposit pricing lags | Long-dated mismatches; embedded options; the full cost of a position |
| What it misses | Risk beyond the horizon: a mismatch that bites in year three does not show | Timing: a large EVE loss may emerge in earnings slowly or not at all if the bank holds to maturity |
| Typical scenarios | Instantaneous shocks and gradual ramps of plus and minus 100 to 400 basis points; yield curve twists | Instantaneous shocks, including the six Basel scenarios |
| Typical limit | Maximum decline in 12-month NII, as a percentage of base NII | Maximum decline in EVE, as a percentage of base EVE or of Tier 1 capital |
| Key assumptions | Deposit betas and lags; prepayment speeds; new volumes and spreads | Deposit average lives and decay; prepayment speeds; discount rates |
| Regulatory anchor | EU outlier test on NII; US supervisory expectation | Basel outlier test at 15 percent of Tier 1 capital; US supervisory expectation |
Neither measure is right and the other wrong. They look at the same balance sheet over different horizons, and a bank that manages only one is managing half its risk.
How NII sensitivity is measured
NII sensitivity is a simulation. The bank’s asset-liability management model projects interest income and interest expense month by month over the horizon, first under a base case, usually the market’s implied forward curve or rates held flat, and then under each rate scenario. The difference between each scenario’s NII and the base NII, expressed as a percentage of base NII, is the sensitivity.
The result depends on the choices built into the simulation:
- Balance sheet treatment. A constant balance sheet replaces maturing positions with like positions; a dynamic balance sheet follows the bank’s plan for growth and mix. Constant balance sheets are easier to compare and are what the Basel disclosure uses; dynamic ones are closer to what management expects.
- Shock or ramp. An instantaneous shock moves all rates at once on day one; a ramp moves them gradually, say over 12 months. Ramps produce smaller first-year effects and are more realistic; shocks show the exposure more starkly.
- Repricing assumptions. Floating-rate loans reprice by contract; administered-rate deposits reprice at management’s discretion, captured through betas and lags. The deposit assumptions often matter more than anything else in the model.
- Optionality. Mortgage prepayments speed up when rates fall and slow when they rise; loan floors and caps switch on and off; depositors move money into time deposits when rates rise.
- Horizon. A 12-month result can look benign when the pain arrives in months 13 to 24, as fixed-rate assets continue at old yields while deposit costs keep catching up. Many banks show both years for that reason.
A typical asset-sensitive regional bank, with floating-rate commercial loans and a large base of non-interest-bearing deposits, shows NII rising in up-rate scenarios and falling in down-rate scenarios. A liability-sensitive bank, funded with deposits that reprice faster than its fixed-rate assets, shows the reverse.
How EVE sensitivity is measured
EVE is a valuation. The model projects every contractual and behavioral cash flow from every asset and liability, discounts each at the appropriate rate, and subtracts the present value of liabilities from the present value of assets. The bank’s own equity is excluded from the cash flows. The calculation is then repeated with each rate scenario applied to the discount curve and to the cash flows that depend on rates, and the change from base EVE is the sensitivity.
Three features of the calculation drive its results:
- Duration mismatch. A 30-year fixed-rate mortgage or a ten-year bond loses much more value when rates rise than a one-year loan. If assets are longer than liabilities, EVE falls when rates rise.
- Non-maturity deposits. Checking, savings and money market accounts have no contractual maturity. For EVE they need an assumed runoff pattern, usually expressed as an average life or decay rate, and the longer the assumed life, the more these low-cost deposits offset the value loss on long assets. This single assumption can move EVE sensitivity more than anything else in the model.
- Optionality and convexity. Prepayment options shorten mortgages when rates fall, which caps their gain; caps, floors and callable funding all change how values respond, so EVE losses are often larger for a rate rise than gains are for an equal fall.
EVE is usually expressed as a percentage change from base EVE, and sometimes, following the Basel outlier test, as a percentage of Tier 1 capital. The two denominators give different numbers for the same exposure, so an ALCO report that does not say which one it uses invites misreading.
Why the two measures disagree
The measures disagree because they look at different parts of the same future. NII looks at the next year or two of earnings, when many assets and liabilities have not yet repriced. EVE looks at the value of every future cash flow, including the years after the NII horizon ends. A position that helps earnings now and hurts them later shows up as an NII gain and an EVE loss.
| Balance sheet position | Effect on 12-month NII if rates rise 200 basis points | Effect on EVE if rates rise 200 basis points |
|---|---|---|
| Long fixed-rate securities funded by low-beta deposits | Up at first: deposit costs lag and floating assets reprice | Down: the securities lose value and the deposits’ value offset depends on their assumed life |
| Floating-rate commercial loans funded by non-interest-bearing deposits | Up: loan yields rise, funding cost does not | Little change: floating assets hold their value |
| Fixed-rate mortgages funded by time deposits | Down: deposits reprice at maturity, mortgages do not | Down: the mortgages lose value; prepayments slow, extending them |
| A pay-fixed interest rate swap hedging fixed-rate assets | Slightly down at first, then up as floating receipts rise | Up: the swap gains value, offsetting losses on the hedged assets |
| Short borrowing funding long bonds | Down: borrowing costs rise immediately | Down: the bonds lose value |
The first row describes most US banks in 2022. Deposit rates lagged the fastest tightening in four decades, so NII rose for most of the year, while the long securities bought with pandemic-era deposit inflows lost value. The EVE result was the warning; the NII result was the reassurance. Banks that watched NII and explained away EVE found out in 2023 which one had been right.
The disagreement also runs the other way. A bank that hedges its EVE exposure with pay-fixed swaps may reduce NII in a falling-rate scenario, because the swap’s fixed payments continue as floating receipts fall. Boards sometimes see that NII effect and conclude the hedge is costing money, when it is doing what it was bought to do.
The rate scenarios regulators expect
The Basel Committee’s 2016 standard on interest rate risk in the banking book, implemented from 2018, prescribes six shock scenarios for EVE, applied to each significant currency:
| Scenario | What moves |
|---|---|
| Parallel up | All maturities rise by the same amount |
| Parallel down | All maturities fall by the same amount |
| Steepener | Short rates fall and long rates rise |
| Flattener | Short rates rise and long rates fall |
| Short rates up | Short rates rise, with the shock fading at longer maturities |
| Short rates down | Short rates fall, with the shock fading at longer maturities |
The shock sizes are set by currency. For US dollar exposures the 2016 calibration used 200 basis points for parallel shocks, 300 for short-rate shocks and 150 for long-rate shocks. The Committee recalibrated the sizes in July 2024, with implementation from 1 January 2026, so check the current table for the currency being reported rather than relying on the original values.
US supervisors do not prescribe a scenario set, but their expectations run in the same direction. The 2010 interagency advisory on interest rate risk management expected banks to look beyond the conventional plus and minus 200 basis point parallel shocks to larger moves of 300 and 400 basis points, to non-parallel changes in the yield curve, and, in a low-rate environment, to put more weight on rising-rate scenarios. Many US banks now run a matrix of instantaneous shocks from minus 200 or 300 to plus 400 basis points, gradual ramps of plus and minus 100 and 200, and at least a steepening and a flattening twist, for both NII and EVE.
A scenario set is only useful if it includes the scenarios that would hurt this bank. A bank with a long securities book needs large up-rate shocks; a bank with heavy mortgage exposure needs down-rate scenarios that accelerate prepayments; a bank with rate floors on its loans needs scenarios where short rates fall toward zero. Internal audit should ask why each scenario is in the set, and which plausible scenario is missing.
Deposit assumptions: where both measures are won and lost
For most banks, the assumptions about non-maturity deposits matter more to both measures than anything else in the model, and they are the assumptions management has most room to choose.
| Assumption | What it means | Which measure it drives | Effect of a more favorable assumption |
|---|---|---|---|
| Deposit beta | The share of a market rate change passed on to the deposit rate | NII, mainly | Lower betas make NII look better when rates rise |
| Lag | How long the deposit rate takes to follow the market | NII | Longer lags make near-term NII look better |
| Decay or average life | How long balances are assumed to stay | EVE, mainly | Longer lives make deposits a larger offset to long assets, reducing EVE sensitivity |
| Migration | How much money moves from non-interest-bearing to interest-bearing accounts, or to time deposits, as rates rise | Both | Less migration makes both measures look better |
Because every one of these assumptions can be tuned, supervisors look hard at how they are set and changed. The 2010 advisory expected banks to document and support their key assumptions, to explain changes between periods and their effect, and to run sensitivity analysis showing which assumptions matter most. It also noted that rate-sensitive, higher-cost deposits, such as brokered and internet deposits, should carry higher decay rates than core relationships.
The Basel standard gives an external reference point. In its standardized framework, a bank may treat only part of its non-maturity deposits as stable, and caps the average repricing maturity it may assume for them: up to 90 percent of retail transactional deposits with a cap of five years, up to 70 percent of retail non-transactional deposits with a cap of four and a half years, and up to 50 percent of wholesale deposits with a cap of four years. US banks are not bound by those caps, but an assumption well beyond them is a question worth asking. How betas and decay rates are estimated, and how to audit the deposit study behind them, is the subject of the site’s guide to deposit beta modeling.
Limits and risk appetite
Most banks set limits for both measures, approved by the board as part of the risk appetite statement and monitored by the asset-liability committee (ALCO).
| Measure | Typical form of limit | Example tiers |
|---|---|---|
| NII sensitivity | Maximum decline in 12-month NII, as a percentage of base NII, by scenario | Minus 5 percent at plus or minus 100 basis points; minus 10 percent at 200; minus 15 percent at 300 |
| NII, second year | The same, for months 13 to 24 | Often looser, or an early-warning trigger rather than a limit |
| EVE sensitivity | Maximum decline in EVE, as a percentage of base EVE, by scenario | Minus 10 percent at 100; minus 15 percent at 200; minus 25 percent at 300 |
| EVE against capital | Maximum decline as a percentage of Tier 1 capital | A trigger at or below the Basel 15 percent outlier level |
The example tiers illustrate the structure only; the right levels depend on the bank’s capital, earnings and strategy. Three features distinguish a limit framework that works. It has limits on both measures, not one. It says what happens at a breach: who is told, by when, and what the options are, including hedging or changing the balance sheet. And it controls the assumptions behind the numbers, so that a limit cannot be met by changing an assumption without the same scrutiny as changing the position.
What the Basel, EU and US frameworks require
| Framework | What it requires | Status for a US bank |
|---|---|---|
| Basel Committee standard on IRRBB (2016, implemented 2018; shock sizes recalibrated in July 2024 for 2026) | Measurement of both EVE and NII; six prescribed shock scenarios; an EVE outlier test comparing the worst-case decline with 15 percent of Tier 1 capital; public disclosure of EVE and NII sensitivities | Not adopted as a rule; a useful benchmark |
| EU Capital Requirements Directive and EBA standards | An EVE outlier test at 15 percent of Tier 1 capital that triggers supervisory powers; since 2024, an NII test in which a one-year decline above 5 percent of Tier 1 capital counts as a “large decline” and prompts supervisory review, without automatic measures | Not applicable; relevant to EU subsidiaries and parents |
| US interagency advisory (2010) and frequently asked questions (2012) | Measurement from both earnings and economic value perspectives; scenarios beyond plus and minus 200 basis points; documented, supported and sensitivity-tested assumptions; limits; independent review | Supervisory expectation for all US banks |
| US examination practice | Interest rate risk assessed as part of the “sensitivity to market risk” component of the CAMELS rating; the OCC’s Comptroller’s Handbook booklet on interest rate risk sets out examination procedures | Applies |
| Model risk guidance (SR 26-2 and OCC Bulletin 2026-13, April 2026) | The ALM model is a model: governance, validation and ongoing monitoring in proportion to its materiality | Most relevant above 30 billion dollars of assets; below that, the bank’s own model risk policy sets the criteria |
The absence of a US outlier test is sometimes read as the absence of a US expectation. It is not: US examiners assess both earnings and economic value exposure, and the 2010 advisory remains the reference point for what they expect to see. The site’s guide to auditing model risk management covers how the ALM model fits into the bank’s model inventory.
Silicon Valley Bank: what happens when one measure is explained away
The Federal Reserve’s April 2023 review of the supervision of Silicon Valley Bank describes a bank whose interest rate risk was visible in its own measures for years. The main points, paraphrased from the review:
- The bank’s long-dated securities portfolio, funded by deposits that could reprice or leave quickly, created a structural mismatch, and the bank breached its internal limits on longer-term interest rate risk on and off from 2017.
- In April 2022, rather than reduce the exposure, the bank changed its modeling assumptions about the duration of its deposits, which reduced the measured risk.
- It removed interest rate hedges that would have protected it as rates rose, favoring near-term earnings.
- Management expected net interest income to benefit from rising rates early in 2022; by October 2022, projections showed NII declining.
- Unrealized losses on securities reached about 15 billion dollars at the end of 2022, while the great majority of its deposits were uninsured.
- Supervisors issued a matter requiring attention on interest rate risk in November 2022 and planned to downgrade the sensitivity to market risk rating; the bank failed on 10 March 2023, a day after customers tried to withdraw more than 40 billion dollars.
Every element in that list is something an internal audit of interest rate risk tests: limit breaches and their resolution, changes to deposit assumptions and who approved them, the removal of hedges, and whether the board saw the EVE result with the same prominence as the NII result. The site’s overview of bank runs puts the 2023 failures in historical context.
How internal audit tests both measures
| Area | What to test | Evidence | Red flags |
|---|---|---|---|
| Governance | Board-approved policy and limits for both NII and EVE; ALCO charter, membership and minutes | Policy, risk appetite statement, ALCO minutes for 12 months | Limits on one measure only; ALCO minutes that record numbers but no discussion |
| Limit monitoring and breaches | Every breach or near-breach, how it was escalated and resolved | Limit reports; breach log; board reporting | Breaches resolved by assumption changes; recurring breaches with no action |
| Data | Completeness and accuracy of the balances and terms loaded into the model | Reconciliation of model inputs to the general ledger and loan and deposit systems | Unreconciled differences; manual adjustments without support |
| Assumptions | Support for deposit betas, lags, decay rates, prepayment speeds and new business assumptions | Deposit studies, prepayment analysis, ALCO approval of assumptions | Assumptions unchanged for years; studies that exclude recent stress periods |
| Assumption changes | Every change in the period: rationale, approval, effect on results, and whether it moved a limit position | Change log; before-and-after results | Changes that coincide with limit pressure; effects not reported |
| Scenarios | Whether the scenario set includes those that would hurt this bank | Scenario rationale; ALCO discussion | Only parallel shocks; no scenarios large enough to test the balance sheet |
| Model validation | Independent validation of the ALM model and its key assumptions, with findings tracked | Validation report; issue log | Validation scope that excluded the deposit assumptions |
| Back-testing | Comparison of projected NII with actual NII, and of assumed deposit behavior with actual | Back-testing reports | No back-testing, or large misses not explained |
| Sensitivity analysis | Results under alternative assumptions for the most influential inputs | Sensitivity reports to ALCO | Single-point results with no range |
| Hedging | Hedge strategy, approvals, effectiveness and any unwinds | Hedge documentation; ALCO and board approvals | Hedges removed to support near-term earnings without board discussion |
| Reporting | Whether the board sees both measures, the limit position, assumption changes and their effect | Board risk committee packs | NII shown prominently and EVE in an appendix |
The assumption-change test is the one most likely to produce a significant finding and the one most often skipped. It asks a simple question: for each assumption change in the period, what would the limit position have been without it? The site’s guide to vendor model validation covers how to test models the bank bought rather than built, which includes most ALM engines.
Reading the ALCO report: ten questions
1. Which measures are shown, and are NII and EVE given equal prominence?
2. What are the limits for each, and where is each result against its limit?
3. Is EVE expressed as a percentage of base EVE or of Tier 1 capital, and is that stated?
4. Which scenarios are shown, and which would hurt this bank most?
5. What deposit betas, lags and average lives are assumed, and when were they last changed?
6. What would the results be under the previous assumptions, or under less favorable ones?
7. How did projected NII compare with actual NII over the last four quarters?
8. Has any limit been breached or approached in the last year, and how was it resolved?
9. What hedges are in place, and has any been added, removed or restructured?
10. Has the model been validated, when, with what findings, and are any open?
A report that can answer all ten is a good report. A board member or auditor who asks the sixth question will learn more about the bank’s interest rate risk than from any single number in the pack.
Worked example: Lakeshore Bancorp’s EVE limit and the deposit assumption
Lakeshore Bancorp, the nine-billion-dollar regional bank with 60 branches used across this site, holds a book of fixed-rate residential mortgages and a securities portfolio funded mostly by retail deposits, a quarter of them non-interest-bearing, alongside floating-rate commercial loans. Internal audit’s 2026 audit of interest rate risk management took the June 2026 ALCO package as its starting point.
The results. The bank’s profile was the classic divergence: NII rose in up-rate scenarios and EVE fell.
| Scenario | 12-month NII change | Limit | EVE change | Limit |
|---|---|---|---|---|
| Plus 300 basis points | +4.2% | -15% | -17.9% | -25% |
| Plus 200 basis points | +3.1% | -10% | -11.6% | -15% |
| Plus 100 basis points | +1.7% | -5% | -5.4% | -10% |
| Minus 100 basis points | -2.6% | -5% | +3.9% | -10% |
| Minus 200 basis points | -5.8% | -10% | +6.2% | -15% |
Every result sat inside its limit. The assumption-change test told a different story.
The assumption change. In the fourth quarter of 2025, the ALM team had extended the assumed average life of interest checking and savings balances, about 2.2 billion dollars, from 4.1 years to 6.8 years, based on a new deposit study. The change had been presented to ALCO and approved. Internal audit re-ran the June results under the old assumption with the model team: EVE sensitivity at plus 200 basis points would have been minus 16.2 percent, outside the minus 15 percent limit, and at plus 300 minus 24.3 percent, just inside its limit. The NII results barely moved, because the assumption drives EVE rather than NII.
What the audit found. The deposit study itself was careful work, but it measured balance retention from 2015 to 2021, a period of near-zero rates and steady deposit growth, and excluded the 2022 to 2024 cycle in which the bank’s own interest checking balances fell by 9 percent as customers moved money into money market accounts and time deposits. The ALCO minutes recorded the approval but no discussion of the effect on the limit position, which had not been presented. The model validation completed in 2025 predated the change and had not covered the study. Expressed against Tier 1 capital, the Basel outlier measure the bank reported to its board as a benchmark, the plus 200 decline was about 15 percent under the new assumption and about 21 percent under the old one. An assumed life of 6.8 years for retail transactional deposits was also well beyond the five-year cap in the Basel standardized framework, which the bank was not bound by but had cited in its own policy as a reference point.
The finding and the response. Internal audit rated the finding High: the change might be justifiable, but the governance around it was not, and it had moved a limit position without anyone deciding that it should. Management agreed to re-run the study including the 2022 to 2024 data, to present EVE results under both assumptions until the study was validated, and to amend the policy so that any assumption change affecting a limit position is reviewed by model risk management and reported to the board risk committee with its before-and-after effect. The re-run study supported an average life of 5.3 years, which put plus 200 EVE sensitivity at minus 14.1 percent. The board risk committee asked for the second-year NII result to be added to the pack, and for a discussion of whether a modest pay-fixed swap program would reduce EVE sensitivity at an acceptable cost to NII.
Common mistakes
| Mistake | Why it matters | Better practice |
|---|---|---|
| Managing NII and treating EVE as a regulatory formality | The EVE loss arrives in earnings eventually, or in a sale at a loss | Limits, reporting and discussion for both measures |
| Unexplained denominators | An EVE decline of 12 percent of EVE and one of 15 percent of Tier 1 can be the same exposure | State the denominator on every report |
| Assumptions that change without their effect being shown | Limits can be met by changing inputs rather than positions | Report before-and-after results for every assumption change |
| Deposit studies built on benign periods | Balances look stickier than they are under stress | Include the most recent rising-rate cycle and stress periods |
| Only parallel shocks | Twists and short-rate shocks can hurt more than parallel moves | Include the scenarios that match this balance sheet’s vulnerabilities |
| A 12-month NII horizon only | Pressure that arrives in year two stays invisible | Show months 13 to 24 |
| No back-testing | Model errors persist unnoticed | Compare projected and actual NII and deposit behavior each quarter |
| Hedges judged only on their NII effect | Protective hedges are unwound because they cost earnings in some scenarios | Evaluate hedges on both measures |
Questions about EVE and NII
What is the difference between EVE and NII?
NII sensitivity measures how net interest income over the next 12 to 24 months would change if rates moved. EVE sensitivity measures how the present value of all assets minus all liabilities would change. NII is a short-term earnings measure; EVE is a long-term value measure.
Why can NII rise when EVE falls?
Because deposit rates often lag market rates and floating-rate assets reprice quickly, so earnings rise at first, while long fixed-rate assets lose value immediately. The EVE loss shows up in earnings later, as the fixed-rate assets continue at old yields while funding costs catch up.
What is the Basel outlier test for EVE?
A supervisory test comparing a bank’s largest EVE decline across the six prescribed scenarios with 15 percent of its Tier 1 capital. Banks above the threshold face closer supervisory scrutiny. US banks are not subject to it, though many use it as a benchmark.
Do US regulators require EVE?
There is no prescribed US EVE calculation or outlier test, but US supervisors expect banks to measure interest rate risk from both an earnings and an economic value perspective, as the 2010 interagency advisory sets out, and assess it as part of the CAMELS sensitivity to market risk component.
Which matters more, EVE or NII?
Neither. NII shows how exposed earnings are over the next year or two; EVE shows the full exposure over the life of the balance sheet. A bank needs limits and reporting for both, and should be most concerned when they point in opposite directions.
What is the most important assumption in an EVE model?
For most banks, the assumed average life or decay rate of non-maturity deposits. Small changes can move EVE sensitivity by several percentage points, which is why assumption changes need the same governance as changes to positions.
Related guides
- Deposit beta modeling — how betas, lags and decay rates are estimated and audited.
- Interest rate risk: an overview — the concepts behind both measures.
- IRR vs IRRBB vs structural interest rate risk — the vocabulary.
- Interest rate risk vs asset-liability management — where the measures sit in ALM.
- The complete guide to interest rate risk — tools and internal audit perspectives.
- Liquidity risk — the other half of the SVB story.
- Auditing model risk management — the ALM model inside the model inventory.
- Vendor model validation — testing an ALM engine the bank bought.
- Risk appetite statements — where the limits belong.
- How to audit treasury — treasury audits outside banking.
- Bank runs in history — the 2023 failures in context.
- The OCC’s risk categories — interest rate risk among the eight categories.
- The interest rate risk library — every guide on this risk.
- How to Audit Model Risk Management: End-to-End Program — a work program for MRM under SR 26-2, from inventory to validation to vendor models
- Internal Audit in Banking: The Complete Guide — what makes bank audit different, the regulators, the universe and the career path
Leave a Reply