Last Updated: 2026-01-27
Rate Risk vs Credit Spread Risk
Duration matching (ALDM) is designed to protect against credit spread risk, not interest rate risk. This distinction is fundamental to understanding when duration matching provides value and when it doesn't.
Why the Distinction Matters
Credit spread risk and interest rate risk behave fundamentally differently:
| Risk Type | Behavior | SSR Response | Duration Matching Applicability |
|---|---|---|---|
| Credit spread widening | Mean-reverting; cyclical | SSR stays flat or falls (flight to quality) | ALDM protects — temporary MTM loss, no cash flow mismatch |
| General rate rise | Can be permanent (regime shift) | SSR rises | ALDM does NOT protect — ongoing cash flow mismatch |
The Mechanism
Why credit spread risk is manageable with duration matching:
- Credit spread widening causes asset prices to fall temporarily
- But SSR (Sky Savings Rate) doesn't rise — it tracks general rate levels, not credit spreads
- During credit stress, USDS often benefits from flight-to-quality, potentially lowering SSR
- The Prime has no cash flow mismatch — it can afford to wait
- As credit spreads compress (empirically mean-reverting), asset prices recover
- Duration matching ensures the Prime has time to wait for recovery
Why rate risk is NOT manageable with duration matching:
- General rate rise (e.g., Fed hikes) causes SSR to rise permanently
- If the Prime holds fixed-rate assets, it earns the old lower rate
- This creates ongoing negative carry: pays SSR + margin, earns old rate
- This isn't a temporary MTM shock — it's permanent cash flow drag
- Duration matching doesn't help because there's no "pull to par" on the rate differential itself
- The Prime will bleed continuously until the asset matures or rates fall
Empirical Support
Credit spreads are mean-reverting:
- Research shows "significant evidence of mean reversion, especially for higher-rated spreads"
- Credit spread indices are modeled using Ornstein-Uhlenbeck processes (mean-reverting systems)
- 2008 GFC: ~6+ months to recover; COVID-19 2020: ~3 weeks after Fed intervention
- Counter-cyclical behavior: widen during contractions, narrow during expansions
Interest rates can shift permanently:
- Monetary policy regime changes are documented (e.g., Volcker era)
- The move from ~15% rates (1980s) to ~0% (2010s) wasn't mean reversion — it was a multi-decade regime shift
- Recent research challenges the assumption that monetary policy is "neutral" in the long run
Rate Hedging Requirement
All Prime fixed-rate exposure must be rate-hedged. Duration matching eligibility requires that positions be rate-neutral relative to SSR.
Methods of Rate Hedging
| Method | Description | When to Use |
|---|---|---|
| Floating-rate assets | Asset yield tracks market rates naturally | Preferred for CLOs (most are floating-rate) |
| Interest rate swaps | Swap fixed receipts for floating | Convert fixed-rate bonds to floating exposure |
| Duration matching | Match asset duration to liability duration | When liabilities have predictable duration |
| Rate hedging capital | Hold extra capital to cover expected rate loss | When hedging instruments unavailable or costly |
Rate Hedging Capital Calculation
If a Prime holds unhedged fixed-rate exposure, it must hold capital to cover the expected loss from rate movements:
Rate Hedge Capital = Fixed Rate Exposure × Duration × Expected Rate Volatility × Confidence Multiplier
Example:
- $100M fixed-rate bonds, 3-year duration
- Expected rate volatility: 200bps at 95% confidence
- Rate Hedge Capital = $100M × 3 × 2% × 1.65 = $9.9M
This capital is in addition to credit risk capital, not a substitute for it.
Duration Matching Eligibility Summary
For an asset to be eligible for matched treatment, it must satisfy both conditions:
- Has stressed pull-to-par ≤ matched liability tier duration (existing requirement)
- Is rate-neutral relative to SSR — either:
- Asset is floating-rate (natural hedge), OR
- Asset is hedged via swap/derivative, OR
- Prime holds rate hedging capital for unhedged fixed-rate portion
Match Eligible = (Has Pull-to-Par) AND (Rate Neutral OR Rate Hedge Capital Held)
What This Means for Asset Types
| Asset | Rate Exposure | Typical Handling | Match Eligible? |
|---|---|---|---|
| JAAA (CLO AAA) | Floating-rate (SOFR + spread) | Natural hedge | ✓ Yes |
| Fixed-rate corporate bonds | Fixed-rate | Must swap to floating or hold rate capital | Conditional |
| T-bills (short duration) | Fixed but short | Minimal rate risk due to short duration | ✓ Yes |
| Long-duration treasuries | Fixed, long duration | Must hedge or hold significant rate capital | Conditional |
| Sparklend | Floating-rate (typically) | N/A — no pull-to-par regardless | ✗ No (no SPTP) |
The Value Proposition of Duration Matching
With rate risk properly hedged, duration matching (ALDM) allows Primes to:
- Avoid hedging credit spread risk on long-duration variable-rate assets
- Take credit spread exposure for yield while managing capital efficiently
- Wait out temporary credit dislocations without forced sales
This is the core value: duration matching lets Primes capture credit spread (which is compensated and mean-reverting) while requiring them to hedge rate risk (which can be permanent and catastrophic if unhedged).
The Matching Principle
Assets can be matched against liability tiers based on their stressed pull-to-par:
Matched Assets (Duration-Matched Treatment)
Condition: Asset stressed pull-to-par ≤ matched liability tier duration
Treatment:
- Forced realization probability is low
- Capital requirement = Risk Weight only
- You only need capital for fundamental risk (credit, smart contract, etc.)
Unmatched Assets (Forced-Loss Treatment)
Condition: Asset stressed pull-to-par > liability tier duration, OR no pull-to-par
Treatment:
- Forced realization probability is high
- Capital requirement = max(Risk Weight, forced-loss capital term)
- For liquid, tradable assets the forced-loss term is the stressed drawdown (FRTB-style)
- For collateralized lending positions the forced-loss term is gap risk (liquidation shortfall)
Matching Example
| Asset | Stressed Pull-to-Par | Required Bucket | If Matched | If Unmatched |
|---|---|---|---|---|
| JAAA | ~1,277 days (912d × 1.4x) | bucket 84 | ~4-5% risk weight | ~10% FRTB (≥ RW) |
| 90-day T-bill | ~90 days (no stress modifier) | bucket 6 | ~0.5% risk weight | ~2% FRTB (≥ RW) |
| 4-week T-bill | ~30 days (no stress modifier) | bucket 2 | ~0.2% risk weight | ~1% FRTB (≥ RW) |
| Sparklend | None | Cannot match | N/A | max(RW, gap risk) |
Partial Matching (Split Treatment)
When an asset position exceeds available duration capacity, the position is split into matched and unmatched portions. Each portion receives its appropriate capital treatment.
Example:
- Hold $500M JAAA (SPTP = 1,260 days, requires bucket 84)
- Cumulative duration capacity at bucket 84+: $300M
| Portion | Amount | Treatment | CRR | Capital Required |
|---|---|---|---|---|
| Matched | $300M | Duration-matched (risk weight) | 5% | $15M |
| Unmatched | $200M | FRTB (drawdown, floored by RW) | 10% | $20M |
| Total | $500M | — | — | $35M |
Calculation:
Matched Amount = min(Position Size, Available Duration Capacity at required bucket)
Unmatched Amount = Position Size - Matched Amount
Capital = (Matched Amount × Risk Weight) + (Unmatched Amount × max(Risk Weight, FRTB Drawdown))
Key property: As duration capacity grows (longer-duration liabilities accumulate), more of each position can be matched, reducing overall capital requirements. This creates natural incentive alignment — sticky liabilities enable more efficient capital deployment.