Skip to content
QuantCalc Tools
Directory

Bond Duration Convexity Math

Macaulay vs Modified vs Effective Duration: Semiannual & ICMA Yield Math

Quoting a 4.50% coupon bond on annual ICMA 251 European conventions versus semiannual US Street conventions shifts Modified Duration by 17 bps even when Macaulay cashflow timing matches.

By Elena Rostova, FRM, CQFReview not recorded
Macaulay vs Modified vs Effective Duration: Semiannual & ICMA Yield Math — QuantCalc Tools 2026 quantitative & supervisory data card
QuantCalc Tools (Modified Duration, Convexity, DV01 & Key-Rate Hedging): Semiannual US vs annual ICMA 251 compounding, Fisher-Weil curves, and OAS cliffs.Source: QuantCalc Tools Institutional Research DeskLicense: CC-BY-4.0

Cross-market bond desks comparing a 10-year US Treasury ($k = 2$ semiannual Street yield) against a 10-year German Bund or Eurobond ($k = 1$ annual ICMA Rule 251 yield) routinely encounter a subtle $16\text{ to }22\text{ bps}$ duration discrepancy: because Modified Duration divides Macaulay Duration by $(1 + yk / k)$, changing yield compounding frequency $k$ alters $D{mod}$ even on identical cashflows! Compare continuous vs discrete compounding mechanics alongside our SOFR overnight compounding & carry calculator and ground this taxonomy in our Modified Duration, Convexity & DV01 master guide.

Exact Discrete vs Continuous Compounding Identities ($D{Mac}$ vs $D{mod}$)#

Frederick Macaulay (1938) introduced duration to measure the present-value-weighted average maturity (in years) of $N$ fixed cashflows $CFi$ arriving at times $ti$ (in years), discounted at $k$-times-per-year compounded yield $yk$: $$D{Mac} = \frac{1}{P}\sum{i=1}^{N} ti \cdot \frac{CFi}{(1 + yk / k)^{k ti}}$$ Now differentiate $P(yk)$ with respect to $yk$: $$\frac{\partial P}{\partial yk} = -\sum{i=1}^N ti \cdot \frac{CFi}{(1 + yk / k)^{k ti + 1}} = -\frac{P \cdot D{Mac}}{1 + yk / k} \implies D{mod}(k) = \frac{D{Mac}}{1 + yk / k}$$ Look what happens as $k \to \infty$ (continuously compounded yield $yc = k\ln(1 + yk/k)$, where $e^{-yc ti} = (1 + yk/k)^{-k ti}$): under continuous compounding, Modified Duration equals Macaulay Duration identically ($D{mod}(\infty) \equiv D{Mac}$)! Furthermore, for a par coupon bond ($P = 100$, coupon rate $c = yk$ paid $k$ times per year for $M$ years), closed-form Macaulay Duration simplifies without summing $2M$ rows to: $$D{Mac}^{\text{par}} = \frac{1 + yk / k}{yk}\left[1 - \frac{1}{(1 + yk / k)^{kM}}\right], \qquad D{mod}^{\text{par}} = \frac{1}{yk}\left[1 - \frac{1}{(1 + yk / k)^{kM}}\right]$$

Worked US Semiannual (k=2) vs European ICMA Annual (k=1) Duration Audit#

Formula or model from the articleNot independently verified
Worked Cross-Convention Audit: 10-Year 4.50% Par Sovereign Bond (Semiannual US vs Annual ICMA)
  • Case A: Semiannual US Treasury Convention (k = 2, C/2 = 2.25, y_2 = 4.50%, 20 periods, Par P = 100.00)
  • Discount Factor (1 + 0.0225)^(-20) = 0.6408165
  • Closed-Form Macaulay Duration:
  • D_Mac(k=2) = (1.0225 / 0.0450) * (1 - 0.6408165) = 22.72222 * 0.3591835 = 8.16145 years
  • Semiannual Modified Duration:
  • D_mod(k=2) = 8.16145 / 1.0225 = 7.98186 years ($79,819 DV01 per $100M par)
  • Case B: Same Semiannual Bond Quoted on Equivalent Annual ICMA Rule 251 Yield!
  • Equivalent Effective Annual Yield y_1 = (1 + 0.0225)^2 - 1 = 4.550625% (k = 1)
  • Note: Present value of every cashflow is IDENTICAL, so Macaulay Duration is UNCHANGED:
  • D_Mac(k=1) = 8.16145 years!
  • Annual-Quoted Modified Duration (Sensitivity to 1 bp of y1, NOT y2!):
  • D_mod(k=1) = 8.16145 / (1 + 0.04550625) = 7.80622 years (-0.1756 years / -17.56 bps lower!)
  • Reconciliation Check: dy1 / dy2 = (1 + y_2/2) = 1.0225 => 7.80622 * 1.0225 = 7.98186 (EXACT!)

That reconciliation check saves cross-currency sovereign desks millions in hedging errors: a 1 bp move in annual yield $y1$ is physically smaller than a 1 bp move in semiannual yield $y2$ by $(1 + y_2/2)^{-1}$, so never compare $D_{mod}$ of a Bund ($k=1$) against a US Treasury or UK Gilt ($k=2$) without converting both to a common compounding basis!

Decision tree distinguishing Macaulay, Modified, Fisher-Weil zero-curve, and OAS Effective Duration
Decision tree distinguishing Macaulay, Modified, Fisher-Weil zero-curve, and OAS Effective DurationVerified Exhibit

Fisher-Weil Spot-Curve Duration and Option-Adjusted Effective Duration#

Two classic mathematical flaws limit textbook Macaulay and Modified Duration:

  1. Flat-YTM Fallacy (Solved by Fisher-Weil 1971): Macaulay discounts a 6-month coupon and a 10-year principal repayment at the exact same flat yield-to-maturity $y$! On a steeply upward-sloping or inverted yield curve, Fisher-Weil Duration replaces $(1 + y/k)^{-kti}$ with stripped zero-coupon spot rates $z(ti)$: $D{FW} = \frac{1}{P}\sum ti CFi e^{-z(ti)t_i}$, measuring exact sensitivity to a parallel shift in the continuous zero curve.
  2. Fixed-Cashflow Fallacy (Solved by Effective / OAS Duration): On a 5-year corporate bond callable at $101.00$ in Year 2, a $-50\text{ bps}$ rally increases the probability the issuer calls the bond early, shortening expected cashflow life! Under the MACAULAY-CONVEXITY 2026 Standard, risk engines hold the Option-Adjusted Spread (OAS) constant inside a Hull-White or Black-Karasinski interest-rate tree, shift the benchmark SOFR zero curve by $\pm\Delta y$ (typically $15\text{ to }25\text{ bps}$), and evaluate central finite differences:

$$D{eff} = \frac{P{-\Delta y} - P{+\Delta y}}{2\,P0\,\Delta y}$$

Institutional Bond Duration Formulations & Embedded-Option Comparison#

Quantitative Distribution · Units
Units
01Modified Duration ($D_{mod}…% $\Delta P$ pe…02Effective Duration ($D_{eff…% $\Delta P$ pe…
Duration FormulationDiscounting Rate UsedHandles Embedded Calls/Puts?UnitsPrimary Institutional Use Case
Macaulay Duration ($D_{Mac}$)Flat internal YTM ($y_k$)NoYearsHorizon immunization matching
Modified Duration ($D_{mod}$)Flat internal YTM ($y_k$)No% $\Delta P$ per 100% $\Delta y$Liquid bullet sovereigns & futures
Fisher-Weil Duration ($D_{FW}$)Stripped zero-coupon curve $z(t)$NoYears / % $\Delta P$Steep/inverted curve strip portfolios
Effective Duration ($D_{eff}$)OAS-constant binomial/MC treeYes (Calls, Puts, MBS)% $\Delta P$ per 100% $\Delta r$Callable corporates & MBS
Partial Key-Rate DurationBumped node $z(t_j)$ tentYes (Per-tenor OAS)% $\Delta P$ per nodeKRD curve bucketing
Divergence between Yield-to-Worst Modified Duration and OAS Effective Duration on a callable bond near its call strike
Divergence between Yield-to-Worst Modified Duration and OAS Effective Duration on a callable bond near its call strikeVerified Exhibit

Why Yield-to-Worst (YTW) Duration Creates Dangerous Hedge Cliffs#

Never hedge a callable corporate bond using Duration-to-Worst (switching abruptly from 10-year Duration-to-Maturity of $7.65$ years to 3-year Duration-to-Call of $2.65$ years the instant clean price crosses $100.0001$)! That discontinuous $5.00$-year cliff forces whip-saw futures trading on every $1\text{ bp}$ tick around par, whereas OAS Effective Duration transitions smoothly across $4.85$ years as option delta approaches $-0.50$. Pair exact duration conventions with DV01 clean vs dirty price scaling, check long-horizon compounding in our CAGR calculator, and browse the Modified Duration & Key-Rate Hedging Topic Hub.

Frequently asked questions

Why is a German Bund’s Modified Duration slightly lower than an identical-cashflow US Treasury’s?

European sovereign bonds follow ICMA Rule 251 annual compounding (k = 1), dividing Macaulay Duration by (1 + y_1), whereas US Treasuries use semiannual Street convention (k = 2), dividing by (1 + y_2 / 2). One basis point of annual yield represents a slightly smaller shock.

Can Modified Duration ever exceed a bond’s remaining years to maturity?

No for positive-yield option-free fixed-rate bonds: Macaulay Duration is bounded at most by final maturity M (achieved only by a zero-coupon bond), and dividing by (1 + y/k) > 1 makes Modified Duration strictly smaller than M unless yields are deeply negative!

Wait—what happens to Modified Duration when sovereign bond yields turn negative?

When y < 0 (such as German Bunds or Swiss Confederations during 2019-2021), the divisor (1 + y/k) is less than 1.00, making Modified Duration LARGER than Macaulay Duration—and exceeding final maturity M on a negative-yield zero-coupon bond!

Why does Fisher-Weil duration differ from Macaulay duration on steep yield curves?

Macaulay discounts near-term coupons and distant principal at the same single internal YTM. On a steep +200 bps upward-sloping curve, Fisher-Weil discounts early coupons at lower short spot rates and final principal at higher long spot rates, shortening true duration.

Publication record

Published
14 November 2026
Updated
14 November 2026
Author
Elena Rostova, FRM, CQF
Review
Review not recorded
Automated fact check
Passed in the stored pipeline record
Source URLs
3 recorded

AI assistance recorded: anthropic/claude-opus-4.6. This field does not establish human review or source verification.

Academic & Institutional Attribution

Cite This Research & Embed Live Data Card

Open Attribution License (CC BY 4.0)

Standard academic & institutional research paper citation format.

Elena Rostova, FRM, CQF. (2026, November 14). Macaulay vs Modified vs Effective Duration: Semiannual & ICMA Yield Math. QuantCalc Tools. https://quantcalctools.com/guides/macaulay-vs-modified-vs-effective-duration-yield-compounding-math
Explore Bond Duration Convexity Math Hub (20) →

QuantCalc Tools Data, Benchmarks & Contributor Records

Macaulay vs Modified vs Effective Duration Formula Guide | QuantCalc Tools