Effective convexity is a measure of how much a bond’s price sensitivity to interest rates changes as those rates move, calculated in a way that allows the bond’s future cash flows to shift with rates rather than staying fixed. It captures the curvature of the price-yield relationship, and it is the right measure whenever a bond has embedded options such as a call, a put, or the prepayment behavior of a mortgage pool. For plain, option-free bonds, simpler measures work; for anything else, effective convexity is the number that reflects reality.
What Effective Convexity Measures
Duration gives you a straight-line estimate of how a bond’s price will react to a change in yields. The real price-yield curve bends, and effective convexity quantifies that bend. Formally, it is the second derivative of the bond’s price with respect to yield, which is a technical way of saying it measures how quickly duration itself changes as rates shift up or down.
A bond with high convexity sees its duration rise rapidly when rates fall and drop rapidly when rates rise. That asymmetry works in the holder’s favor: the bond gains more in a rally than it loses in a selloff of the same size. A bond with low or negative convexity does the opposite. Knowing which category a holding falls into is the point of running the calculation.
Effective Convexity vs. Modified Convexity
Modified convexity assumes cash flows are locked in regardless of what rates do. That assumption is fine for Treasury bonds and non-callable corporates, where coupon and principal payments follow a fixed schedule. The math needs only the bond’s contractual terms and the current yield.
Effective convexity drops the assumption. For callable bonds, putable bonds, and mortgage-backed securities, the payments an investor actually receives depend on the path rates take. A falling-rate environment can prompt an issuer to call a bond early. A rising-rate environment can prompt a holder to put a bond back. Mortgage borrowers refinance or prepay at speeds nobody can pin down in advance. In each case, cash flows shift, and any convexity number that ignores those shifts will mislead. If the bond has any embedded optionality, effective convexity is the tool.
The Formula
The calculation requires three prices: the bond’s current price (P₀), its estimated price if yields rise by a small amount (P₊), and its estimated price if yields fall by that same amount (P₋). The yield shift (Δy) is typically written as a decimal, so 25 basis points becomes 0.0025. The formula is:
Effective Convexity = (P₋ + P₊ − 2 × P₀) / (P₀ × Δy²)
The numerator captures how far the two shifted prices, taken together, deviate from what a duration-only estimate would predict. The denominator scales that deviation by the current price and the square of the yield change, producing a number you can compare across bonds and maturities.
Where the Prices Come From
For option-free bonds, P₊ and P₋ can be calculated directly by discounting the fixed cash flows at the new yield. For bonds with embedded options, the shifted prices need to come from a model that reprices the option at each yield level. Binomial interest rate trees and Monte Carlo simulations are the two common approaches. The model generates many potential rate paths, decides whether the option gets exercised along each path, and averages the resulting prices. The effective convexity number is only as reliable as the pricing model producing those inputs.
Choosing the Yield Shift
The size of Δy matters. A 10-basis-point shift is standard for routine portfolio analytics because the duration-plus-convexity approximation works well for small moves. Once the shift grows past roughly 100 to 200 basis points, the approximation loses precision, and logarithmic methods may produce better results. Using too small a shift creates the opposite problem: rounding errors in the model prices get magnified when you divide by a tiny Δy². Twenty-five basis points is a practical middle ground for most investment-grade bonds. Whatever value you use, apply it symmetrically in both directions.
A Worked Example
Suppose you hold a callable corporate bond currently priced at $980. The pricing model says that if yields drop by 25 basis points, the price rises to $990.50. If yields rise by 25 basis points, the price falls to $968.70. Plugging in:
Effective Convexity = ($990.50 + $968.70 − 2 × $980) / ($980 × 0.0025²)
The numerator is $990.50 + $968.70 − $1,960 = −$0.80. The denominator is $980 × 0.00000625 = $0.006125. The result is roughly −130.6.
That negative value tells you the bond’s upside is capped. Yields fell by 25 basis points and the price only rose $10.50, while a 25-basis-point rise pushed the price down $11.30. The call option is compressing the gain. A non-callable bond with similar duration would show positive convexity, meaning its upside move would exceed its downside move.
Combining Duration and Convexity to Estimate Price Changes
Duration by itself understates gains when rates fall and overstates losses when rates rise for positively convex bonds. The convexity adjustment corrects the estimate:
%ΔP ≈ (−Duration × Δy) + (½ × Convexity × Δy²)
The first term is the linear duration estimate. The second term is the convexity correction. For a positively convex bond, the correction is always additive regardless of which way rates move, because Δy is squared. For a negatively convex bond, the correction works against the holder, cutting gains and deepening losses.
The practical difference is real. For a bond with moderate duration, a two-percentage-point rate increase might cause a duration-only estimate to overstate the price decline by nearly a full percentage point. Adding the convexity adjustment brings the estimate within a few cents of the actual repriced value. For smaller moves of 25 to 50 basis points the correction is smaller, but still worth including in any portfolio with meaningful rate exposure.
Positive Convexity in Standard Bonds
Most option-free bonds exhibit positive convexity. The price-yield curve bows upward: the price rises faster as yields fall and declines slower as yields rise. That asymmetry is a favorable property, and higher convexity makes it more pronounced.
Longer-maturity zero-coupon bonds tend to have the highest convexity among conventional bonds because all of the cash flow sits at one distant point, making the price extremely sensitive to the curvature of the discount function. Coupon-paying bonds spread cash flows over time, which dampens the effect. Two bonds can have identical durations and very different convexities depending on how their cash flows are structured, which is why convexity deserves its own line on a risk report.
Negative Convexity in Callable Bonds
A callable bond lets the issuer redeem the bond early at a set call price. When rates fall well below the coupon rate, the issuer has a strong incentive to call and refinance. That creates a price ceiling: buyers know the bond is likely to be pulled, so its market price won’t climb much above the call price.
The ceiling is the source of negative convexity. Below a certain yield, the price-yield curve flattens and eventually bends the wrong way. Instead of accelerating upward as rates fall further, the price stalls near the call price. Duration would predict a bigger gain than actually shows up. When rates rise, the call option is worthless and the bond behaves like an ordinary fixed-rate instrument, fully exposed to price declines. Capped upside, uncapped downside.
Most of the mispricing in this area comes from investors who rely on modified convexity for a callable bond. In a falling-rate environment they overestimate the bond’s value and allocate capital based on gains that will never materialize.
Mortgage-Backed Securities and Prepayment Risk
Mortgage-backed securities have their own convexity profile because it is driven by homeowner behavior rather than a clean contractual call. When rates drop, borrowers refinance, prepayments surge, and the investor receives principal back early and must reinvest at lower rates. When rates rise, prepayments slow to a trickle and the investor is stuck holding a below-market-rate asset for longer than expected, which is extension risk.
The Public Securities Association (PSA) prepayment model provides a standardized framework for estimating those speeds. As PSA speeds rise, effective convexity falls: a pool that shows convexity above 80 at 100 PSA may fall below 50 at 200 PSA. Faster prepayments compress the security’s upside while leaving the downside intact. MBS positions often carry deeply negative effective convexity during rate rallies, and hedging that exposure requires either options overlays or deliberate allocation to positively convex assets that offset the drag.
Putable Bonds and Price Floors
Putable bonds run in the opposite direction. The holder has the right to sell the bond back to the issuer at the put price, typically par, on specified dates. When rates rise and the market value falls, the put becomes valuable because the investor can force a sale at the put price rather than accept the depressed market value.
That creates a price floor. The bond’s price won’t fall much below the put price because the option kicks in, and the effective convexity is positive. The upside stays uncapped when rates fall, since the issuer has no right to call. The result is significantly more upside potential than downside risk.1CFA Institute. Valuation and Analysis of Bonds with Embedded Options
Putable bonds are rarer than callable bonds and usually carry lower yields, since the embedded put has value to the holder. Effective convexity is the tool for quantifying how much protection the put provides at different yield levels.
Barbell vs. Bullet Portfolios
Two portfolios can share the same duration and have very different convexity profiles. A bullet portfolio concentrates holdings around a single maturity. A barbell portfolio splits holdings between short-term and long-term bonds, with little in the middle. The barbell has higher convexity because spreading cash flows across distant points increases the curvature of the portfolio’s price-yield relationship.
For small, parallel shifts, the two portfolios perform similarly. The difference shows up during larger moves, where the barbell’s extra convexity generates better returns in both directions. The barbell also tends to outperform during yield curve flattening. The bullet holds an edge mainly when the curve steepens moderately.
Convexity matching extends the idea into liability-driven investing. Immunizing a portfolio against interest rate risk requires matching asset duration to liability duration, but that alone is not enough for large rate shifts. The asset portfolio’s convexity must equal or exceed the liability portfolio’s convexity, and the asset cash flows should be more widely dispersed in time than the liability cash flows. Meeting both conditions is what keeps large rate swings from opening a hole in the funding ratio.
Tax Consequences When a Callable Bond Gets Redeemed
Negative convexity is not just a pricing abstraction. When an issuer actually calls a bond, the tax bill arrives. The IRS treats the redemption as a sale or exchange, so any difference between the call price and your adjusted basis produces a capital gain or loss reported on Schedule D.2Internal Revenue Service. Publication 550, Investment Income and Expenses
If you bought the bond at a premium above the call price, the call crystallizes a capital loss. If you bought at a discount, you recognize a gain. Interest that accrued between the last coupon payment and the call date is taxed as ordinary interest income, separate from the capital transaction. For bonds purchased with original issue discount, your basis is adjusted upward by the OID already included in income over the holding period, which reduces or eliminates the gain at redemption.2Internal Revenue Service. Publication 550, Investment Income and Expenses
Callable bonds bought at a premium also raise a premium amortization question. If using an earlier call date produces a smaller amortizable premium than using the maturity date, the holder must amortize based on the call date. That forces a slower amortization schedule than the holder might otherwise choose, cutting the annual deduction against interest income. The rule applies even if there is no realistic chance the issuer will call on that date.