Definition
Duration estimates a bond or bond fund's price sensitivity to a change in interest rates. An ETF with a duration of 7 years would be expected to lose roughly 7% if comparable yields rose by one percentage point, or gain roughly 7% if yields fell by one point. That is an estimate, not a promise, because large rate moves and changing credit spreads make the relationship nonlinear.
Duration is not the same as maturity. Maturity says when principal is due; duration also accounts for when coupons arrive. Higher coupons return cash sooner and usually reduce duration.
Bond funds quote effective duration for the whole portfolio. It is a weighted blend of the holdings, updated as bonds roll off and new ones are bought. A target-maturity fund's duration drifts down as the end date approaches; an open-ended fund such as BND keeps duration in a stated range by replacing maturing bonds.
Why It Matters
A bond ETF can have a modest yield and still carry substantial price risk. Duration lets investors compare that risk before assuming every bond fund is a stable cash substitute. It also helps match a holding to a time horizon: money needed soon generally should not depend on long-duration prices.
A T-bill fund such as SGOV has duration measured in *months*. Its price barely moves when the 10-year yield jumps. A long Treasury fund such as TLT can have duration in the high teens. The same one-point rate rise that is a rounding error on SGOV can be a double-digit drawdown on TLT.
Income investors feel this when they buy "the highest SEC yield" in the bond sleeve. Longer duration often *is* the extra yield. That is a trade, not a free upgrade. See bond basics for the coupon-versus-price picture.
Example
The rates below are illustrative, not live quotes.
Suppose a bond ETF has an effective duration of 6.5 years. If market yields rise from 4% to 4.75%, the duration estimate is -6.5 × 0.75%, or about a 4.9% price decline. Income continues to arrive, and the fund gradually reinvests at higher yields, but the immediate price effect matters.
If yields instead fall 0.75 points, the same estimate is about a 4.9% price gain. That symmetry is a first-order approximation. Convexity makes large rallies a bit friendlier and large selloffs a bit worse than the straight-line guess.
A cash-like sleeve with 0.3 years of duration would be expected to move about 0.2% on that same 0.75-point shift. The yield gap versus the 6.5-year fund is the price of that stability.
How to Use Duration
- Match the money's job. Next year's tuition does not belong in a 16-year duration fund, even if the SEC yield is higher.
- Compare duration before comparing yield. Two "intermediate" bond ETFs can differ by years of duration.
- Add credit separately. Duration does not tell you what happens if spreads blow out. A high-yield fund can fall when Treasuries rally.
- Recheck after large rate moves. Effective duration itself changes as yields change and as the portfolio turns over.
Building a Bond Sleeve
A simple three-bucket sleeve uses duration on purpose:
- Cash-like: SGOV (months of duration) for money you may need this year.
- Core: BND (intermediate duration) for ballast that still has rate sensitivity.
- Long: TLT only if you *want* a large move when long rates fall — not as a "safer" version of BND.
Write the job next to the duration number. If the job is "next year's property tax," TLT is the wrong tool even when its SEC yield is higher. If the job is "I want duration as a hedge if growth and yields fall," TLT is a satellite with a loss budget, not a cash account.
Add credit separately. Duration does not protect a high-yield sleeve when spreads blow out. See credit risk.
Common Mistakes
- Treating maturity and duration as interchangeable.
- Assuming a bond ETF cannot lose money because its holdings eventually mature.
- Comparing yields without comparing duration and credit quality.
- Using the estimate as an exact forecast for a large rate move.
- Buying a long-duration fund for "safety" because bonds feel safer than stocks.
FAQ
Is lower duration always better?
No. Lower duration reduces rate sensitivity, but longer duration can provide more upside when rates fall and may better match a long-dated liability.
Does duration include credit risk?
Not fully. Duration isolates rate sensitivity; changing credit spreads can move corporate bonds in addition to the Treasury-rate effect.
Why didn't my fund move exactly duration times the yield change?
Duration is a local, linear estimate. Large moves, curve twists (short rates vs long rates), and credit-spread changes all push the result off the simple line.
Can I mix SGOV and TLT and call it intermediate?
Only if you measure the blend. A 50/50 mix of 0.3-year and 16-year duration is about 8 years — closer to a long-intermediate fund than to cash. Weights matter more than the average of the ticker names.
Does a target-maturity ETF's duration fall over time?
Yes. As the end date approaches, duration rolls down. That is useful when you need a known payoff year. An open-ended fund such as BND does not roll down; it buys new bonds and keeps duration in range.
