# Agency MBS Negative Convexity & 10Y Treasury Spillover

> Quantify UMBS 30-year current-coupon mortgage spreads, negative convexity duration extension, servicer delta-hedging flows, and 10-year Treasury amplification.

Canonical: https://alphapulsefin.com/analysis/mbs-current-coupon-convexity-extension-treasury-volatility-spillover
Last updated: 2026-11-14T23:02:55.003Z

**Publication:** AlphaPulse Finance
**Category:** Fomc Treasury Term Premium
**Published:** 2026-11-14T11:15:25.915Z
**Author:** Dr. Marcus Vance, CFA (CFA, PhD in Financial Economics (Wharton))
**Reviewer:** Not recorded

## Answer-first summary

Negative convexity across $9.1T of agency MBS forces mortgage servicers and money managers to shed duration as yields rise, amplifying 10-year US Treasury volatility during macro rate selloffs.

## Key takeaways

- A +50 bps Treasury selloff extends duration on the $432B actively hedged 5.5%-6.5% UMBS cusp cohort by ~0.98 years, forcing ~$51.7B of 10Y-equivalent hedge selling.
- Deep-discount 2.0%-3.0% pandemic MBS pools ($4.85T) sit 300+ bps out-of-the-money and exhibit positive convexity, muting economy-wide extension compared to 2003.
- Current-coupon UMBS spreads (+136 bps over Treasuries) decompose into 88 bps of volatility-driven prepayment option cost and +48 bps of clean OAS.
- Falling swaption implied volatility (MOVE index compression) directly tightens current-coupon MBS nominal spreads even if OAS stays constant.

## Related financial entities

- **Federal Reserve System** (FED) — company: The central banking system of the United States responsible for FOMC monetary policy, reserve management, and systemic financial stability.
- **SOMA Reserve Management Purchases (RMP) T-Bill Operations** (SOMA-RMP) — product: SOMA Reserve Management Purchases (RMP) T-Bill Operations (SOMA-RMP) refers to an institutional financial benchmark and regulatory standard monitored by Federal Reserve SOMA Desk (US), currently calibrated at $3.12T (Bank Reserves) as of October 2026.
- **U.S. Treasury 10-Year Benchmark Note Auction Liquidity** (UST-QRA-10Y) — product: U.S. Treasury 10-Year Benchmark Note Auction Liquidity (UST-QRA-10Y) refers to an institutional financial benchmark and regulatory standard monitored by U.S. Department of the Treasury (US), currently calibrated at 4.32% (Par Yield) as of October 2026.

## Analysis

Standing at $9.1 trillion outstanding, the US agency mortgage-backed securities (MBS) market represents the second-largest fixed-income asset class in the world—and embeds a massive short Bermudan swaption held by American homeowners, who may prepay fixed-rate 30-year mortgages at par without penalty whenever rates drop. When [10-Year Benchmark Treasury Yields](/entities/ust-qra-10y-u-s-treasury-10-year-benchmark-note-auction-liquidity) sell off by 50 bps across key refinancing cusps, expected mortgage cash flows lengthen ("extension risk"), forcing mortgage servicing rights (MSR) hedgers and active MBS portfolios to pay fixed on swaps or sell 10-year Treasury futures—creating a pro-cyclical feedback loop into sovereign term premia.

## Seasoned Lock-In Pools vs High-Coupon 5.5%-6.5% Cusp Convexity

Post-2022 rate dispersion bifurcated the UMBS 30-year universe into two distinct convexity cohorts. Roughly 62% of outstanding pools carry deep discount 2.0%-3.5% coupons held heavily inside [Federal Reserve](/entities/federal-reserve-system) SOMA (~$2.18T) and commercial bank Held-to-Maturity (HTM) portfolios under [OCC](/regulators/office-of-the-comptroller-of-the-currency-u-s-155) and [FDIC](/regulators/federal-deposit-insurance-corporation-u-s-158) supervision; sitting 250+ bps out-of-the-money for refinancing, those deep discounts exhibit mild positive convexity like bullet bonds.

Conversely, the $2.4 trillion vintage of 5.50%, 6.00%, and 6.50% UMBS pools originated between 2023 and 2026 sits squarely "on the refinancing cusp" when primary 30-year mortgage rates trade near 6.05%-6.35%. Across cusp coupons, empirical Option-Adjusted Duration (OAD) swings violently from 3.4 years at a 3.80% 10-year Treasury yield to 6.8 years at a 4.80% 10-year yield.

![Negative convexity feedback loop from rising 10-year Treasury yields to MBS duration extension](/api/figure?type=flow&title=Agency%20MBS%20Convexity%20Extension%20%26%20Treasury%20Feedback%20Loop&subtitle=How%20rising%2010Y%20yields%20slow%20CPR%20prepayments%20and%20trigger%20mechanical%20duration%20selling&steps=Initial%20%2B25%20Bps%20UST%20Shock%3A%3A10Y%20yield%20rises%20from%204.32%25%20to%204.57%25%3B%20mortgage%20rates%20top%206.45%25%7CRefi%20Incentive%20Drops%3A%3AProjected%20CPR%20on%20UMBS%206.0%25%20pools%20falls%20from%2016%25%20to%208%25%7COAD%20Extension%20%28%2B0.85%20Yr%29%3A%3ACusp%20MBS%20effective%20duration%20stretches%20from%204.6%20to%205.45%20years%7CServicer%2FREIT%20Delta%20Hedge%3A%3ADesks%20sell%20~%2436M%20DV01%20of%20TY%2FTN%20futures%2C%20pushing%2010Y%20higher&metrics=UMBS%20Current%20Coupon%3A%3A5.68%25%3A%3A%2B136%20bps%20over%205Y%2F10Y%20UST%7COption%20Cost%20Wedge%3A%3A88%20bps%3A%3AEmbedded%20refi%20call%20value%7CCurrent-Coupon%20OAS%3A%3A%2B48%20bps%3A%3AVol-adjusted%20net%20spread%7CCusp%20Pool%20Cohort%3A%3A%242.40T%3A%3AUMBS%205.5%25%20-%206.5%25%20stack)

## Current-Coupon Nominal Spread vs OAS Decomposition

Par current-coupon UMBS 30-year yields (~5.68% secondary TBA) trade roughly +136 bps above interpolated 5Y/10Y Treasury blends—wider than the +95 bps pre-2020 norm. Decomposing that +136 bps nominal spread reveals ~88 bps of prepayment option cost (driven by elevated swaption implied volatility) leaving +48 bps of pure Option-Adjusted Spread (OAS) compensation for bank and real-money buyers executing TBA trades reported to [FINRA](/regulators/finra) TRACE.

Because [SOMA Reserve Management](/entities/soma-rmp-soma-reserve-management-purchases-rmp-t-bill-operations) passively lets $13.8B/month of MBS roll off while commercial banks cap HTM additions, active money managers and REITs absorb net TBA supply through primary dealers like [Cantor Fitzgerald](/institutions/cantor)—hedging convexity dynamically in [CFTC](/regulators/cftc)-cleared swaps and futures.

```formula
Nominal_MBS_Spread = OAS + Option_Cost(Implied_Swaption_Vol)
Delta_DV01_Hedge = -Notional_Hedged * Option_Adj_Convexity * Delta_y * 0.01

Worked +50 Bps Selloff Extension Flow ($432B Actively Delta-Hedged Cusp Cohort, OAC = -1.95):
- Duration Extension per $100 Par = -(-1.95) * (+0.50) = +0.975 years of added OAD
- Incremental Market DV01 Created = $432B * 0.975 * 0.0001 = $42.12M of unwanted DV01
- Equivalent 10-Year Note Selling Needed (ModDur 8.15) = $42.12M / 0.000815 = $51.68B 10Y-Equiv
```

## Convexity Hedging DV01 Amplification Formula

Quantifying aggregate market duration extension per 25-bps Treasury selloff shows how delta-hedgers inject mechanical selling pressure into the 10-year point.

![Option-adjusted duration and convexity profile across UMBS 30-year 2.5% to 6.5% coupon pools](/api/figure?type=bars&title=UMBS%2030-Year%20Coupon%20Stack%3A%20OAD%20%26%20Negative%20Convexity&subtitle=Option-adjusted%20convexity%20across%20seasoned%20lock-in%20discounts%20vs%20active%20TBA%20production&steps=UMBS%202.5%25%20%28%242.9T%20Lock-In%29%3A%3A7.4%20yr%20OAD%2C%20%2B0.28%20convexity%20-%3E%20trades%20like%20a%20positive-gamma%20note%7CUMBS%204.0%25%20%28%241.6T%20Belly%29%3A%3A6.3%20yr%20OAD%2C%20-0.35%20convexity%20-%3E%20mild%20extension%20sensitivity%7CUMBS%205.5%25%20%28%241.3T%20Cusp%29%3A%3A5.1%20yr%20OAD%2C%20-1.65%20convexity%20-%3E%20peak%20refi%20S-curve%20inflection%7CUMBS%206.0%25%20%28%241.1T%20Active%29%3A%3A4.6%20yr%20OAD%2C%20-2.10%20convexity%20-%3E%20heaviest%20MSR%20%26%20REIT%20hedging&metrics=Total%20Agency%20MBS%3A%3A%249.10T%3A%3AUMBS%20%2B%20GNMA%20universe%7CActive%20Hedged%20Share%3A%3A~18%25%3A%3AMSRs%2C%20REITs%20%26%20hedge%20funds%7C25-Bps%20Hedge%20Flow%3A%3A%2436M%20DV01%3A%3A10Y-equiv%20futures%20selling%7CTRACE%20TBA%20Liquidity%3A%3A%24290B%2Fday%3A%3A2nd%20most%20liquid%20bond%20mkt)

## UMBS 30-Year Coupon Stack Convexity & Spread Profile

Comparing deep-discount SOMA pools against active TBA production coupons highlights where extension gamma concentrates across the coupon stack.

| UMBS 30Y Cohort | Outstanding Float | Option-Adj Duration | Option-Adj Convexity | Primary Holder & Regulatory Regime |
| --- | --- | --- | --- | --- |
| UMBS 2.0% - 3.0% Discount | $4.85T | 7.2 - 7.6 yrs | +0.15 to +0.32 | SOMA & Bank HTM ([OCC](/regulators/office-of-the-comptroller-of-the-currency-u-s-155) / [FDIC](/regulators/federal-deposit-insurance-corporation-u-s-158)) |
| UMBS 3.5% - 4.5% Seasoned | $1.85T | 5.9 - 6.6 yrs | -0.25 to -0.65 | Core Bond Mutual Funds ([SEC](/regulators/sec) 1940 Act) |
| UMBS 5.5% Production TBA | $1.30T | 5.10 yrs | -1.65 | TBA Forward Dealers ([Cantor](/institutions/cantor) / [FINRA](/regulators/finra)) |
| UMBS 6.0% - 6.5% Cusp | $1.10T | 4.1 - 4.6 yrs | -2.10 to -2.45 | MSR Servicers & REITs ([CFTC](/regulators/cftc) Swap Hedges) |

## Why Modern Convexity Selloffs Are Smaller Than 2003 Yet Still Potent

During July-August 2003, aggressive delta-hedging by Fannie Mae and Freddie Mac retained portfolios drove 10-year Treasury yields up 145 bps in six weeks. Post-conservatorship GSE portfolio caps eliminated GSE-driven convexity cascades, leaving dynamic MSR servicers and leveraged MBS REITs (~18% hedged share of the $2.4T cusp cohort) as the residual channel—adding an estimated +6 to +11 bps of excess 10-year yield amplification during rapid 50-bps selloffs.

Desk replication notes span [FOMC Policy Path, SOMA Runoff & US Treasury Term Premia](/analysis/fomc-soma-runoff-us-treasury-term-premium-architecture), [ICE BofA MOVE Index & 1y10y Swaption Volatility](/analysis/move-index-swaption-vol-surface-term-premium-feedback-loops), and [10-Year US Treasury Auction Tails, Indirect](/analysis/ten-year-treasury-auction-tail-indirect-bidder-stop-out-analytics), organized across [FOMC Policy Path](/topics/fomc-treasury-term-premium) and [tracked macro entities](/entities).

## Frequently asked questions

### What is negative convexity in agency mortgage-backed securities?

Because homeowners refinance when mortgage rates drop and keep existing low-rate loans when rates rise, an MBS bond shortens in duration during rallies (capping price upside above par) and lengthens during selloffs (accelerating price losses).

### Why do mortgage servicing rights (MSR) holders buy Treasuries when yields fall and sell when yields rise?

MSRs act like interest-only (IO) strips whose value rises when prepayments slow; to keep net portfolio duration stable, MSR hedgers sell Treasury futures when yields rise and buy duration back when yields fall.

### What is the difference between MBS nominal Z-spread and Option-Adjusted Spread (OAS)?

Nominal spread compares static single-scenario MBS cash flows to Treasuries, whereas OAS runs Monte Carlo interest-rate paths to subtract the fair market value of the homeowner’s prepayment call option.

### Why does TBA (To-Be-Announced) trading make agency MBS nearly as liquid as US Treasuries?

TBA rules collapse tens of thousands of individual Freddie Mac and Fannie Mae UMBS mortgage pools into fungible forward contracts by coupon and settlement month, concentrating $290B+ of daily volume reported to FINRA TRACE.


## Source URLs

- https://www.finra.org/filing-reporting/trace/securitized-products-data
- https://www.newyorkfed.org/markets/ambs/ambs_schedule
- https://www.federalreserve.gov/econres/feds/mortgage-backed-securities-and-interest-rate-volatility.htm

## Citation

- APA: Dr. Marcus Vance, CFA. (2026). Agency MBS Current-Coupon Spreads & Negative Convexity Hedging into 10Y Treasury Selloffs. AlphaPulse Finance. https://alphapulsefin.com/analysis/mbs-current-coupon-convexity-extension-treasury-volatility-spillover
- Canonical article: [Agency MBS Current-Coupon Spreads & Negative Convexity Hedging into 10Y Treasury Selloffs](https://alphapulsefin.com/analysis/mbs-current-coupon-convexity-extension-treasury-volatility-spillover)
- Site research index: [AlphaPulse Finance](https://alphapulsefin.com/llms-full.txt)
