IRRBB - Interest Rate Risk in Banking Book
Understanding Basel III Interest Rate Risk in the Banking Book framework and its application to JGB portfolios
Introduction: Why IRRBB Matters for JGB Portfolios
Interest Rate Risk in the Banking Book (IRRBB) is the Basel III framework for measuring and managing interest rate risk in banks' non-trading portfolios. For Japanese banks holding significant JGB positions, IRRBB represents one of the most critical regulatory risk metrics.
Unlike trading book capital (covered by market risk rules), banking book positions are typically:
- Held-to-maturity JGBs purchased for yield and regulatory purposes
- Available-for-sale (AFS) JGBs providing liquidity and asset-liability management
- Loan portfolios hedged with JGB positions or interest rate swaps
IRRBB measures how these positions' economic value and earnings change under standardized interest rate shock scenarios—critical for banks operating under the BOJ's unconventional monetary policy environment.
The Two Pillars of IRRBB Measurement
The Basel Committee on Banking Supervision (BCBS) requires banks to measure IRRBB using two complementary metrics:
1. Economic Value of Equity (ΔEVE)
Definition: The change in the net present value of the bank's assets minus liabilities under interest rate shocks.
\[\Delta EVE = PV_{Assets,shocked} - PV_{Liabilities,shocked} - (PV_{Assets,base} - PV_{Liabilities,base})\]What it captures: Immediate balance sheet impact if interest rates move today
Perspective: Economic (mark-to-market), not accounting
2. Net Interest Income (ΔNII)
Definition: The change in net interest income (interest earned on assets minus interest paid on liabilities) over a one-year forward period.
\[\Delta NII = (Interest_{Assets,shocked} - Interest_{Liabilities,shocked}) - (Interest_{Assets,base} - Interest_{Liabilities,base})\]What it captures: Impact on earnings over the next 12 months
Perspective: Accrual accounting (how P&L changes)
Key Distinction: EVE measures economic capital impact; NII measures earnings impact. A bank can have favorable NII (higher earnings) while suffering negative EVE (economic value loss).
The Six Standard Shock Scenarios
The BCBS prescribes six standard interest rate shock scenarios that banks must apply to their portfolios. These are designed to capture different types of curve movements:
| Scenario | Description | JGB Context |
|---|---|---|
| Parallel Up | All maturities rise by same amount (e.g., +200bp) | BOJ exits YCC, raises all rates simultaneously |
| Parallel Down | All maturities fall by same amount (e.g., -200bp) | Return to aggressive easing (negative rates deepen) |
| Steepener | Short rates unchanged, long rates rise significantly | BOJ holds short rates, but fiscal concerns hit long-end |
| Flattener | Short rates rise more than long rates | BOJ hikes aggressively, long-end anchored by growth concerns |
| Short Rates Up | Only short-end (≤1Y) rises sharply | BOJ raises policy rate from -0.1% to +0.5% |
| Short Rates Down | Only short-end (≤1Y) falls sharply | BOJ cuts rates further into negative territory |
Important: The magnitude of shocks is calibrated to each currency. For JPY, given the historically low rate environment, the prescribed shocks are smaller than for USD or EUR.
Example Shock Calibration for JPY (Illustrative)
| Maturity | Parallel Up | Parallel Down | Steepener | Flattener |
|---|---|---|---|---|
| O/N - 1M | +150bp | -100bp | 0bp | +200bp |
| 3M - 6M | +150bp | -100bp | +25bp | +175bp |
| 1Y | +150bp | -100bp | +50bp | +150bp |
| 2Y - 5Y | +150bp | -100bp | +100bp | +100bp |
| 10Y | +150bp | -100bp | +150bp | +50bp |
| 20Y+ | +150bp | -100bp | +200bp | 0bp |
(Note: Actual BCBS calibrations are updated periodically and vary by supervisory authority. Japanese banks use FSA-prescribed shocks.)
Worked Example: ΔEVE Calculation for a JGB Portfolio
Let's calculate the EVE sensitivity for a simplified regional bank balance sheet with a ¥100 billion JGB portfolio.
Baseline Balance Sheet
Assets:
- ¥50bn in 5Y JGBs @ 0.3% yield, price ¥99.50, duration 4.85
- ¥50bn in 10Y JGBs @ 0.7% yield, price ¥97.20, duration 9.50
- Total Asset Value: ¥98.35bn
Liabilities:
- ¥90bn in retail deposits @ 0.01% (repricing immediately, effective duration ≈ 0.1)
- ¥10bn in 3Y wholesale funding @ 0.25%, duration 2.90
- Total Liability Value: ¥100bn
Net Equity (Economic): ¥98.35bn - ¥100bn = -¥1.65bn (accounting shows positive equity via HTM classification, but economic value is negative)
Scenario: Parallel Up (+150bp Shock)
Step 1: Reprice Assets
For 5Y JGBs with duration 4.85:
\[\%\Delta P \approx -D_{mod} \times \Delta y = -4.85 \times 0.015 = -7.275\%\]New 5Y value: ¥49.75bn × (1 - 0.07275) = ¥46.13bn
For 10Y JGBs with duration 9.50:
\[\%\Delta P \approx -9.50 \times 0.015 = -14.25\%\]New 10Y value: ¥48.60bn × (1 - 0.1425) = ¥41.67bn
Total Assets (shocked): ¥46.13bn + ¥41.67bn = ¥87.80bn
Step 2: Reprice Liabilities
Retail deposits (duration 0.1):
\[\%\Delta P \approx -0.1 \times 0.015 = -0.15\%\]New deposit value: ¥90bn × (1 - 0.0015) = ¥89.865bn
3Y wholesale funding (duration 2.90):
\[\%\Delta P \approx -2.90 \times 0.015 = -4.35\%\]New funding value: ¥10bn × (1 - 0.0435) = ¥9.565bn
Total Liabilities (shocked): ¥89.865bn + ¥9.565bn = ¥99.43bn
Step 3: Calculate ΔEVE
Net equity (shocked): ¥87.80bn - ¥99.43bn = -¥11.63bn
ΔEVE = -¥11.63bn - (-¥1.65bn) = -¥9.98bn loss
EVE Impact as % of Tier 1 Capital: If the bank has ¥15bn Tier 1 capital, this represents a 66.5% erosion of capital—well above typical regulatory thresholds of 15-20%.
Why This Matters
This bank faces massive duration mismatch risk:
- Long-duration assets (JGBs with 4-10 year duration)
- Short-duration liabilities (deposits repricing immediately)
- When rates rise, asset values fall far more than liability values
This is the core risk Japanese regional banks faced during the 2024 YCC exit, when 10Y JGB yields surged from 0.25% to 1.0%.
NII vs. EVE: The Earnings Paradox
Interestingly, the same +150bp shock that devastates EVE can improve NII for banks with positive duration gaps.
NII Impact of +150bp Shock (Same Bank)
Before Shock (Annual NII):
- Asset income: (¥50bn × 0.3%) + (¥50bn × 0.7%) = ¥500 million
- Liability cost: (¥90bn × 0.01%) + (¥10bn × 0.25%) = ¥34 million
- Net Interest Income: ¥466 million
After Shock (Year 1 Forward NII):
Assuming the 5Y and 10Y JGBs are held-to-maturity and don't reprice, but deposits reprice immediately:
- Asset income: ¥500 million (unchanged—existing bonds locked in)
- Liability cost: (¥90bn × 1.51%) + (¥10bn × 0.25%) = ¥1,384 million
- Net Interest Income: -¥884 million (becomes negative!)
ΔNII = -¥884M - ¥466M = -¥1,350 million
Wait—this shows negative NII impact! Why?
The Key: In this example, deposits reprice immediately (paying higher rates to customers) while JGB coupon income stays fixed. This is the nightmare scenario for banks: economic value loss (EVE) and earnings collapse (NII).
Alternative Scenario: Steepener Shock
Under a steepener (short rates flat, long rates +200bp), the bank might see:
- EVE deteriorates (long-dated JGBs fall in value)
- NII stable or improves (deposit costs unchanged, asset reinvestment at higher yields)
This asymmetry—EVE and NII moving in opposite directions—is why IRRBB requires measuring both metrics.
Regulatory Thresholds and Supervisory Outliers
Under Basel III Pillar 2, banks are flagged as "outliers" if:
\[\frac{\text{Maximum } \Delta EVE \text{ across 6 scenarios}}{\text{Tier 1 Capital}} > 15\%\]For Japanese banks, the FSA typically uses a 20% threshold given the historically low rate environment.
Consequences of Outlier Status
- Enhanced reporting: Monthly IRRBB disclosures to FSA
- Capital add-ons: Pillar 2 surcharges on minimum capital requirements
- Remediation plans: Must submit detailed plans to reduce IRRBB exposure
- Public disclosure: Large banks must disclose IRRBB metrics in annual reports
Example: Outlier Calculation
Our example bank had:
- Maximum ΔEVE: -¥9.98bn (parallel up scenario)
- Tier 1 Capital: ¥15bn
- Ratio: 9.98 / 15 = 66.5%
This is far above the 20% threshold—the bank is a severe outlier and faces immediate supervisory action.
Managing IRRBB: Hedging and Behavioral Assumptions
Banks have several tools to reduce IRRBB exposure, but each comes with trade-offs.
1. Duration Matching via Asset Allocation
Strategy: Shorten asset duration by shifting from 10Y JGBs to 2Y-5Y JGBs.
Impact on IRRBB:
- ✅ Reduces ΔEVE significantly (assets and liabilities closer in duration)
- ❌ Lowers carry income (shorter bonds yield less)
- ❌ May violate liquidity requirements (need long-dated HQLA)
2. Interest Rate Swaps (IRS)
Strategy: Enter pay-fixed, receive-floating swaps to convert fixed-rate JGB exposure to floating.
Example: Bank pays 0.8% fixed on ¥50bn notional swap, receives 3M JPY TIBOR.
Impact on IRRBB:
- ✅ Reduces duration of asset side (floating-rate exposure has low duration)
- ✅ Improves ΔEVE under rising rate scenarios
- ❌ Introduces counterparty credit risk and swap margin costs
- ❌ Accounting treatment can be complex (hedge accounting rules)
3. Behavioral Assumptions for Deposits
The most controversial aspect of IRRBB is modeling non-maturity deposits (NMDs) like retail checking accounts.
The Challenge: Legally, deposits are callable overnight (duration ≈ 0), but behaviorally, retail customers don't withdraw en masse even if rates rise.
Core vs. Non-Core Deposits:
| Deposit Type | Legal Maturity | Behavioral Maturity | Effective Duration |
|---|---|---|---|
| Core retail (stable) | Overnight | 5-7 years (sticky) | 3-5 years |
| Non-core retail | Overnight | 1-2 years | 0.5-1.5 years |
| Wholesale/corporate | Overnight | 3-6 months (rate-sensitive) | 0.1-0.3 years |
By assuming core retail deposits have longer effective duration, banks can report lower IRRBB. However:
- FSA scrutinizes these assumptions heavily
- Banks must provide statistical evidence (deposit flow analysis, regression studies)
- Conservative assumptions are safer but worsen reported IRRBB
4. Grid Point Sensitivity Hedging
As discussed in Section 3.5, GPS-based hedging allows banks to neutralize specific portions of the curve:
- Hedge 10Y exposure with 10Y futures
- Hedge 30Y exposure with cash 30Y JGBs or swaps
- Leave 2Y exposure unhedged if deposits provide natural offset
This targeted approach minimizes hedging costs while addressing the most problematic IRRBB scenarios (e.g., steepeners).
Connection to Negative Yield Environments
Japan's experience with negative rates from 2016-2024 created unique IRRBB challenges:
Elevated Duration at Negative Yields
As discussed in Section 3.1, duration increases when yields are negative. A 10Y JGB at -0.10% yield has higher duration than the same bond at +0.50% yield.
IRRBB Implication: Banks holding JGBs during negative rate periods faced amplified ΔEVE when yields rose from -0.10% to +0.50% (YCC exit 2024).
Convexity Effects
See Section 3.4: During the 2024 YCC exit, banks suffered not just duration losses but negative convexity losses on premium bonds purchased during the negative rate era.
Bonds bought at ¥105-110 (trading above par due to negative yields) experienced outsized price declines when yields normalized, compounding IRRBB impacts.
Key Takeaways
- IRRBB measures two distinct risks: Economic value (ΔEVE) and earnings (ΔNII)—both are critical
- Six standard shock scenarios capture parallel shifts, steepeners, flatteners, and short-rate shocks
- Outlier threshold: ΔEVE > 15-20% of Tier 1 Capital triggers enhanced supervision
- Duration mismatch is the core driver—long-duration JGBs vs. short-duration deposits creates massive EVE sensitivity
- Hedging options: Asset allocation, interest rate swaps, GPS-targeted hedging, and behavioral deposit modeling
- Negative yields amplified risk: Higher durations and premium bond losses during 2016-2024 made IRRBB worse
- GPS framework essential: Non-parallel shocks (steepeners/flatteners) require grid-point analysis from Section 3.5
IRRBB is the primary regulatory framework governing JGB portfolio risk for Japanese banks. Understanding the EVE/NII distinction, shock scenarios, and hedging strategies is essential for anyone managing banking book interest rate risk in the post-YCC era. The metrics and concepts from Sections 3.1-3.5 all feed directly into IRRBB calculations and management.