Guide

Treasury bonds (T-bonds) explained

Harbor Capital's government sleeve looked balanced on paper: T-bills for liquidity, a note ladder for intermediate duration, and TIPS for inflation. When the 30-year Treasury yield fell from 4.8% to 3.6% over eighteen months, the portfolio gained — but pension actuaries reported liabilities still rose faster. The mismatch: Harbor's effective duration topped out near 6 years while liability duration was 14. The CIO had intermediate bonds, not long bonds. Adding a dedicated 20–30 year Treasury bond sleeve closed half the funding gap in one rebalancing move.

Treasury bonds (T-bonds) are U.S. government securities with original maturities of 20 or 30 years. Like notes, they pay fixed semiannual coupons; unlike notes, they carry the highest nominal duration and convexity on the government curve — amplified price swings when long yields move. They anchor pension discount rates, mortgage duration hedges, and the global “long bond” narrative. This guide covers T-bond mechanics, the long end of the yield curve, term premium, liability matching, Harbor Capital's long-duration refactor, a technique decision table, pitfalls, and a checklist. Pair with duration and rate risk for the math behind drawdowns.

T-bond maturity taxonomy

The Treasury issues two long tenors. Both trade in a deep secondary market, settle through Fedwire, and carry the same full faith and credit guarantee as bills and notes. The practical difference is how much rate sensitivity you buy per dollar of face value.

Tenor	Typical role	Modified duration (approx.)	Who holds it
20-year	Long intermediate; less volatile than 30s	~13–15 years at par	Pensions, insurers, barbell portfolios
30-year	Maximum nominal government duration	~18–20 years at par	LDI programs, duration hedges, macro funds

T-bonds differ from STRIPS and zero-coupon bonds, which pay no coupons and maximize duration for a given maturity. They differ from TIPS, which adjust principal for CPI. T-bonds are nominal: if inflation surprises to the upside, real purchasing power erodes even when prices rally on falling yields.

Notes vs bonds: where the curve breaks

The 10-year note is often called “the long bond” on financial TV, but duration math says otherwise. A 30-year bond with a 4.5% coupon near par has roughly double the rate sensitivity of a 10-year note. Investors moving from notes to bonds are not merely extending maturity by a few years — they are stepping into a different risk regime where convexity and term premium dominate returns.

Coupon mechanics, convexity, and price behavior

New 30-year auctions set a coupon near the clearing yield. Semiannual payments follow until maturity; principal returns at par unless the U.S. defaults, which markets price as near-zero probability. When yields fall after issuance, price appreciation on long bonds typically exceeds what duration alone predicts — that extra lift is positive convexity. When yields spike, losses accelerate but convexity cushions the downside relative to a linear duration estimate.

Example intuition: a +1% parallel yield rise might mark a 30-year bond down roughly 15–18% before coupons; a −1% move might gain 18–22%. The asymmetry matters for pensions betting on falling discount rates and for retirees who cannot rebalance into cheaper bonds after a selloff.

Reinvestment and income stability

Long bonds pay higher coupons than shorter Treasuries when the curve is upward-sloping, but most total return still comes from price moves at the long end. Coupon reinvestment at uncertain future rates adds a second variable. Hold-to-maturity investors ignore mark-to-market; traders and ETF holders live in mark-to-market land daily.

Term premium and the long end of the curve

Long bond yields decompose into expected short-rate paths plus a term premium — compensation for holding duration risk. When the premium is high, 30-year yields offer extra income over rolling T-bills; when it compresses toward zero or negative, long bonds rally even if the Fed is on hold. Harbor Capital monitors the ACM-style term premium estimate quarterly: add long bonds when premium is above its 20-year median, trim when premium is deeply negative and duration already exceeds liability targets.

The long end also reflects supply dynamics. Heavy Treasury issuance at 20 and 30 years can steepen the curve; pension and foreign official buying can offset. Unlike corporates, there is no credit spread — pure rate and liquidity risk. During flight-to-quality episodes, long Treasuries often outperform equities dramatically, but 2022 showed they can sell off alongside stocks when inflation shocks dominate.

Liability-driven investing and who needs T-bonds

Liability-driven investing (LDI) matches asset duration to promised payments decades out. Defined-benefit pensions, life insurers, and some endowments with perpetual horizons are natural holders. Retail investors rarely have 30-year spending certainty; a retirement portfolio drawing in five years usually should not load on 30-year bonds unless it is an explicit rate bet with risk capital.

Pension de-risking — buy long bonds when funded status improves to lock discount-rate gains.
Convexity hedge — pair with equities or credit when seeking crisis ballast (correlation is not guaranteed).
Barbell — T-bills plus 30-year bonds; avoids the middle of the curve while keeping flexibility.
Duration overlay — futures or ETFs to synthetically extend a note portfolio to liability duration.

See liability-driven investing for pension-style frameworks and bond ladders for simpler staggered-maturity approaches that may stop at 10-year notes.

How to buy Treasury bonds

Channel	Best for	Watch-outs
TreasuryDirect	Hold-to-maturity; no ongoing fees	Illiquid before maturity; long commitment
Brokerage secondary market	Specific 20y/30y issues; ladder construction	Markups on odd lots; accrued interest at settlement
Long Treasury ETF (TLT-style)	One-ticker duration exposure; easy rebalance	Duration drifts down as bonds age; not constant 30y
Extended-duration ETF (EDV-style)	STRIPS-like duration without buying zeros	Higher volatility; understand underlying holdings
Treasury futures	Institutional overlays and hedges	Margin, roll, and basis risk

Auction calendars publish months ahead. Noncompetitive bids suit buy-and-hold investors; competitive bids need yield limits. On-the-run 30-year bonds define the benchmark long yield quoted in macro headlines.

Harbor Capital long-duration sleeve refactor

After the note ladder upgrade, Harbor still faced a 8-year duration gap versus pension liabilities. The refactor added a long bond sleeve:

Target 25% of government assets in 20–30 year bonds when term premium is above median; 15% when premium is compressed.
Split 30s and 20s 70/30 to avoid concentrating entirely on the most volatile point.
Rebalance on funded-status triggers — add long bonds after equity rallies lift funded ratio; trim after rate rallies reduce discount rates and close the gap.
Keep T-bill buffer unchanged — long bonds are not emergency liquidity.

The first year delivered +12% on the long sleeve as yields fell, offsetting underperformance in short cash. The CIO noted the trade is path-dependent: a repeat of 2022-style inflation would have produced double-digit drawdowns. Documentation now requires a +100 bps rate-shock stress test before each increase to the long bond weight.

Technique decision table

Approach	Best when	Weak when
30-year T-bonds	Long liabilities; term premium attractive; deflation or rate-cut bet	Spending within 10 years; cannot tolerate 15%+ drawdowns
20-year T-bonds	Need long duration with slightly less volatility than 30s	Still too much rate risk for near-term goals
10-year T-notes	Intermediate horizon; mortgage-rate sensitivity	Cannot match 20+ year liabilities
STRIPS / zero-coupon	Maximum duration per dollar; dated cash flows	Phantom income tax complexity; extreme volatility
TIPS (long)	Inflation-linked liabilities; real return targeting	Negative real yields; deflation floor limits rally
Long Treasury ETF	Simple implementation; daily liquidity	Duration drift; fees; not a maturity match

Common pitfalls

Treating long bonds as “safe cash plus yield” — drawdowns can exceed equity corrections in bad rate years.
Confusing the 10-year with true long duration — notes are not bonds for LDI purposes.
Ignoring convexity in stress tests — linear duration underestimates gains in rallies and can understate losses in sharp selloffs depending on curve shape.
Buying TLT for a 5-year goal — instrument mismatch guarantees regret if rates rise before spending.
Forgetting ETF duration decay — funds roll down the curve; rebalance or accept shrinking rate exposure.
Chasing last year's rally — low yields embed asymmetric risk if inflation re-accelerates.
Taxable accounts without planning — coupons taxed annually; selling at a loss may still leave net pain after years of income tax.
Assuming negative stock-bond correlation — regimes change; long bonds are not a free hedge.

Portfolio checklist

Measure liability duration (or spending horizon) before sizing long bonds.
Run +100 bps and +200 bps parallel shock on proposed holdings.
Check term premium vs historical median before overweighting the long end.
Separate liquidity sleeve (bills) from long bond strategic allocation.
Choose individual bonds vs ETF based on lot size and rebalance discipline.
Read ETF effective duration and average maturity monthly.
Document hold-to-maturity lots vs trading sleeve.
Pair nominal long bonds with TIPS if inflation risk is material.
Rebalance when funded status or duration gap moves more than 1 year.
Compare 20y vs 30y on yield pickup and volatility, not yield alone.
Account for accrued interest on secondary purchases.
Review after major supply announcements or Treasury refunding changes.

Key takeaways

T-bonds are the long-duration government anchor — 20 and 30 years, not the 10-year note.
Convexity amplifies both rallies and selloffs relative to intermediate notes.
Match bonds to liabilities — retail savers rarely need full 30-year exposure.
Term premium tells you when the long end is paying for risk.
ETFs simplify access but drift down the curve over time.