Bond Convexity: Price-Yield Curvature, Duration Adjustment, and Negative Convexity

What is Convexity?

Bond convexity measures the curvature of a bond's price-yield relationship.

Modified Duration provides a first-order, straight-line estimate of how price changes when yield changes. Convexity adds a second-order adjustment that recognizes the actual relationship is curved.

A common approximation is:

1ΔP / P
2≈
3-D_mod × Δy
4+
50.5 × Convexity × (Δy)^2

where:

• ΔP / P is the estimated percentage price change;
• D_mod is modified duration;
• Δy is the change in yield in decimal form; and
• Convexity is the convexity measure using a compatible convention.

For an ordinary option-free fixed-rate bond, convexity is generally positive.

That means the bond tends to gain more when yields fall than a duration-only line predicts and lose less when yields rise than that same line predicts.

Why duration alone is incomplete

Bond prices are present values of future cash flows.

As the discount rate changes, the price does not move along a perfect straight line.

If you plot price on one axis and yield on the other, a conventional option-free bond produces a curved relationship.

Modified duration approximates the slope at the current yield.

That works well for a small movement near the starting point.

But the farther the yield moves, the more the tangent line and the actual curve separate.

Convexity measures that curvature and helps correct the error.

A useful hierarchy is:

1Duration
2-> first-order sensitivity
3-> local slope
4
5Convexity
6-> second-order sensitivity
7-> curvature around that slope

Neither replaces a full repricing model, but together they give a more informative approximation than duration alone.

Positive convexity creates asymmetric price behavior

Suppose an option-free bond has the same magnitude yield move in either direction.

With positive convexity:

1Yield falls 50 bps
2-> actual price gain tends to be larger than duration-only estimate
3
4Yield rises 50 bps
5-> actual price loss tends to be smaller than duration-only estimate

This is favorable curvature.

It does not mean the bond cannot lose money. A sufficiently large rise in yields can still cause a substantial price decline.

Convexity means the shape of that response is better than a straight symmetric duration line would imply.

A hypothetical duration-plus-convexity example

Assume a bond has:

1Modified duration: 7.0
2Convexity:         60
3Yield change:      +100 bps

Convert the yield move:

1100 bps = 0.0100

The duration-only estimate is:

1-7.0 × 0.0100 = -7.00%

The convexity adjustment is:

10.5 × 60 × (0.0100)^2
2= 0.30%

The combined approximation is therefore:

1-7.00% + 0.30% = -6.70%

If yields instead decline 100 bps:

1Duration effect:   +7.00%
2Convexity effect:  +0.30%
3Combined estimate: +7.30%

The same positive convexity adjustment improves both sides of the duration-only estimate.

This example is hypothetical and uses a simplified annual convexity convention. Investors should use duration and convexity figures calculated on compatible bases.

Convexity matters more for larger yield changes

Notice that the convexity term uses:

1(Δy)^2

If the yield change doubles, the second-order term grows by a factor of four.

That means convexity has relatively little impact for very small yield changes but becomes more important as the shock grows.

Convexity also tends to matter more for longer-maturity and lower-coupon bonds because their price-yield curves can exhibit more curvature.

This is why a duration-only approximation that looks excellent for a 10 bp move can be noticeably wrong for a 150 bp move.

Option-free fixed-rate bonds usually have positive convexity

For a standard fixed-rate bond without embedded options, price generally rises at an increasing rate as yields fall and falls at a decreasing rate as yields rise.

That produces positive convexity.

All else equal, CFA Institute's fixed-income framework notes relationships similar to duration:

• longer maturity tends to increase convexity;
• lower coupon tends to increase convexity; and
• lower yield tends to increase convexity.

Those relationships help explain why long-duration bonds can require a larger convexity correction than short-duration bonds.

But the word "positive" should not be confused with a statement that the bond is attractive at any price.

More convexity is not free

Investors often prefer positive convexity, all else equal.

The phrase all else equal does a great deal of work.

If two otherwise identical bonds offered the same yield and duration but one had more positive convexity, the higher-convexity bond would have a more favorable price response to large yield moves.

Markets generally do not give valuable features away without affecting price or yield.

A bond with more desirable convexity may trade at a higher price or lower yield.

An investor should therefore avoid a simplistic ranking:

1Higher convexity = automatically better investment

The relevant question is whether the investor is adequately compensated for the full package of yield, duration, credit risk, liquidity, optionality, and convexity.

Callable bonds can have negative convexity

Embedded options can radically alter the price-yield curve.

A callable bond gives the issuer the right to redeem the bond early.

When yields fall, the issuer's incentive to call a high-coupon bond can increase.

That limits the investor's upside because the attractive cash-flow stream may end early.

As a callable bond approaches the region where the call option is economically important, its price can rise less when yields fall than it falls when yields rise.

That behavior can produce negative convexity.

A simplified picture is:

1Option-free bond
2-> falling yields can produce strong price appreciation
3
4Callable bond near call region
5-> falling yields increase value of issuer's call option
6-> price appreciation becomes capped
7-> convexity can turn negative

This is a core reason Yield to Worst and Option-Adjusted Spread matter for callable securities.

Mortgage-backed securities can also exhibit negative convexity

Mortgage borrowers often have the right to prepay their loans.

When rates fall, refinancing activity can accelerate.

For an investor in mortgage-backed securities, that can return principal sooner just when higher-coupon cash flows have become more valuable.

When rates rise, refinancing can slow, extending the expected life of the cash flows when lower prices are undesirable.

That pattern can create negative convexity:

1Rates fall -> prepayments can accelerate -> upside constrained
2Rates rise -> prepayments can slow       -> duration can extend

This makes static yield-based duration and convexity less appropriate for many mortgage securities. Option-aware valuation and effective risk measures become important.

Effective convexity versus yield convexity

Like duration, convexity comes in more than one form.

Yield convexity examines curvature relative to changes in a bond's own yield under a fixed-cash-flow framework.

Effective convexity can be estimated by revaluing a bond after upward and downward shifts in a benchmark yield curve while allowing modeled cash flows to respond.

A common numerical structure is:

1Effective Convexity
2≈
3(P_down + P_up - 2P_0)
4------------------------
5 P_0 × (Δcurve)^2

where the precise scaling depends on convention.

For option-sensitive securities, effective measures are generally more informative because the expected cash flows are not fixed.

The important investor habit is to confirm which convexity measure is being displayed before comparing securities.

Convexity is not the same as volatility

Convexity describes the shape of price sensitivity to yield changes.

It does not tell you how volatile yields themselves will be.

A high-convexity bond in a calm rate environment can experience less realized price volatility than a lower-convexity bond exposed to a much more volatile spread or curve segment.

Likewise, a bond can have positive convexity and significant credit risk.

Convexity is a response function, not a probability distribution for future rate moves.

Credit spreads can move independently of benchmark rates

A corporate bond's yield can be decomposed conceptually into a benchmark rate plus a spread.

If Treasury yields fall but Credit Spread widens sharply, the corporate bond's own yield may not fall by the same amount.

A duration-plus-convexity estimate based only on one assumed yield shock can therefore miss the source and shape of the actual market move.

Portfolio managers may use separate curve, spread, and empirical risk measures when those distinctions matter.

This is especially important during market stress, when benchmark-rate changes and credit-spread changes can behave differently from a simple parallel-shift assumption.

Convexity at the portfolio level

For a portfolio of conventional bonds, managers often summarize convexity with market-value-weighted component measures.

That can provide a practical approximation, but it still assumes the underlying yield moves are represented consistently.

Real yield curves can steepen, flatten, twist, or move differently by sector.

A single portfolio convexity number does not capture every key-rate exposure or spread interaction.

Portfolio convexity is therefore useful as a second-order summary, not a complete scenario engine.

Why convexity is a good candidate for interactive analysis

Convexity is difficult to appreciate from a definition alone.

The concept becomes clearer when a reader can change:

• starting yield;
• duration;
• convexity; and
• yield shock

and compare:

1duration-only estimate
2versus
3duration-plus-convexity estimate

That is a strong future Research Tool opportunity for the encyclopedia.

TC11 deliberately does not create that tool yet. Research Tool authority remains separately reviewed so an interactive module can be designed once, tested carefully, and reused across the fixed-income concept graph rather than being improvised inside one article.

How investors should use convexity

A useful fixed-income workflow is:

1. Start with Yield to Maturity or YTW to understand the yield framework.
2. Use duration to estimate first-order rate sensitivity.
3. Add convexity when the yield move is material or the bond has substantial curvature.
4. Use effective measures when cash flows can change with rates.
5. Separate interest-rate changes from credit-spread changes where possible.
6. Evaluate the yield or price cost of desirable convexity rather than ranking the measure in isolation.

For broader macro and valuation context, investors can continue into the Macroeconomic Conditions Index and the Cyclically Adjusted Risk Premium. These are separate research surfaces, not live bond-sensitivity calculators for this article.

Sources and further reading

• CFA Institute, 2026, Yield-Based Bond Convexity and Portfolio Properties: https://www.cfainstitute.org/insights/professional-learning/refresher-readings/2026/yield-based-bond-convexity-and-portfolio-properties
• CFA Institute, 2026, Yield-Based Bond Duration Measures and Properties: https://www.cfainstitute.org/insights/professional-learning/refresher-readings/2026/yield-based-bond-duration-measures-and-properties
• CFA Institute, 2026, Curve-Based and Empirical Fixed-Income Risk Measures: https://www.cfainstitute.org/insights/professional-learning/refresher-readings/2026/curve-based-and-empirical-fixed-income-risk-measures
• CFA Institute, 2026, Valuation and Analysis of Bonds with Embedded Options: https://www.cfainstitute.org/insights/professional-learning/refresher-readings/2026/valuation-analysis-bonds-embedded-options

Explore more topics in the Financial Research Encyclopedia.