Skip to content
QuantCalc Tools
Directory

Bond Duration Convexity Math

DV01-Neutral Barbell vs Bullet Portfolio: Convexity Pickup vs Yield Math

Matching market value and Modified Duration between a 2Y/30Y Treasury barbell and a 10Y bullet boosts portfolio convexity by +114% while introducing calculable yield-give-up and 2s30s twist exposure.

By Elena Rostova, FRM, CQFReview not recorded
DV01-Neutral Barbell vs Bullet Portfolio: Convexity Pickup vs Yield Math — QuantCalc Tools 2026 quantitative & supervisory data card
QuantCalc Tools (Modified Duration, Convexity, DV01 & Key-Rate Hedging): Cash- and DV01-neutral 2Y/30Y weights, +114% convexity pickup, and 2s30s twist exposure.Source: QuantCalc Tools Institutional Research DeskLicense: CC-BY-4.0

Portfolio managers seeking positive tail convexity without altering aggregate interest-rate beta swap a concentrated intermediate Bullet (100% in 10-year notes) into a cash-and-duration-matched Barbell combining short 2-year notes and ultra-long 30-year bonds. Why does a barbell always carry higher Macaulay Duration, Modified Duration & Second-Order Convexity than an equal-duration bullet? Because Modified Duration is a linear weighted average of maturities whereas Convexity grows quadratically ($\propto t^2$): by Jensen's inequality, $wS DS^2 + wL DL^2 > (wS DS + wL DL)^2$! Test funding carry differentials in our SOFR overnight compounding & carry calculator and ground this strategy in our Modified Duration, Convexity & DV01 master guide.

Solving Exact Cash-Neutral & DV01-Neutral Barbell Weights#

Given target Bullet market value $VB = \$100\text{M}$ and Modified Duration $DB$, we allocate market-value weights $w2$ and $w{30}$ across the 2-year wing ($D2, C2$) and 30-year wing ($D{30}, C{30}$) by solving the $2 \times 2$ linear system: $$w2 + w{30} = 1, \qquad w2 D2 + w{30} D{30} = DB \implies w{30} = \frac{DB - D2}{D{30} - D2}, \quad w2 = 1 - w{30}$$ Once $w2$ and $w{30}$ are locked in, portfolio convexity aggregates linearly by market-value weight: $C{\text{barbell}} = w2 C2 + w{30} C{30}$. On a normal concave upward-sloping yield curve, however, linear yield interpolation lies below the curved 10-year belly yield, creating a negative static carry spread ($\Delta y{\text{giveup}} = y{\text{barbell}} - yB < 0$). Solving $\frac{1}{2}(C{\text{barbell}} - CB)(\Delta y{BE})^2 = |\Delta y{\text{giveup}}|$ gives the exact breakeven parallel yield move required over 1 year for the barbell to outperform!

Worked $100M 2Y/30Y Barbell Weighting & Breakeven Volatility Audit#

Formula or model from the articleNot independently verified
Worked $100,000,000 DV01-Neutral 2Y/30Y Barbell vs 10Y Bullet Trade Audit
  • Market Benchmark Inputs (Par Clean/Dirty = 100.00):
  • 2Y Treasury Wing: YTM y2 = 4.10%, Dmod = 1.890 yrs, Convexity C_2 = 4.50 yrs^2
  • 10Y Bullet Target: YTM y10 = 4.40%, Dmod = 7.820 yrs, Convexity C_10 = 73.40 yrs^2
  • 30Y Treasury Wing: YTM y30 = 4.65%, Dmod = 16.950 yrs, Convexity C_30 = 392.00 yrs^2
  • 1) Exact Market-Value & DV01-Neutral Barbell Weights:
  • w_30 = (7.820 - 1.890) / (16.950 - 1.890) = 5.930 / 15.060 = 39.3758% ($39,375,830 in 30Y)
  • w_2 = 1 - 0.393758 = 60.6242% ($60,624,170 in 2Y)
  • Check Barbell Duration: 0.606242(1.890) + 0.393758(16.950) = 1.1458 + 6.6742 = 7.8200 yrs (EXACT!)
  • Check Portfolio DV01: Both portfolios = $78,200 / bp!
  • 2) Net Portfolio Convexity Pickup (Delta_C):
  • C_barbell = 0.606242 * (4.50) + 0.393758 * (392.00) = 2.728 + 154.353 = 157.081 yrs^2
  • Net Convexity Gain Delta_C = 157.081 - 73.400 = +83.681 yrs^2 (+114.01% MORE CONVEXITY!)
  • 3) Static Yield Give-Up & 1-Year Breakeven Parallel Yield Shift (DeltayBE):
  • Market-Value Weighted Yield y_barbell = 0.606242(4.10%) + 0.393758(4.65%) = 4.3166%
  • 1-Year Static Yield Give-Up vs 10Y Bullet (4.40%) = 4.3166% - 4.4000% = -8.34 bps (-0.000834)
  • Breakeven Parallel Shift: 0.5 * 83.681 * (DeltayBE)^2 = 0.000834
  • DeltayBE = sqrt( 0.001668 / 83.681 ) = sqrt(0.000019933) = +/- 0.004465 = +/- 44.65 bps!

Look at Step 3: the 2Y/30Y barbell gives up $-8.34\text{ bps}$ of static yield if the Treasury curve freezes completely unchanged for 12 months, yet outperforms the 10Y bullet whenever parallel yields move more than $\pm 44.65\text{ bps}$ in EITHER direction! At a $\pm 120\text{ bps}$ parallel shock, the barbell beats the bullet by $+0.5 \times 83.681 \times (0.012)^2 - 0.000834 = +60.25 - 8.34 = \mathbf{+51.91\text{ bps}}$ ($+\$519,100$ pure convexity alpha)!

DV01-neutral barbell vs bullet portfolio construction and curve-twist stress-testing workflow
DV01-neutral barbell vs bullet portfolio construction and curve-twist stress-testing workflowVerified Exhibit

The Hidden Curve Trap: Why a Cash-Neutral Barbell Is an Implicit 2s30s Flattener!#

Look closely at the fourth KPI card above (30Y WING DV01 SHARE = 85.35%)! Even though 60.62% of cash sits in the 2-year note and only 39.38% of cash sits in the 30-year bond, 85.35% of the barbell's DV01 ($0.39376 \times 16.95 / 7.82 = 85.35\%$, or $$66,742/\text{bp}$ out of $$78,200/\text{bp}$) sits out at the 30-year maturity! Check Key Rate Duration (KRD) & Partial DV01 Yield Curve Bucketing across non-parallel curve twists:

  • Bear Flattener / Bull Flattener (2Y yields rise relative to 30Y yields): The barbell CRUSHES the 10Y bullet because $85.4\%$ of its DV01 sits in the outperforming 30Y wing!
  • Bear Steepener (30Y yields jump $+40\text{ bps}$ while 2Y stays anchored): Even if 10Y yields rise $+20\text{ bps}$, the barbell loses $-\$66.74\text{k} \times 40 = -\$2.67\text{M}$ while the 10Y bullet loses only $-\$78.20\text{k} \times 20 = -\$1.56\text{M}$—wiping out 13 years of convexity pickup in one afternoon!

Barbell Minus Bullet Active Return Across Parallel & Twist Scenarios#

Quantitative Distribution · Barbell Minus Bullet Excess Return
Barbell Minus Bullet Excess Return
01Bear Flattener (FOMC Hiking…$+154.80\text{ …02Bear Steepener (Fiscal Defi…$-176.40\text{ …
Curve Scenario (1-Year Horizon, $100M Book)$\Delta y_{2Y}$$\Delta y_{10Y}$$\Delta y_{30Y}$Barbell Minus Bullet Excess Return
Static Unchanged Curve$0\text{ bps}$$0\text{ bps}$$0\text{ bps}$$-8.34\text{ bps}$ (Carry give-up; excludes rolldown)
Parallel Bull Rally ($-120\text{ bps}$)$-120\text{ bps}$$-120\text{ bps}$$-120\text{ bps}$$+51.91\text{ bps}$ (Convexity surge wins)
Parallel Bear Selloff ($+120\text{ bps}$)$+120\text{ bps}$$+120\text{ bps}$$+120\text{ bps}$$+51.91\text{ bps}$ (Convexity cushion wins)
Bear Flattener (FOMC Hiking Cycle)$+100\text{ bps}$$+55\text{ bps}$$+25\text{ bps}$$+154.80\text{ bps}$ (Long 30Y wing wins big!)
Bear Steepener (Fiscal Deficit Shock)$+10\text{ bps}$$+45\text{ bps}$$+80\text{ bps}$$-176.40\text{ bps}$ (85.4% 30Y DV01 punished!)
Barbell minus Bullet active performance attribution across unchanged, +/-120bp parallel, flattener, and steepener scenarios
Barbell minus Bullet active performance attribution across unchanged, +/-120bp parallel, flattener, and steepener scenariosVerified Exhibit

Upgrading to Regression / PCA-Neutral Barbells and Rolldown Accounting#

Also remember rolldown: a 10-year bullet rolling down to 9 years on a steep 7s10s slope often earns $+12\text{ to }+18\text{ bps}$ of annual rolldown price gain, whereas a 30-year bond rolling to 29 years on a flat 25s30s tail earns near zero—see leveraged Treasury repo carry and rolldown math! To isolate pure curvature without unintended 2s30s slope bets, weight wings via 50/50 vs regression-weighted 2s5s10s butterflies and Redington liability dispersion rules, and explore the Modified Duration & Key-Rate Hedging Topic Hub.

Frequently asked questions

Why does a barbell always have higher convexity than a bullet with the exact same Modified Duration?

Modified Duration weights cashflow maturities linearly (t), whereas Convexity weights cashflow maturities quadratically (t*(t + 1/k)). Spreading cashflows to 2Y and 30Y adds variance (cashflow dispersion M^2) on top of squared mean duration, boosting convexity.

Why does a DV01-neutral 2Y/30Y barbell lose money against a 10Y bullet during a yield-curve steepening?

Because the 30Y bond has nearly 9x the duration of the 2Y note, roughly 85% of the barbell’s dollar DV01 resides in the 30Y sector and only 15% in the 2Y sector. If 30Y yields rise more than 10Y yields, the heavy 30Y DV01 drag overwhelms 2Y gains.

Can you construct a barbell that is BOTH cash-neutral, parallel-DV01-neutral, AND slope-neutral?

Not with only two instruments against one bullet (2 equations fix 2 weights). Adding a third maturity (creating a 3-leg butterfly or 2Y/5Y/30Y structure) or allowing repo leverage provides the third degree of freedom needed to neutralize PCA slope!

When does a barbell earn HIGHER static yield than a bullet AND higher convexity simultaneously?

Whenever the yield curve exhibits positive butterfly curvature (or humped/inverted wings where w_2*y_2 + w_30*y_30 > y_10), the barbell delivers both positive static carry pickup AND +114% higher convexity—a classic relative-value arbitrage signal!

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). DV01-Neutral Barbell vs Bullet Portfolio: Convexity Pickup vs Yield Math. QuantCalc Tools. https://quantcalctools.com/guides/dv01-neutral-barbell-vs-bullet-portfolio-convexity-pickup-math
Explore Bond Duration Convexity Math Hub (20) →

QuantCalc Tools Data, Benchmarks & Contributor Records

DV01-Neutral Barbell vs Bullet Convexity Formula Guide | QuantCalc Tools