Macaulay Duration
Macaulay Duration is the weighted-average time, expressed in years, it takes to receive all of a bond's cash flows - coupon payments plus the return of principal - where each payment's weight is the present value of that payment relative to the bond's total price. It answers a simple question: "on average, how long do I have to wait to get my money back from this bond?" A bond that pays no coupons at all (a zero-coupon bond) has a Macaulay Duration exactly equal to its time to maturity, because the entire cash flow arrives on one single date. Any bond that pays periodic coupons has a duration shorter than its maturity, because some cash comes back earlier.
The concept was introduced by Canadian economist Frederick Macaulay in his 1938 study of interest rates, bond yields and stock prices in the United States, where he was trying to find a better single number than "years to maturity" to describe a bond's exposure to interest-rate changes. Two bonds can both mature in 10 years, but if one pays a high coupon and the other pays none, the high-coupon bond returns cash to the investor sooner on average - so its price is less sensitive to a change in interest rates. Macaulay Duration was his way of capturing that difference in a single, comparable figure.
The formula sums, for every cash flow, the time until that payment multiplied by its present value, then divides the total by the bond's current price (the sum of all present values):
Macaulay Duration = Σ [t × PV(CFt)] ÷ Bond Price, where t is the time until each cash flow and PV(CFt) is that cash flow discounted at the bond's yield to maturity.
A concrete example: a 5-year bond with a $1,000 face value, a 6% annual coupon ($60 per year) and a yield to maturity of 6% (so it trades at par, $1,000) has a Macaulay Duration of approximately 4.47 years - shorter than its 5-year maturity because the coupon payments return some cash before year 5. Compare that to a 5-year zero-coupon bond with the same $1,000 face value and 6% yield: its Macaulay Duration is exactly 5.00 years, because the investor receives nothing until maturity. The higher the coupon rate (holding maturity and yield constant), the lower the Macaulay Duration, because more cash arrives earlier in the bond's life.
For a trader or investor, duration matters because it is the standard first-pass measure of interest-rate risk: a bond or bond portfolio with a longer duration will see its price move more for a given change in interest rates than one with a shorter duration. A closely related figure, Modified Duration, is derived directly from Macaulay Duration (Modified Duration = Macaulay Duration ÷ (1 + yield/n)) and is used to estimate the actual percentage price change for a given change in yield - for example, a bond with a Modified Duration of 4.3 would be expected to lose roughly 4.3% of its price if yields rose by 1 percentage point, all else equal. Portfolio managers use duration to match assets to liabilities (pension funds try to match the duration of their bond holdings to the duration of their future payout obligations), and traders use it to compare interest-rate risk across bonds with very different coupon structures and maturities.
The measure needs a known schedule of future payments, which a bond has and a currency, a commodity or a token does not, so it belongs to fixed income and to the funds that hold it. Equities are the interesting exception: a share has no contractual schedule, and analysts nonetheless describe a company whose value rests on profits expected far in the future as long duration, because it behaves like a long bond when interest rates move. That is an analogy rather than a calculation, and it explains a great deal about which shares fall hardest when yields rise. The concept does resurface indirectly wherever fixed-income logic applies across markets - for instance, bond-like crypto lending products or corporate debt issued by mining and energy firms tied to commodities can be analyzed with the same duration framework, but the term itself is a bond-market concept first and foremost.
The estimate above is a straight line laid against a curve, and the distinction stops being academic on a large move. The relationship between a bond's price and its yield bends, so the simple calculation is accurate for small changes and progressively wrong for big ones, always in the same direction: it overstates the loss when yields rise a long way and understates the gain when they fall a long way. That curvature has its own name, convexity, and it is why a long-dated bond in a severe sell-off falls somewhat less than the arithmetic implies, and why a portfolio hedged on duration alone is not actually hedged once the move becomes interesting.
What makes 2022 worth remembering is not that the measure was imperfect but that it worked. A government bond fund holding paper with a duration near twenty years was going to lose roughly a fifth of its value for every percentage point that yields rose. That was not a forecast, a scenario or an estimate of risk; it was the arithmetic definition of what the fund held, printed on the factsheet, available to anyone who read it. Yields rose, the losses arrived, and a great many holders were astonished by the behaviour of what they had been told was the safe part of their portfolio. The number had been there the whole time, and the lesson is less about bonds than about how rarely a disclosed risk is read as a description of what will happen.