Rethinking global aggregate bonds for today’s markets

The Bloomberg Global Aggregate Bond Index has long been the default benchmark for global fixed income. But some of the conditions that supported long duration through the multi‑decade disinflationary era have changed. Sovereign financing needs are higher, while the artificial intelligence (AI) build-out is adding to corporate duration supply. At the same time, central banks are absorbing less duration, and traditional sources of long‑end demand have weakened. Uncertainty around inflation and long‑term rates has also risen. For long duration bonds, this greater uncertainty around inflation, real rates, and term premia can translate into a wider range of price outcomes.

Against this backdrop, this paper examines whether a more deliberate approach to Global Aggregate beta—focused on the 1- to 10‑year universe—could offer advantages. This segment retains a broadly similar sector mix and has historically captured most of the parent benchmark’s annualized return, with materially lower volatility and shallower drawdowns. The principal difference is substantially less exposure to long‑end duration.

This is not an argument for excluding long duration bonds from the opportunity set. Rather, the current environment provides a reason to reconsider how much long‑end duration should be embedded structurally in strategic beta. A 1- to 10‑year benchmark could offer an alternative—retaining broad exposure to the Global Aggregate while allowing investors to make a more deliberate decision about how much long‑end duration to own, and when.

A shifting backdrop for long‑term rates

For much of the past four decades, duration benefited from a powerful set of supports. The monetary policy response to the Great Inflation of the 1970s helped anchor inflation expectations, while globalization and expanding productive capacity added further disinflationary pressure. Later, quantitative easing (QE) removed substantial duration from private markets. Against this backdrop, inflation and long‑term yields trended lower for much of the period.

Today, the supply‑demand balance for long duration bonds looks different. Fiscal deficits are larger and more persistent, increasing sovereign financing requirements. The AI infrastructure build‑out is adding another source of duration supply as large technology companies raise debt, including at longer maturities. Economists at the Federal Reserve Bank of Dallas note that Wall Street estimates of AI‑related investment‑grade issuance are centered around $300 billion in 2026, potentially generating as much as $360 billion of new duration supply in 10‑year equivalents.¹

At the same time, some traditional sources of long‑end demand have weakened. Central banks are absorbing less duration than during the QE era, defined benefit pension systems have declined in importance, and aging populations are shifting from saving toward dissaving in many countries. Reserve accumulation by central banks and sovereign wealth funds also provides less structural demand than in the past.

The inflation outlook has also become more uncertain. Supply chain resilience, geopolitical risks, and energy shocks can add to inflationary pressures. AI cuts both ways: Productivity gains could be disinflationary over time, while the near‑term infrastructure build‑out is resource intensive.

Together, these shifts point to a wider range of potential outcomes for inflation, real rates, and term premia than during much of the multi‑decade disinflationary era, with near‑term inflation risks more skewed to the upside. For long duration bonds, greater uncertainty around these variables likely translates into a wider range of potential price outcomes.

How much long‑end exposure is embedded in the Global Aggregate Bond Index?

Global Aggregate provides broad diversification across countries, issuers, and sectors. Its market‑value weights, however, can obscure where interest rate risk is concentrated. As of July 31, 2026, the 10+ year segment represented approximately 20% of benchmark market value but accounted for around 44% of its duration exposure.²

This means almost half of the benchmark’s duration risk is concentrated in the part of the curve most sensitive to long‑term inflation expectations, real rates, and term premia. With uncertainty around these variables elevated and near‑term inflation risks skewed to the upside, the consequences of carrying that exposure structurally are greater.

The changing risk and return experience at the long end

For much of the multi‑decade decline in yields, long duration was well rewarded. Greater interest rate sensitivity generated capital appreciation as yields fell, particularly following the global financial crisis. But that sensitivity worked in reverse as yields reset higher after the pandemic. By the end of July 2026, the rolling five‑year annualized return for the 10+ year segment had fallen to −3.8% (see Figure 2).

Higher volatility at the long end is expected given its greater duration. More notable is the magnitude of the recent increase. By the end of July 2026, rolling five‑year volatility had risen to 9.5% for the 10+ year segment, compared with 3.7% for the 1- to 10‑year segment (see Figure 2).

Long duration bonds have also traditionally played an important defensive role when growth weakens and central banks ease. Recent experience illustrates that this relationship is not automatic. The long end lagged 1- to 10‑year bonds not only through the 2022 to 2023 tightening cycle, but also during the subsequent easing cycle.

Policy easing affects the front end directly but does not necessarily lead to a rally in long‑term bonds. Long‑term yields remain sensitive to inflation expectations, real rates, and term premia, as well as the supply‑demand balance for long duration assets. When these forces keep long‑term yields elevated, rate cuts may not translate into long duration outperformance.

Not all duration is rewarded equally

The current yield curve offers additional income for extending maturity, but the incremental yield is modest relative to the additional interest rate exposure required. As of July 31, 2026, extending from the 1- to 3‑year segment to the 10+ year segment adds around 1.3 percentage points of yield, but requires roughly seven times the duration and has historically been associated with almost six times the realized volatility (see Figure 3).

The pattern holds across the maturity spectrum. Much of the yield available in the Global Aggregate can be accessed without taking the full duration exposure of the long end.

These measures are not expected‑return or relative‑value forecasts. Rather, they illustrate the additional duration exposures currently required to capture the incremental yield available at longer maturities.

A different balance between income and price return

Maturity changes the balance between income and price as sources of return. Across the full period from January 1, 2006, to July 31, 2026, income was the dominant positive return source for both 1- to 10‑year and 10+ year bonds. The key difference lies in price sensitivity: Annualized price‑return volatility was approximately 7.4% for 10+ year bonds, compared with 2.4% for 1- to 10‑year bonds, while coupon volatility was only around 0.2% for both.

The two contrasting rate environments in Figure 4 illustrate how this price sensitivity can work for or against investors. From 2006 to 2019, falling yields allowed price appreciation to add to long‑end returns. From 2020 through July 2026, rising yields had the opposite effect, with price declines overwhelming income at the long end, while the impact on 1- to 10‑year bonds was much smaller.

A 1- to 10‑year maturity constraint reduces how much of the return outcome depends on the more volatile price component associated with long duration exposure. This can result in a more income‑anchored strategic beta, while preserving the ability to add long duration exposure as an active allocation.

Refining the Global Aggregate: what changes?

Moving from the full Global Aggregate to its 1- to 10‑year universe changes the maturity and duration profile much more than the broad sector composition. As Figure 5 shows, the 1‑ to 10‑year universe retains a broadly similar sector mix to the parent benchmark.

The maturity constraint primarily reduces exposure to long‑dated sovereign and corporate bonds—the parts of the market where the supply‑demand shifts discussed earlier may be particularly relevant.

The historical trade‑off

Historical data may provide a useful way to assess the trade‑offs involved in moving from the full Global Aggregate to its 1- to 10‑year universe.

Over more than 20 years, the 1‑ to 10‑year index captured approximately 97% of the Global Aggregate’s annualized return, with 28% lower volatility and a maximum drawdown that was 4.2 percentage points shallower.

Historically, reducing structural long‑end exposure changed the risk profile much more than the return outcome. The different maturity profile resulted in approximately 1.16% annualized tracking error relative to the Global Aggregate.

Refining, rather than replacing, the Global Aggregate

The Global Aggregate remains a useful representation of global fixed income. However, under a market‑value weighting approach, the amount of duration embedded in the benchmark reflects the maturity profile of outstanding debt, rather than the prospective compensation for taking that duration risk.

A 1- to 10‑year benchmark offers one potential refinement. It retains broad exposure to the Global Aggregate while reducing the amount of long‑end duration held structurally in strategic beta. The historical evidence suggests that this can materially change the risk profile without materially changing the return outcome.

Long duration bonds remain an important part of the opportunity set. The case for a 1- to 10‑year approach is not to exclude them, but to make long duration exposure an active allocation decision—adding it when the prospective compensation warrants the additional risk rather than holding it structurally through the benchmark by default.