Editorial note: this article is educational content explaining a general market mechanism. It is not investment advice and does not describe any specific current event, company, or security.
Two bonds can carry the identical 5% yield and still behave like entirely different instruments the moment interest rates move. One might lose 2% of its value, the other 18%. The yield printed on the label reveals none of this. The number that does is duration, a measure that sounds technical but answers a very practical question: how much pain, or gain, does this bond deliver for a given shift in interest rates.
Duration is a measure of time and sensitivity at once
Duration was originally conceived, by economist Frederick Macaulay in the 1930s, as a way to express a bond’s average maturity in a more meaningful way than its final payoff date. Rather than simply counting the years until a bond repays its face value, Macaulay duration weighs each cash flow, every coupon payment along the way plus the final principal, by how much of the bond’s total value it represents and by how far in the future it arrives. A bond that pays no coupons and simply returns principal at the end (a zero-coupon bond) has a duration equal to its maturity, because all the value arrives at one moment. A bond that pays generous coupons every six months has a shorter duration than its maturity date suggests, because a meaningful share of its value returns to the investor well before the final payment.
The more commonly quoted figure in markets today is modified duration, a close cousin that converts this time-weighted average into a direct estimate of price sensitivity. Modified duration answers the question investors actually care about: for every one-percentage-point move in prevailing interest rates, roughly what percentage will this bond’s price move, in the opposite direction. A bond with a modified duration of 7 will tend to lose about 7% of its market value if rates rise by one percentage point, and gain roughly 7% if rates fall by the same amount. This inverse relationship, prices falling as rates rise and vice versa, is the mechanical foundation on which duration rests, since a bond’s fixed coupon becomes less attractive relative to newly issued debt paying a higher rate, so its market price must adjust downward to keep its effective yield competitive.
Why maturity alone is a poor substitute
It is tempting to assume that a 10-year bond is simply “riskier” to rate changes than a 5-year bond, and in general that intuition holds, longer-dated bonds do tend to have higher duration. But maturity and duration are not the same thing, and the gap between them matters. Coupon size, payment frequency, and any embedded features (such as a call provision that lets the issuer redeem the bond early) all pull duration away from raw maturity. A 30-year bond with a very high coupon can have a materially lower duration than a 30-year bond paying a minimal coupon, because more of the high-coupon bond’s value is returned to the investor sooner.
This is precisely why professional bond investors and portfolio managers lean on duration rather than maturity or yield when assessing rate exposure. Yield tells you the income a bond currently generates. Maturity tells you when the principal is scheduled to return. Neither tells you how the bond’s market price will react if the interest-rate environment shifts before that maturity date arrives. Duration fills that specific gap, and it does so in a single, comparable number that works across bonds of different structures, coupons, and issuers.
Duration’s limits and why context still matters
Duration is a linear approximation, and interest-rate-driven price changes are not perfectly linear, a property described by convexity. For small rate movements, duration estimates tend to track actual price changes closely. For larger swings, the estimate can diverge from what actually happens, which is why more sophisticated analyses often pair duration with a convexity adjustment for a fuller picture. Duration also assumes a parallel shift across the entire yield curve, meaning short-term and long-term rates move by the same amount together, a simplification that does not always hold in real markets, where short and long rates can move independently or even in opposite directions.
None of this diminishes duration’s usefulness as a starting point. It converts an abstract worry, “what happens to my bond if rates move”, into a concrete, comparable estimate, which is precisely why it remains one of the first figures fixed-income analysts and portfolio managers examine, well before yield, when sizing up how a bond or a bond portfolio might respond to a changing rate environment.
