OCC Interest Rate Risk handbook: a practitioner read

The OCC Interest Rate Risk Handbook: Practitioner's Playbook

If you work at a national bank, the OCC Interest Rate Risk Handbook is your Bible. It's 40 pages of guidance on how to measure, manage, and govern interest rate risk. For a junior ALM analyst, reading and understanding this Handbook is essential. It's the OCC's definition of what "sound" interest rate risk management looks like.

The Handbook is not a regulation (it's not binding law). But OCC examiners use it as the standard. If your bank deviates from it, examiners will ask why. You'd better have a good answer.

The Five Core Sections

1. Management of Interest Rate Risk: Core Concepts

The Handbook starts with philosophy:

"Management of interest rate risk is essential to the safe and sound operation of commercial banks. Interest rate risk arises from the timing of repricing of assets, liabilities, and off-balance-sheet items. Changes in interest rates can adversely affect the bank's earnings (earnings risk) and its economic value (market risk)."

Key concepts:

• Repricing gap: The difference between assets and liabilities repricing in a given time bucket. A positive gap (assets > liabilities repricing) means you benefit from rates going up.


• Duration: How sensitive is the market value of an asset to interest rate changes? A 30-year mortgage with 5% coupon has high duration (market value is very sensitive to rate moves). A floating-rate loan with no coupon lock has low duration.


• Earnings risk: How much will NII change if rates move? If you're asset-sensitive, lower rates hurt NII. If you're liability-sensitive, higher rates hurt NII.


• Economic (market) value risk: How much will the bank's equity value change if rates move? This is EVE analysis.

The Handbook emphasizes: "Banks should not take interest rate risk passively. Interest rate risk should be actively managed to support the bank's strategic objectives."

Translation: You should have a view on rates. You should position accordingly (within policy limits). But you should do it deliberately and understand the risks.

2. Interest Rate Risk Measurement

The Handbook identifies four methodologies:

a) Gap Analysis

Create time buckets (0-3 months, 3-6 months, 6-12 months, 1-3 years, 3-5 years, 5+ years).
For each bucket, calculate:

• Rate-sensitive assets: loans and securities repricing in that bucket


• Rate-sensitive liabilities: deposits and wholesale funding repricing in that bucket


• Gap: RSA - RSL

Example:
``
0-3 months: RSA = $10B, RSL = $8B, Gap = +$2B (asset-sensitive)
3-6 months: RSA = $9B, RSL = $7B, Gap = +$2B
6-12 months: RSA = $8B, RSL = $10B, Gap = -$2B (liability-sensitive)
`

If rates rise, you earn more on the +$2B gap in months 0-3 (your assets reprice up faster than your liabilities). In months 6-12, you lose (your liabilities reprice up faster than your assets).

Gap analysis is intuitive but crude. It assumes all repricing is immediate and doesn't account for duration.

b) Duration Analysis

Duration is more sophisticated. It measures the weighted average time to repricing of cash flows.

Formula:
`
Duration = sum(t * CF_t) / sum(CF_t)
where t = time period, CF_t = cash flow in period t
`

Example: A 30-year mortgage with 5% coupon and current yield of 5%:

• Pays $50k/year for 30 years


• Duration ≈ 8-9 years (the cash flows are weighted toward early years due to amortization)

If rates rise 1%, the market value of the mortgage falls ~8-9%. This is duration risk.

Duration gap:
`
Duration Gap = Duration(Assets) - Duration(Liabilities)
``

If duration of assets is 5 years and duration of liabilities is 2 years:

• A 1% rate rise causes assets to fall 5% in value


• Liabilities fall 2% in value


• Net effect: equity falls 3% in value

This is more accurate than gap analysis because it accounts for the timing of cash flows.

c) Scenario Analysis

The Handbook recommends stress-testing under various scenarios:

• Parallel shift: All rates move up or down equally (Fed raises rates uniformly)


• Non-parallel shifts: Short rates move different from long rates (yield curve flattens or steepens)


• Bank-specific scenarios: Based on the bank's actual risk profile

Example:

• Scenario 1: +200 bps parallel shift (all rates up 2%)


• Scenario 2: -200 bps parallel shift (all rates down 2%)


• Scenario 3: Yield curve flattens (short rates +100, long rates -50)


• Scenario 4: Yield curve steepens (short rates -50, long rates +100)


• Scenario 5: Bank-specific (our bank's deposit betas rise from 70% to 85%; our mortgages prepay faster)

For each scenario, calculate EVE impact and NII impact.

d) Value-at-Risk (VAR)

Optional methodology. VAR models the probability distribution of outcomes. "We are 95% confident that a 1-day loss from interest rate moves will not exceed $X."

VAR is more sophisticated but not commonly used for ALM (it's more trading-focused). The Handbook mentions it but doesn't require it.

3. Management and Governance

The Handbook is very clear on governance:

"The board of directors should establish a clear interest rate risk policy that outlines the bank's tolerance for interest rate risk. The policy should be specific, measurable, and enforceable."

Specifically:

• Board approval: The IRR policy must be approved by the board (not just ALCO)
• Specific limits: "EVE should not decline by more than X% of capital under a Y bps shock." Not vague.
• Independent oversight: Someone other than the risk-takers (Treasury) must oversee the policy (Risk Committee)
• Regular reporting: ALCO should review metrics at least quarterly (monthly is better)
• Annual review: The policy should be reviewed annually
• Escalation: Breaches must be escalated and remediated

"Management should ensure that the measurement and management of interest rate risk are integrated throughout the bank. This includes loan pricing, funding decisions, and investment portfolio management."

Translation: Every business line should understand how their decisions affect balance sheet interest rate risk.

4. Assumptions and Estimation

This section is critical. The Handbook outlines assumptions that should be validated:

Deposit repricing and behavior:

• "Demand deposits should be treated conservatively. Even if they are non-interest-bearing, they will reprice if competitors offer rates."


• "The bank should estimate deposit beta based on historical experience and peer comparison."


• "In stress scenarios, deposit behavior may change materially."

Loan repricing:

• "Fixed-rate loans reprice only at maturity or prepayment."


• "ARMs reprice according to index and margin. The bank should model repricing lag (e.g., 1-3 month lag between index move and rate adjustment)."


• "Prepayment assumptions should be based on historical experience and industry models."

Mortgage prepayment:

• "As rates fall, prepayments accelerate. The bank should model prepayment risk."


• "Factors affecting prepayment: interest rate change, seasoning of mortgage, seasonal patterns, economic conditions."


• "A reasonable prepayment model uses S-curve relationships (slow acceleration of prepayments until rates fall significantly, then rapid acceleration)."

Customer behavior:

• "Customers may not reprice in lockstep with market. Model repricing lags."


• "Relationship depositors (small business) are stickier than rate-chasers (large corporations)."

5. Risk Limits and Tolerances

The Handbook provides guidance on reasonable limits:

EVE sensitivity:

• "A limit of 10-15% of capital for a 200 bps shock is typical for regional banks."


• "Large, sophisticated banks with active hedging programs may have tighter limits (5-10%)."


• "Community banks may have looser limits (15-20%)."

NII sensitivity:

• "A limit of 5% for a 200 bps shock is reasonable."


• "Some banks use earnings-at-risk (EAR) limits instead: NII volatility should not exceed X% of expected net income."

Duration gap:

• "Some banks use duration gap limits: the absolute value of duration gap should not exceed 2-3 years."


• "This is a secondary metric, not as important as EVE."

How to Use the Handbook as an ALM Professional

1. Read it. Seriously. It's well-written and foundational.

2. Build your models using Handbook guidance. Your EVE model should follow Handbook methodology (discounted cash flow, proper scenario selection, reasonable assumptions).

3. Document your assumptions against the Handbook. "Our deposit beta is 70%, based on [historical data]. The Handbook suggests deposit betas of 50-80% depending on competitive environment; our 70% is reasonable."

4. When examiners ask questions, reference the Handbook. If an examiner says, "Your IRR policy doesn't have a specific EVE limit," you can say, "You're right, we'll add one. The Handbook suggests 10-15% of capital; we'll propose 12%."

5. Use it to push back on overly conservative constraints. If business lines say your funding allocation is too tight, you can model the EVE impact using Handbook methodology, show it's within policy, and argue for higher allocation.

Takeaway

The OCC Interest Rate Risk Handbook is:

• The gold standard for interest rate risk management in national banks


• Specific on measurement methodologies


• Thorough on assumptions and limitations


• Clear on governance and limits


• A document you should reference regularly

If your bank follows the Handbook, you have solid ALM. If you deviate, be prepared to explain why.

<h2>The OCC Handbook: From Theory to Practical Application in Real ALM Scenarios</h2>

<h3>Understanding the OCC Handbook as Your Operating Manual</h3>

<p>The Office of the Comptroller of the Currency's Handbook on interest rate risk is far more than a regulatory document—it's a practical operating manual that outlines the methodology for measuring, modeling, and managing interest rate risk in commercial banks. For national banks and federal savings associations under OCC supervision, this handbook defines the standard for how interest rate risk should be measured and communicated. The handbook is approximately 40 pages of detailed technical guidance that covers risk measurement methodologies, stress scenarios, modeling assumptions, governance requirements, and disclosure practices. Mastering this document is not optional for a national bank; it's the foundation of your ALM framework.</p>

<h3>Building an EVE Model to Handbook Standards</h3>

<p>The OCC Handbook provides step-by-step guidance on constructing an Economic Value of Equity (EVE) model. Let me walk through how to build a model that meets these standards, using a realistic example from a mid-size regional bank.</p>

<p><strong>Step 1: Creating a Detailed Repricing Schedule</strong></p>

<p>Under the OCC Handbook, you must create a comprehensive repricing schedule for each major asset and liability category. This schedule specifies when interest rates on each instrument will change—at maturity, at the next reset date, or in response to customer behavior. Here's a realistic example:</p>

<pre>
ASSETS
Category | Book Value | Interest Rate | Repricing Period | Rate Sensitivity
30-year Mortgages | $20,000M | 6.50% | At maturity (30 yr) | Fixed; prepayment risk
ARM Mortgages | $3,000M | 5.50%+2.5% | 6 months | Float; index-sensitive
5-year Commercial | $8,000M | 7.00% | 5 years | Fixed; some prepayment
Floating Commercial | $4,000M | Prime+1.5% | 1 month | Float; Fed funds tied
Treasury Securities | $2,000M | 4.50% | 1,3,10 years mixed | Fixed; maturity schedule
Agency MBS | $3,000M | 4.75% | ~5 yr (average) | Fixed; prepayment risk
Total Assets | $40,000M | | |

LIABILITIES & EQUITY
Category | Book Value | Interest Rate | Repricing Period | Rate Sensitivity
Demand Deposits | $15,000M | 0.50% (MMDA) | Instantaneous | Float; deposit beta
Time Deposits (3-yr) | $8,000M | 4.50% | 3 years | Fixed at maturity
Time Deposits (1-yr) | $5,000M | 4.75% | 1 year | Fixed at maturity
Wholesale CDs | $3,000M | 4.80% | 1 year | Fixed at maturity
Senior Debt (5-yr) | $6,000M | 4.50% | 5 years | Fixed at maturity
Equity | $3,000M | N/A | N/A | Absorbs all losses
Total Liabilities | $40,000M | | |
</pre>

<p>This repricing schedule is the foundation of your model. It answers the critical questions: When will rates change? By how much? How sensitive is each instrument to rate movements? The quality of this schedule determines the quality of your EVE calculation.</p>

<p><strong>Step 2: Defining Scenarios Aligned with Federal Reserve Standards</strong></p>

<p>The OCC Handbook directs banks to use the Federal Reserve's standard scenarios for interest rate shocks. These scenarios represent the regulatory baseline for assessing interest rate risk. The primary scenarios are:</p>

<p><strong>Scenario 1: +200 Basis Point Parallel Shift</strong></p>

<p>In this scenario, interest rates across the yield curve rise uniformly by 200 basis points. Current mortgage rates of 6.50% become 8.50%; deposit rates that are currently 4.50% become 6.50% (assuming a 100% deposit beta—that is, deposits reprice in line with market rates). The impact on your EVE depends on your duration gap. If you have more long-duration assets than liabilities (you're asset-sensitive), rising rates hurt you because asset values fall more than liability values. For a bank with $20B in mortgages and $15B in deposits, the asset sensitivity is significant.</p>

<p><strong>Scenario 2: -200 Basis Point Parallel Shift</strong></p>

<p>Rates fall uniformly across all maturities by 200 basis points. Mortgages fall from 6.50% to 4.50%; deposits fall from 4.50% to 2.50% (or to a floor, since deposit rates can't go negative). This scenario helps illustrate prepayment risk: when rates fall, mortgage holders prepay their mortgages at the old higher rate and refinance at the new lower rate. The bank loses the long-duration asset and must reinvest the prepayment proceeds at lower rates. For EVE, falling rates hurt a duration-long bank (assets are long, liabilities are short).</p>

<p><strong>Scenario 3: Yield Curve Twist</strong></p>

<p>In a twist scenario, different parts of the yield curve move differently. Short-term rates might rise 200 basis points while long-term rates rise only 50 basis points (a flattening). Alternatively, short rates might fall while long rates rise (a steepening). These non-parallel shifts capture the reality that yield curves don't always move in lockstep. A bank's EVE sensitivity to a twist depends on whether it has concentrated exposure to specific points on the curve.</p>

<p><strong>Step 3: Projecting Cash Flows Under Each Scenario</strong></p>

<p>For each asset and liability, under each scenario, you project all future cash flows. This is where prepayment modeling and deposit repricing behavior become critical. Let's use 30-year mortgages as an example.</p>

<p>In the base case (current rates), a $100 million mortgage at 6.50% generates annual payments of approximately $127.2 million (principal plus interest). You calculate cash flows for all 30 years, tracking principal paid down annually and interest earned.</p>

<p>But in the -200 basis point scenario, things change dramatically. The mortgage at 6.50% is now underwater relative to market rates of 4.50%. Rational borrowers refinance. The OCC Handbook requires you to model this prepayment behavior realistically. Historical data shows that when rates fall this significantly, prepayment rates accelerate. A reasonable model might assume 50% of mortgages prepay in Year 1 of a -200 scenario (called the "conditional prepayment rate" or CPR). Your cash flow projections then reflect this: instead of receiving a stream of payments over 30 years, you receive a large portion of principal back in Year 1, which you must reinvest at the new lower rates.</p>

<p><strong>Step 4: Discounting Cash Flows at New Rate Scenarios</strong></p>

<p>Once you've projected cash flows under each scenario, you discount them back to present value using the rates implied by that scenario. In the +200 scenario, you discount using higher rates (8.50% for mortgages, 6.50% for deposits). The higher discount rate reduces present values: long-duration assets like mortgages lose more value when discount rates rise.</p>

<p>Let's work through a simple example. If you have $100 in cash flow due in Year 1, its present value under the +200 scenario at 8.50% discount rate is: $100 / 1.085 = $92.17. Compare that to the base case at 6.50%: $100 / 1.065 = $93.90. The difference ($1.73) is the loss in value due to rising rates.</p>

<p>For all assets and liabilities across all years and all scenarios, you perform this discounting exercise. Sum all the present values: the total PV of assets minus the total PV of liabilities equals EVE.</p>

<p><strong>Step 5: Calculating EVE Impact and Sensitivity</strong></p>

<p>Now you have EVE under multiple scenarios. Your base case EVE (at current market rates) might be $2,700 million. Under the +200 scenario, EVE might fall to $1,000 million. The impact is -$1,700 million, or -56.7% of your $3,000 million in capital.</p>

<p>This is a significant loss! Most banks' policies limit EVE sensitivity to 12% of capital (some allow up to 15%). If your EVE loss under a +200 scenario is 56.7% of capital, you're wildly out of policy. What's driving this? In this example, it's the duration gap: long-duration mortgages funded with short-duration deposits. When rates rise, mortgages lose 20-30% of value, but deposit liabilities barely change in value because they're short-duration. The net loss is substantial.</p>

<p><strong>Managing EVE: The Feedback Loop</strong></p>

<p>When EVE sensitivity exceeds policy limits, ALM must identify remediation. The bank might:</p>

<p>Sell $10 billion of the mortgage portfolio to a mortgage bank (reduces exposure to long-duration assets); Securitize $5 billion of mortgages (removes them from balance sheet, reduces EVE sensitivity); Enter interest rate swaps to convert fixed-rate mortgages to floating rate (reduces duration of assets); or Extend the maturity profile of deposits or wholesale funding (increases duration of liabilities, reducing the duration gap).</p>

<p>Each option has trade-offs: securitization costs money; hedging has accounting implications; funding maturity extension means higher interest expense. But the principle is clear: if you're out of policy, you must take action to get back in.</p>

<h3>Deposit Behavior and the Post-SVB Handbook Lessons</h3>

<p>The OCC Handbook emphasizes that "in stress scenarios, deposit behavior may change materially." This is now the most important lesson from SVB. The handbook says deposits are not equally stable across scenarios. Insured deposits (those under $250,000 in FDIC coverage) tend to be stickier—they're less likely to flee in a crisis because they're protected. Uninsured deposits, especially those from sophisticated customers or institutional sources, can be flight-prone in stress scenarios.</p>

<p>Pre-SVB, many banks modeled uninsured deposit runoff at 10-25% in stress scenarios. SVB, with 92% uninsured deposits, experienced 80%+ runoff. Post-SVB, examiners now ask: "Is your uninsured deposit runoff assumption realistic? If your bank is 70% uninsured deposits, and those deposits are from technology companies or venture capital funds, wouldn't they run at higher rates in stress?"</p>

<p>The handbook approach is to segment deposits by stability and model them accordingly. Insured retail deposits might have 5% runoff in stress; uninsured business deposits might have 50% or more. The handbook also notes that deposit beta (how much deposit rates reprice in response to market rate changes) can spike in stress scenarios. In normal times, deposit beta might be 70%; in stress, it could spike to 90% or higher if competitors are aggressively recruiting deposits.</p>

<h3>Model Governance Requirements from the Handbook</h3>

<p>The handbook is explicit about governance: "The board should ensure that interest rate risk models are independently validated at least annually. Validation should include: (1) back-testing the model against actual results; (2) benchmarking assumptions against peer banks and industry data; (3) stress-testing the model itself to identify where it fails or becomes unreliable; and (4) assessing the model's limitations and communicating them to management."</p>

<p>This governance standard has a specific implication: you need a Model Risk Management function that is separate from the Treasury team that builds and uses the models. The Model Risk team validates; Treasury builds and runs. This separation ensures independence and prevents the risk team from becoming captured by the business line.</p>

<p>Validation should be ongoing. Annual validation is a minimum. If your business model changes materially—say, you shift from mortgages to commercial loans, or from deposits to wholesale funding—you should re-validate assumptions. If actual deposit repricing differs from your model's predictions by more than a small tolerance, you should investigate and adjust the model.</p>

<h3>The Handbook on Hedging: Documentation and Economic Intent</h3>

<p>The OCC Handbook approves of hedging as a tool to manage interest rate risk but requires clear documentation and economic intent. Say you have $20 billion in fixed-rate mortgages. You want to reduce your interest rate risk by hedging half of that amount through interest rate swaps.</p>

<p>Step 1: Document the intent. Write a memo that says: "We are entering interest rate swaps to reduce our EVE sensitivity from 12% to 8% of capital under the -200 basis point scenario. This will free up EVE capacity for business growth in our mortgage origination."</p>

<p>Step 2: Identify the hedge instrument. "We will enter pay-fixed, receive-floating swaps with a notional of $10 billion, 5-year tenor, at a fixed rate of 4.50%, receiving Fed funds flat. This converts the economic exposure of $10B of fixed-rate mortgages into floating-rate exposure, reducing interest rate risk."</p>

<p>Step 3: Model the impact. Rerun your EVE model with the hedges in place. Show that EVE sensitivity has improved from 12% to 8%.</p>

<p>Step 4: Account for it properly. Is this a "hedge" under accounting standards (FAS 133)? Or is it an "economic hedge" that doesn't qualify for hedge accounting? Communicate clearly to your accounting team and your auditors.</p>

<p>The handbook doesn't require hedge accounting—it allows economic hedging even if accounting doesn't recognize it as a hedge. But you must be explicit about what you're hedging and why.</p>

<h3>The Handbook on Limits: Setting Realistic Policies</h3>

<p>The handbook provides guidance on limit-setting: "Most banks establish an EVE limit in the range of 10-15% of capital for a 200 basis point interest rate shock. The appropriate limit depends on bank size, balance sheet complexity, risk appetite, and earnings volatility." A large complex bank might set a 10% EVE limit (tighter, more conservative). A mid-size bank might set 12%. A small community bank with stable deposits might set 15%.</p>

<p>The handbook also notes: "Limits should be calibrated to be breached occasionally (1-2 times per year) but not regularly. A limit that is never breached is too loose. A limit that is always breached is meaningless." This is an important insight. If your limit is so loose that you never breach it, the limit isn't constraining behavior or forcing decisions. If your limit is so tight that you breach it constantly, the limit becomes a nuisance rather than a governance tool.</p>

<h3>Applying the Handbook: A Real ALCO Scenario</h3>

<p>Let's walk through a realistic ALCO discussion that applies the handbook:</p>

<p><strong>Mortgage Business Head:</strong> "We want to book $2 billion in mortgages next quarter. We have the pipeline and the capability. But ALM is saying we can only do $800 million due to EVE limits. The handbook says limits should be 10-15%. We're at 11%. Why can't we do more?"</p>

<p><strong>ALM Lead:</strong> "Good question. Here's the math: Currently, our EVE sensitivity to a -200 scenario is 11% of capital, which is within our policy limit of 12%. If we add $2 billion in mortgages, our EVE sensitivity would climb to 14%, exceeding our limit. We can accommodate about $800 million in new mortgages and stay within policy. Beyond that, we'd be in breach."</p>

<p><strong>Mortgage Business Head:</strong> "Could ALCO change the policy to 14%?"</p>

<p><strong>ALM Lead:</strong> "That's possible. The handbook says 10-15% is reasonable, so 14% is within guidance. But it would require board approval. And you should think about the trade-off: a 14% EVE limit means that if rates fall 200 basis points, our equity value falls by 14% of capital. Is that acceptable to the board and to shareholders?"</p>

<p><strong>Mortgage Business Head:</strong> "I can take that to the board. What would the impact be on earnings?"</p>

<p><strong>ALM Lead:</strong> "With $2 billion in new mortgages, net interest income would probably increase by $25-30 million per year. Offset by increased hedging costs of maybe $2-3 million. Net benefit: $22-27 million per year. But the downside risk is that in a falling-rate scenario, that earnings boost gets partially offset by prepayment losses. Also, if we ever want to securitize mortgages or access the mortgage market for funding, a high EVE sensitivity might limit our options. It's a business decision, but it has trade-offs."</p>

<p><strong>Mortgage Business Head:</strong> "Let me put together a proposal for the board."</p>

<p>Notice what happened: The handbook provided a framework for the conversation. It wasn't "no, we can't do this." It was "yes, we can, but here's the policy, here's the trade-off, and here's who needs to approve it." The handbook created a common language and set of principles that allowed the business head and the ALM lead to have a productive conversation rather than a standoff.</p>

<h3>The Handbook as a Tool for Examiner Conversations</h3>

<p>When OCC examiners come, they will reference the handbook. If your bank is using handbook-compliant methodology, you're on strong ground. If examiners have concerns about your model, you can refer to the handbook guidance. For example, if an examiner questions your deposit beta assumption, you can say: "The handbook suggests deposit beta should be validated quarterly against actual experience. Here's our quarterly validation showing that actual deposit beta in 2024 has been 72%, and our model assumption of 70% is within reasonable variance. If we see continued drift, we'll adjust the assumption."</p>

<p>The handbook is your shield against arbitrary demands. It's not a complete defense—examiners can always ask for more—but it establishes standards that examiners themselves reference.</p>

<h3>Conclusion: Mastering the Handbook</h3>

<p>For national banks, the OCC Handbook is not optional. It's the operating manual for interest rate risk management in the United States. Master it. Use it. Reference it in your ALCO discussions. Ensure your models are handbook-compliant. Document your assumptions. Validate your models annually. Govern yourself to the handbook standard, and you'll be well-positioned for any OCC examination.</p>