Convexity-Adjusted Yield Spread

What Is Convexity-Adjusted Yield Spread?

The convexity-adjusted yield spread is a refined fixed income metric that modifies a raw yield spread by accounting for the dollar convexity differential between two instruments. When comparing bonds with meaningfully different convexity profiles—such as a callable corporate versus a bullet Treasury, or a mortgage-backed security versus a swap—raw yield spreads systematically overstate or understate the true carry advantage because they ignore how differently each instrument's price responds to parallel yield curve shifts.

Formally, the adjustment subtracts the convexity benefit (expressed in yield-equivalent terms) that a higher-convexity instrument provides from its nominal spread. The intuition is direct: if bond A yields 50 basis points over bond B but also carries significantly higher positive convexity, bond A's investors receive embedded insurance against large rate moves—insurance with a quantifiable market price. Ignoring this produces misleading relative value conclusions, particularly when comparing instruments with embedded optionality such as callable bonds, puttable notes, or agency mortgage-backed securities against plain-vanilla benchmarks.

The convexity value is approximated as: Convexity Value ≈ ½ × Convexity × σ², where σ is the annualized yield volatility assumption. This term represents the expected price gain from convexity per unit of rate variance—and at elevated volatility levels, it is far from trivial. At a MOVE Index reading of 150 and a 30-year Treasury carrying a convexity of roughly 350, the convexity value alone can exceed 40 basis points in yield-equivalent terms, dwarfing typical sovereign credit differentials.

Why It Matters for Traders

For relative value fixed income managers and macro traders expressing curve or spread views, failing to convexity-adjust spreads leads to systematic mispositioning. In a high-volatility rate environment, the convexity premium embedded in long-duration or positively-convex instruments can be worth 20–60+ basis points in yield-equivalent terms, rendering superficially wide spreads almost entirely explained by the convexity differential rather than genuine credit or liquidity risk compensation.

The spread is particularly critical when comparing agency MBS—which carry negative convexity due to prepayment optionality—against investment-grade corporate bonds or interest rate swaps. In MBS analysis, the convexity-adjusted spread is closely related to the option-adjusted spread (OAS), but the two diverge when the model's volatility input differs from the analyst's forward volatility assumption. Similarly, when analyzing callable sovereign debt issued by emerging market governments against hard-currency bullet bonds, the embedded call option can account for 30–50 basis points of nominal spread at typical EM rate vol levels, leaving far less residual compensation for credit or liquidity risk than the raw number implies.

Portfolio managers running duration-neutral relative value trades—long negative-convexity MBS against long positive-convexity Treasuries, for instance—must monitor this metric continuously, as a shift in the vol regime can instantly reprice which leg is cheap on an adjusted basis without any change in credit fundamentals.

How to Read and Interpret It

A positive convexity-adjusted spread after stripping out the convexity premium suggests genuine excess compensation—an attractive long opportunity on a risk-adjusted basis. A spread that collapses to zero or turns negative after adjustment signals that the nominal spread is entirely a convexity artifact, with no residual credit or liquidity premium remaining.

Practical thresholds vary by sector and regime. In agency MBS versus swaps, a convexity-adjusted spread below 20 basis points has historically indicated expensive valuations; levels above 60 basis points post-adjustment have corresponded to attractive entry points for real money buyers and bank treasury accounts. For callable investment-grade corporates, a post-adjustment spread sitting below the OAS of comparable bullet bonds from the same issuer typically indicates richness and warrants a switch into the bullet. Conversely, when the adjusted spread is materially wider than the OAS of the bullet equivalent, the callable bond is offering genuine value beyond mere optionality compensation—a signal that has historically preceded strong total return performance for callable IG credit over 6–12 month horizons.

Historical Context

The relevance of this metric became acute during the 2022 Fed tightening cycle, one of the most aggressive in modern history. Rate volatility, as measured by the MOVE Index, surged from approximately 80 in early 2022 to over 160 by autumn—a near doubling. Agency MBS nominal spreads to Treasuries widened sharply to around 150 basis points by late 2022, levels that appeared superficially attractive to yield-seeking investors accustomed to the 2010–2021 low-vol regime.

However, on a convexity-adjusted basis, with implied vol running hot and prepayment speeds collapsing (extending effective duration sharply), the negative convexity embedded in current-coupon MBS consumed roughly 60–80 basis points of that nominal spread. The true adjusted carry was far less compelling than headline numbers suggested. Investors who anchored to nominal spread widened their MBS allocations, only to suffer disproportionate mark-to-market losses as duration extension amplified price declines. Fed officials and primary dealers who tracked convexity-adjusted spreads had flagged this disconnect publicly as early as mid-2022.

An earlier episode occurred during the 2003 mortgage convexity unwind, when rapid prepayment speeds forced mortgage servicers and GSE portfolios to simultaneously buy duration to rehedge, compressing convexity-adjusted MBS spreads to near zero and contributing to a violent bear steepening of the Treasury curve in a matter of weeks.

Limitations and Caveats

The adjustment depends critically on the volatility assumption used. Different desks applying historical vol, at-the-money implied vol, or regime-conditioned vol will arrive at materially different convexity-adjusted spreads for identical instruments—differences that can easily span 20–40 basis points. There is no market consensus on the correct input, which means two rational analysts can reach opposite relative value conclusions from the same bond prices.

Additionally, the standard formula assumes a parallel yield shift, ignoring convexity contributions from non-parallel curve moves—twist and butterfly reshapings—which in practice can be equally important. For instruments with path-dependent optionality such as prepayable mortgages or extendable notes, the simple analytical convexity calculated from a yield-price schedule may poorly approximate true effective convexity under full Monte Carlo simulation, particularly in fat-tailed rate distributions. Practitioners should treat the analytical formula as a first-order approximation and calibrate it against model-based OAS frameworks when precision matters.

Finally, the metric captures price convexity but not cash flow convexity—the variability of coupon timing under different rate paths. For floating-rate instruments with embedded caps or floors, a convexity adjustment based on duration alone will be fundamentally incomplete.

What to Watch

• MOVE Index levels — higher rate vol mechanically increases the convexity value embedded in long-duration and positively-convex instruments, compressing adjusted spreads even when nominal spreads are unchanged
• Agency MBS OAS versus convexity-adjusted spreads — the divergence between these two metrics signals whether the market's implied vol is above or below the analyst's vol assumption, a useful cross-check on positioning
• Callable corporate bond issuance volumes — surge issuance from highly-rated borrowers can rapidly reprice the convexity premium across the IG credit market, widening adjusted spreads on outstanding callables
• Interest rate skew and vol surface shape — right-skewed vol (rates rising more than falling) asymmetrically affects the adjustment for callable versus puttable bonds and deserves separate treatment in high-inflation regimes
• Duration extension risk in MBS — as prepayment speeds slow, effective duration lengthens and negative convexity worsens, directly degrading convexity-adjusted spreads independent of any nominal spread movement