Convexity Hedging

What Is Convexity Hedging?

Convexity hedging is the forced, mechanical rebalancing of interest rate risk by holders of mortgage-backed securities (MBS), instruments that carry negative convexity. Unlike standard bonds, where duration (price sensitivity to rate changes) behaves predictably and symmetrically, MBS holders face a shape-shifting risk profile driven entirely by homeowner behavior. When rates rise, homeowners prepay less and stay in their mortgages longer, extending the bond's effective duration. When rates fall, homeowners refinance faster, shortening duration. This asymmetric, path-dependent behavior means large institutional holders, banks, insurance companies, the GSEs (Fannie Mae, Freddie Mac), and historically the Federal Reserve, must constantly rebalance their duration exposure using Treasury bonds and interest rate swaps.

The mechanics are both straightforward and powerful. When yields sell off, MBS duration extends and holders find themselves longer duration than their mandates allow. To rebalance, they must sell Treasuries or enter pay-fixed interest rate swaps, pushing yields higher still. In a rally, the reverse: falling rates trigger prepayment acceleration, shortening MBS duration and forcing holders to buy duration, amplifying the move lower in yields. This self-reinforcing feedback loop is the defining feature of convexity hedging and the reason it commands attention far beyond the fixed income desk.

Why It Matters for Traders

Convexity hedging flows represent one of the most significant structural, non-fundamental forces in the U.S. Treasury market. They can move 10-year yields by 20–40 basis points in a single session without any change in economic data or Fed communication, a fact that repeatedly confounds macro traders anchored to fundamentals. The effect is most pronounced in the 5- to 10-year sector, where the bulk of MBS duration is concentrated, though the 30-year point can also be impacted during extreme episodes.

For rates traders, these flows matter because they create violent, momentum-amplifying moves that can trigger stop-loss cascades across leveraged positions. For equity and multi-asset traders, the channel runs through real yields and financial conditions: a self-reinforcing Treasury selloff driven by convexity hedging tightens conditions abruptly, compressing price-to-earnings multiples and hitting rate-sensitive sectors, technology, utilities, REITs, with particular severity. The MOVE Index, which tracks implied volatility across Treasury maturities, often spikes sharply during convexity events with no corresponding spike in the VIX, a divergence that serves as a diagnostic signal.

The dynamic is also crucial for understanding swap spreads. When MBS holders pay fixed in swaps at scale, they depress swap rates relative to Treasuries, driving spreads sharply negative, a pattern that recurred notably during the 2019–2020 period and again in 2022.

How to Read and Interpret It

Because convexity flows are not directly reported, traders must infer them from a constellation of market signals:

• MOVE Index spiking disproportionately to VIX: The clearest real-time signature. A MOVE/VIX ratio above 7–8x historically coincides with bond technical dynamics rather than macro repricing.
• Rate of change threshold: When the 10-year yield moves more than 20–25 basis points intraday or 50+ basis points over a 5-day window, convexity rebalancing is almost certainly contributing to the velocity.
• Current coupon MBS spreads widening rapidly: Widening option-adjusted spreads (OAS) on agency MBS signal that holders are being forced to shed duration, increasing Treasury supply pressure.
• Swap spreads moving sharply negative in the 5–10 year sector: A reliable footprint of large-scale pay-fixed hedging. In early 2020, 10-year swap spreads briefly touched -30 basis points, an extreme reading consistent with massive MBS hedging flows.
• MBA Refinance Index surging or collapsing: Sharp changes in refinancing activity alter prepayment expectations and reprice the embedded optionality in MBS, triggering fresh hedging needs.
• Fed MBS portfolio size: During active QE, the Federal Reserve absorbs convexity risk from the market; during quantitative tightening, it returns that risk to private holders.

Historical Context

The most dramatic convexity event on record unfolded in spring–summer 2003, when the 10-year Treasury yield surged from approximately 3.1% to 4.6%, a 150 basis point move, in roughly six weeks following unexpectedly strong economic data. Federal Reserve research at the time estimated that MBS convexity hedging amplified the selloff substantially, with portfolio managers forced to sell an estimated $50–100 billion in Treasuries and pay-fixed swaps into a thinly bid market. The episode left lasting scar tissue across fixed income desks and prompted major dealers to build dedicated convexity risk models.

The 2013 Taper Tantrum replicated the dynamic at scale. Between May and September 2013, the 10-year yield jumped approximately 130 basis points after then-Chairman Bernanke signaled potential tapering of asset purchases. With the Fed's MBS portfolio already large but not yet dominant, private holders still carried meaningful convexity exposure, and forced selling amplified the initial move well beyond what fundamental repricing alone would justify.

More recently, the 2022 rate shock, the most aggressive Fed tightening cycle in four decades, reintroduced convexity hedging as a persistent amplifier. With the 10-year yield moving from 1.5% to over 4.2% across the year, MBS duration extended dramatically. The Fed's simultaneous pivot to QT meant the central bank was no longer absorbing this convexity risk, leaving private holders to manage the full burden. MBS OAS widened from near-zero in early 2022 to over 60 basis points by October, reflecting the scale of forced repositioning.

Limitations and Caveats

Convexity hedging flows are among the most model-dependent signals in fixed income, they are estimated through prepayment and duration models rather than directly observed, and model assumptions vary significantly across dealers. During periods of heavy Federal Reserve MBS ownership, the central bank effectively insulates the market from convexity dynamics; between 2020 and early 2022, with the Fed holding over $2.7 trillion in agency MBS, private-sector convexity hedging was near-dormant.

The signal also fails when credit and liquidity risk dominate, as in March 2020, when forced selling of all assets, including Treasuries, overwhelmed the more nuanced convexity mechanic. Additionally, the diffuse nature of MBS ownership means individual institutional hedging decisions are uncoordinated; the aggregate flow is real, but its precise timing and magnitude remain uncertain even to sophisticated participants.

What to Watch

For traders building convexity hedging into their market framework, the highest-priority indicators are:

• Fed's agency MBS portfolio size and monthly roll-off pace under QT (published in the Fed's H.4.1 release)
• MBA Refinance Index weekly, sudden directional breaks signal shifting prepayment assumptions
• Current coupon MBS OAS versus historical norms (Bloomberg, JPMorgan MBS analytics)
• 5- and 10-year swap spreads for evidence of large-scale pay-fixed flow
• MOVE Index level and rate of change, particularly relative to the VIX
• 30-year primary mortgage rate minus 10-year Treasury yield, a widening spread signals mortgage market stress and constrained MBS demand, which can itself dampen hedging activity by reducing new issuance

When several of these indicators align simultaneously, rising yields, widening MBS spreads, negative swap spreads, and an elevated MOVE, traders should treat the move in rates as having a mechanical, self-reinforcing component that may overshoot fair value before exhausting itself.