Duration: Measuring How Much a Rate Move Will Cost You
Two bonds maturing within a year of each other can lose wildly different amounts when the Federal Reserve raises rates, one shedding 2% of its value and another shedding 15%. Duration is the single number that predicts this gap in advance, and almost nobody checks it before buying.
The core mechanism: why bond prices move opposite to yields
A bond is a contract that promises a fixed stream of future cash payments: periodic coupons and a final repayment of principal (the "face value" or "par value"). Once that stream of payments is locked in at issuance, the only thing that can change is the price a buyer is willing to pay today for the right to receive it. When newly issued bonds start offering a higher yield than an existing bond's fixed coupon, that existing bond becomes less attractive by comparison, and its price must fall until its effective yield, given its fixed payments, catches up to the new market rate. This inverse relationship, price falls when yields rise and price rises when yields fall, is the single most important mechanical fact in fixed income investing, and it holds for every conventional bond regardless of issuer or credit quality.
The question that matters for a portfolio is not whether a bond's price moves opposite to yields, since that is always true, but by how much. This is where duration comes in. Duration measures a bond's price sensitivity to a change in interest rates, expressed in years, and it answers a precise question: for a small change in yield, what percentage does the bond's price move? Two forces drive duration higher. First, longer time to maturity generally means higher duration, because a payment scheduled 20 years out is more sensitive to a change in the discount rate than a payment scheduled 2 years out, in the same way a distant target is harder to hit precisely with a small aiming error. Second, for a given maturity, a lower coupon rate means higher duration, because more of the bond's total value is concentrated in the single final principal payment rather than spread across a stream of intermediate coupons that arrive, and can be reinvested, sooner. A 20-year zero-coupon bond, which pays nothing until maturity, has a duration equal to its full 20-year maturity, the maximum possible for that term. A 20-year bond carrying a generous coupon might carry a duration closer to 12 or 13 years, because a meaningful share of its value returns to the investor well before the 20-year mark.
Formally, duration is a weighted average of the times until each cash flow is received, with the weights set by the present value of each payment as a share of the bond's total present value. That definition, known as Macaulay duration, is the conceptual foundation, but the number investors actually use for risk management is modified duration, a close variant that converts the Macaulay figure into a direct estimate of percentage price sensitivity. The working approximation, and the one worth memorizing, is: percent price change ≈ negative modified duration x change in yield. A negative sign is built into the relationship because price and yield move in opposite directions; in practice, investors usually just remember that a rate increase produces a price decrease of roughly that magnitude, and vice versa.
The math: two worked examples of duration in action
Worked example 1: estimating the loss on a single bond from a rate shock. Suppose an investor holds a corporate bond with a face value of $10,000 and a modified duration of 7.5 years. The Federal Reserve raises its policy rate, and market yields for bonds of this type rise by 0.75 percentage points, or 75 basis points. Using the approximation, the expected percentage price change is -7.5 x 0.75% = -5.625%. Applied to the $10,000 position, the estimated dollar loss is $10,000 x 5.625% = $562.50. The investor still owns the same bond, still receives the same coupons, and will still get par back at maturity if the issuer does not default, but the market value of the position today has fallen by roughly $562.50, a real economic cost for anyone who needs to sell before maturity or who is marking a portfolio to market for reporting purposes.
Worked example 2: comparing two bonds with the same maturity but different durations. Consider two bonds, both maturing in 15 years. Bond A carries a 6% coupon and has a modified duration of 10.2 years. Bond B is a 15-year zero-coupon bond with a modified duration of 14.6 years, close to its full maturity since it pays nothing until the end. Suppose yields across the board rise by 1 percentage point. Bond A's estimated price change is -10.2 x 1% = -10.2%. Bond B's estimated price change is -14.6 x 1% = -14.6%. On $20,000 invested in each, Bond A loses approximately $20,000 x 10.2% = $2,040, while Bond B loses approximately $20,000 x 14.6% = $2,920, a difference of $880 between two bonds that mature on the exact same date. The gap exists entirely because of how each bond's cash flows are timed, and it is invisible to an investor who looks only at the maturity date printed on the bond.
It is also worth extending example 1 to see what happens with rates moving in the investor's favor, since the same approximation works symmetrically for a first-order estimate. If yields on that same $10,000, 7.5-duration bond fell by 0.75 percentage points instead of rising, the estimated price gain would be -7.5 x (-0.75%) = +5.625%, or roughly $562.50 of appreciation. This symmetry is precisely why long-duration bonds are often used deliberately: an investor who expects rates to fall, or who wants a hedge against a recession (a scenario in which central banks typically cut rates), can use long-duration bonds as a source of expected capital appreciation, not just a source of risk.
What market history shows about rate shocks
The mechanical relationship between duration and price is not a theoretical curiosity; it has played out repeatedly and painfully in observable market history. During the rate-hiking cycle that began in 2022, as central banks raised policy rates at one of the fastest paces in decades to combat inflation running well above target, long-duration government bond funds recorded some of their steepest price declines on record, with certain long-maturity Treasury funds falling by 25% to 30% in a single calendar year, a magnitude of loss more commonly associated with equities than with instruments widely marketed to conservative investors as "safe." The mathematics of duration explains this precisely: a long-duration Treasury fund with an average duration near 18 years, facing a roughly 1.5 to 2 percentage point rise in long-term yields over the period, produced losses squarely in line with the duration approximation, even though every underlying bond continued paying its coupons on schedule and carried effectively zero default risk.
The historical record also illustrates the limits of the linear approximation used above. Duration is a straight-line, or "first-order," estimate, and it works best for small to moderate yield changes. For larger moves, the true price change deviates from the duration estimate because the price-yield relationship is actually curved, not straight, an effect called convexity, covered in more detail elsewhere. In practice this means duration slightly overstates the loss from a large rate increase and slightly understates the gain from a large rate decrease, a modest but real asymmetry that becomes more important the larger the rate move under consideration.
Applying duration in a real portfolio
For a working professional building a portfolio around a defined savings goal, retirement in 25 years, a home purchase in 5 years, a child's education in 12, duration is the tool that lets a bond allocation be matched deliberately to that horizon rather than chosen by habit. An investor with a near-term, fixed-dollar liability, tuition due in three years, for example, is generally better served by a short-duration bond fund or an actual bond ladder timed to mature around that date, since a short-duration position is far less exposed to a rate spike disrupting the plan. An investor with a genuinely long horizon and no near-term need for the money can rationally accept more duration risk in the bond portion of a portfolio, since price fluctuations along the way matter far less than the total return earned by the time the money is actually needed.
This distinction matters most for people managing sizeable fixed income allocations inside retirement accounts, where the instinct to treat "bond fund" as synonymous with "safe" can be costly. A target-date fund or a generic aggregate bond fund often carries a duration in the 5 to 7 year range, which is moderate, but some long-Treasury or long-corporate funds marketed for yield carry durations of 15 years or more, exposing a supposedly conservative sleeve of a portfolio to double-digit percentage swings. Checking a fund's stated average duration, a figure every fund is required to disclose in its literature, takes under a minute and prevents a genuinely unpleasant surprise the next time rates move sharply.
Duration also interacts directly with how a portfolio is drawn down in retirement. An investor withdrawing a fixed dollar amount each year from a long-duration bond sleeve during a rising-rate period is effectively forced to realize losses at exactly the wrong moment, selling depreciated shares to fund living expenses rather than waiting for the eventual price recovery that duration-matched holding periods would otherwise deliver. This is one reason many retirement income plans deliberately hold several years of near-term spending needs in short-duration instruments, cash, short Treasuries, or a short-term bond ladder, insulated from a rate shock, while allowing the remainder of the fixed income allocation to carry a longer duration appropriate to the portfolio's full time horizon. The split is not arbitrary; it is duration matching applied directly to a spending plan rather than to a single lump-sum liability.
It is also worth noting that duration is dynamic, not fixed. As a bond ages and moves closer to maturity, its duration mechanically shortens, and as market yields themselves change, a bond's duration shifts slightly as well, an effect closely related to convexity. A bond fund's reported duration is therefore a snapshot, not a permanent characteristic, and an investor holding a fund for many years should expect its duration to drift as the underlying portfolio turns over, particularly during periods when the fund manager is actively repositioning in anticipation of a changing rate environment.
Actionable breakdown
- Understanding your exposure
- Look up a bond fund's average duration before buying.
- Multiply duration by an expected rate change for a rough loss estimate.
- Compare durations, not just maturities, across similar bonds.
- Managing the risk deliberately
- Shorten duration when you expect rates to rise soon.
- Extend duration when you want a recession hedge.
- Ladder maturities to smooth out duration exposure over time.
- Matching duration to your life
- Match bond duration to a known, dated future expense.
- Keep near-term cash needs in short-duration instruments only.
- Accept longer duration only with a genuinely long horizon.
Common pitfalls
Confusing duration with maturity: a 15-year coupon bond and a 15-year zero-coupon bond can differ in duration by four years or more, so the maturity date printed on a bond tells you little about its actual rate sensitivity.
Treating the duration approximation as exact for large moves: it is a linear estimate; for big rate swings, the true price change bends away from the straight-line prediction, an effect covered separately under convexity.
Assuming all bond funds are low risk: a long-duration fund can lose more in a single year than many equity funds lose in a mild correction, and the fund's name alone rarely signals this.
Ignoring duration drift inside a fund: a bond fund's average duration changes as its holdings mature and are replaced, so a fund that was short-duration two years ago is not guaranteed to be short-duration today.
The bottom line
Duration converts an abstract interest rate forecast into a concrete number: check it before you buy, because it is a far better predictor of a bond's real-world risk than its maturity date or its credit rating.
All articles · Convexity · Passive bond management · Bonds fundamentals · Duration (glossary)