The Bloomberg US Aggregate Bond Index, the standard benchmark for the investment-grade US bond market, returned negative 13.01% in calendar year 2022. This was the worst calendar year in the index's history, which goes back to 1976. Mainstream bond ETFs that track the index, including the iShares Core US Aggregate Bond ETF (AGG) and the Vanguard Total Bond Market ETF (BND), produced very similar negative returns over the same period.
For long-dated US Treasury funds the picture was worse. The iShares 20+ Year Treasury Bond ETF (TLT), which holds Treasuries with at least twenty years remaining to maturity, peaked above $170 per share in 2020 and traded as low as the low $80s in October 2023. The peak-to-trough decline was approximately 50%, in an instrument that holds nothing but securities issued by the United States government and is widely described in retail materials as "safe."
The bonds inside these funds did not default. There was no credit event. The decline was caused by interest rates rising, and the mechanism by which rising rates cause a bond fund to fall is one of the most useful pieces of fixed income arithmetic a retail investor can learn. This piece walks through it from first principles, contrasts the behaviour of a bond fund with the behaviour of an individual bond held to maturity, and ends with a practical decision framework for when each is the right tool.
How a bond price moves when rates change
A bond is a contract to pay specified cash flows on specified dates. A ten-year US Treasury issued at par with a 4% coupon pays $20 every six months for ten years and then returns the $1,000 face value at maturity. The cash flows are fixed at issuance and do not change with the market.
What changes is the price an investor is willing to pay for those cash flows. If, the day after issuance, prevailing yields on equivalent ten-year Treasuries rise to 5%, no buyer in the secondary market will pay $1,000 for a bond paying $20 every six months when they could buy a new bond paying $25. The price of the existing bond falls until the implied yield to maturity of the cash flows matches the new market rate of 5%. The arithmetic of how much the price falls is captured by the bond's duration.
Modified duration, expressed in years, is the approximate percentage change in a bond's price for a 1 percentage point change in yield. A bond with a modified duration of seven years will fall approximately 7% in price if its yield rises by 1 percentage point. A bond with a modified duration of seventeen years will fall approximately 17%. The relationship is not exact (the second-order term, called convexity, makes the actual move slightly less negative for large rate increases and slightly more positive for large rate decreases), but for typical rate moves the duration approximation is close enough to be useful.
The rule of thumb that follows from this is simple: longer-dated bonds are more sensitive to rate changes than shorter-dated bonds. A two-year Treasury fund will move very little when ten-year yields rise by 1 percentage point. A twenty-year Treasury fund will move a great deal. The investor who buys a long-dated bond fund is, whether they know it or not, taking a substantial duration position.
What rates actually did between 2020 and 2023
The 10-year US Treasury yield, available daily from the Federal Reserve's data release at H.15 Selected Interest Rates, fell to roughly 0.5% in mid-2020 (the lowest level in the series' history). It then rose, in roughly straight-line fashion through 2022 and into 2023, peaking around 5.0% in October 2023. The total move was approximately 4.5 percentage points over three years.
For an aggregate bond index with a duration around six years, that move predicts roughly a 27% cumulative price decline from the rate move alone, partially offset by the coupon income earned during the period. The actual cumulative total return of the iShares Core US Aggregate Bond ETF (AGG) from its 2020 peak through October 2023 was approximately negative 20%, consistent with the duration arithmetic plus coupon income.
For a long-dated Treasury fund with a duration around 17 years, the same 4.5 percentage point rate move predicts roughly a 76% price decline. The actual peak-to-trough decline of TLT was approximately 50%, less extreme than the linear approximation predicts (because of convexity, and because the yield rise was spread over three years during which the fund earned coupon income). The order of magnitude is right.
None of this required any of the bonds inside the funds to default. It required only that interest rates rose. The mechanism is mathematical and is the same for every bond fund in every currency in every jurisdiction.
Why an individual bond held to maturity behaves differently
An individual bond, held by an investor who intends to hold it until it matures, has a different risk profile from a bond fund holding the same bond. The bond pays its scheduled coupons and returns its face value at maturity, regardless of what happens to interest rates in the meantime. If the investor bought the bond at $1,000, holds it for ten years, and the issuer does not default, the investor receives the contracted cash flows and the principal. The market value of the bond will fluctuate during the holding period, sometimes substantially, but the fluctuation does not affect the realised outcome if the investor simply waits.
A bond fund, by contrast, never matures. The fund holds a portfolio targeted to a specific duration, and as bonds inside the portfolio approach maturity they are sold and replaced with longer-dated bonds. The fund's duration stays roughly constant over time. An investor who buys a 17-year-duration Treasury fund and holds it for 30 years is, throughout that period, exposed to 17-year duration risk. There is no point at which the fund "matures" and returns the investor's principal.
This is the most important practical distinction between the two formats. An investor with a known liability at a known future date (a planned house purchase in five years, a child's tuition in ten years) can match the liability with an individual bond, or with a defined-maturity bond ETF such as the iShares iBonds series, and lock in the outcome regardless of what happens to rates in the interim. An investor in a constant-duration bond fund cannot. The constant-duration fund is making a perpetual bet on the path of interest rates, even if the investor never thought of it that way.
Why bond funds are still the right tool most of the time
None of the above is an argument against bond funds. For most retail investors with diversified, multi-decade portfolios, the constant-duration bond fund is the right tool, for three reasons.
First, building a properly diversified individual bond portfolio across credit qualities, sectors, durations and issuers is impractical for retail investors. The minimum trade sizes in the corporate bond market often start at $10,000 or higher per bond, and the secondary market spreads for retail-sized trades are wide. A retail investor with $50,000 to allocate to bonds cannot replicate the diversification of a bond fund holding 8,000 issues at a fraction of a basis point of marginal cost.
Second, the rebalancing problem with individual bonds is real. As bonds mature, they need to be reinvested. Doing that across a portfolio of dozens of bonds is operational work. A bond fund handles it automatically.
Third, the duration risk of a bond fund is a feature, not a bug, in a portfolio context. A long-duration bond fund will fall hard when rates rise but will rise hard when rates fall, which is the period in which equity portfolios usually do worst. The negative correlation of long-duration Treasuries with equities, observable in most (though not all) recent recessions, is precisely what makes them useful as a portfolio hedge. The investor who owns the bond fund for that hedging property has to live with the fact that during periods when rates are rising, the hedge is going to lose money.
The decision framework
For a retail investor reading this, the practical question is: bond fund or individual bond? The honest answer depends on what the bond is for.
- Money you need on a specific date: individual bond, defined-maturity bond ETF (such as the iBonds series), or a bank fixed deposit with a matching maturity. Avoid constant-duration bond funds for this purpose.
- The bond allocation in a long-term portfolio: constant-duration bond fund or ETF, sized to the duration the investor is comfortable holding. Most multi-asset model portfolios use intermediate-duration bond funds (around 6 to 8 years duration) for this role.
- An equity-portfolio hedge: long-duration Treasury fund (such as TLT or the European equivalent IDTL). The investor has to accept that this hedge underperforms badly during rising-rate periods and is most valuable during recessions when rates fall.
- Cash you might need at any moment: money market fund or short-duration bond fund (1 to 3 year maturity). The duration risk on these is small enough that even a sharp rate move produces only a few percent of price impact.
The investor who matches the tool to the purpose will not be surprised by a 50% decline in a long-duration Treasury fund, because they will not be using a long-duration Treasury fund for any purpose where a 50% decline would be unacceptable. The investor who reads about "safe" bonds in a generic personal finance article and buys whatever the broker's app puts at the top of the bond category is the investor who will be surprised. Avoiding that surprise is the point of understanding duration.