Global Markets Daily: Re-Assessing the Value of Duration for Balanced Portfolios

In “GOAL Post: The name is bond — rethinking duration in balanced portfolios” we set out a framework for assessing the role of bonds in multi-asset portfolios after the sharp rise in yields. Here we highlight the key investment implications, focusing on the drivers of bond allocations across tactical, medium-term and strategic horizons and where investors can find attractive sources of carry, convexity and diversification.

Global bond yields have risen sharply YTD, driven by a combination of cyclical support, rising energy prices due to the Middle East war, crowding out from AI debt issuance and fiscal concerns. This follows the large bond sell-off in 2022. As a result, over the past 5 years US 10-year bonds delivered the worst return in more than a century – in real terms, returns were nearly as low as after WW1 and WW2 and in the 1970s.

However, despite the material bond sell-off, current levels of bond yields are not unusually high compared with long-run history – they were just unusually low during the post-GFC cycle. In fact, the current US 10-year yield, at 5%, is close to the average of 4.7% over the last 250 years (Exhibit 1, excluding the 1970s results in an average of 4.5% but during WW2 the Fed anchored long-dated bond yields via yield curve control, suggesting bond yields would have been much higher). Admittedly, such a long history is not comparable to now, as macro conditions have been more friendly for the current generation of bond investors. But, the macro regime has materially shifted, with multiple cyclical and structural headwinds for bonds.

Are bonds becoming more attractive again for multi-asset portfolios? Based on Markowitz’ Modern Portfolio Theory the optimal bond weight in a balanced portfolio is a function of (1) relative return, (2) relative risk, (3) diversification benefits and (4) the return on cash. With higher equity/bond correlations the optimal bond weight decreases, and that means lower Sharpe ratios and higher portfolio volatility. The optimal bond allocation also decreases with better relative equity returns vs. bonds and higher relative risk. Higher returns on cash can also drive a higher equity allocation as bonds become less attractive vs. cash. Importantly, a more positive equity/bond correlation amplifies all those shifts.

While higher bond yields suggest higher total returns for bonds in the medium term, the return relative to equities will be driven mostly by equities, equity/bond correlations are likely to be less negative compared with the post-GFC cycle, and term premia look set to remain relatively low in a long-run context.

In the near term, monetary policy and energy prices should remain the key drivers of both equities and bonds (Exhibit 2). So far, higher bond allocations have added volatility rather than buffered portfolios. Further yield increases, especially sharp moves driven by energy prices or more aggressive central bank tightening, would likely weigh more on equities than bonds; historically, increases above 2sd over 3 months have mattered most (Exhibit 3). Conversely, lower energy prices, continued core inflation normalisation and less hawkish central banks — consistent with our economists’ baseline — should help bonds stabilise into year-end and support equities. We are tactically Neutral across assets in our 3m allocation.

On a 6-12m horizon, the case for higher bond allocations would strengthen if growth slows materially or AI-driven earnings tailwinds fade. For now, strong earnings growth still favours equities over bonds: prospective global equity carry - which combines the starting dividend yield plus the subsequent change in earnings - has increased sharply YTD supported by the AI investment boom, and our equity strategists still expect slower but still strong earnings growth in 2027. This should keep equity risk-adjusted returns attractive versus bonds, even as relative returns moderate. We remain OW equities, N bonds and UW credit on a 12m horizon.

Over longer horizons, higher starting yields should lift optimal bond allocations from the unusually low levels of the past five years. Higher yields improve carry and medium- to long-term buy-and-hold returns, while providing more cushion against further yield increases. This strengthens the case for higher strategic bond allocations than in the past 5 years, when bonds had limited value in portfolios, but mainly reflects a normalisation back towards the long-run optimal range of 35-55%. However, even over 10-year horizons most of the variation across possible long-term scenarios is driven by equities, similar to the near term.

Managing duration risk and convexity

As we wrote in Balancing Innovation and Inflation in Portfolios, in the coming years we believe investors need to maintain exposure to innovation, protect against inflation and improve risk mitigation in portfolios. While longer duration bonds historically provided more convexity during recessions and ‘risk off’ periods, the potential for large declines in long-dated bond yields might be more limited in the next crisis, and near-term rates volatility could remain somewhat elevated. Our rates strategists see the largest potential for rates relief in the front end vs. forwards, pointing to positions in the belly of the yield curve – that said, those would offer less convexity.

Rates receiver swaptions on shorter-dated rates can provide a more convex hedge for a recession, with central banks likely to remain very responsive. With front-end forward rates having increased strongly YTD but rates volatility somewhat anchored, the entry point for front-end receiver swaptions appears attractive. This is both due to the central bank put in the event of growth risks but also near-term potential rates relief in the event of lower energy prices – during previous recessions, a receiver swaption on short-dated rates offered higher payoffs than S&P 500 OTM puts (Exhibit 4, based on current pricing). Also, long curve vol expressions can hedge against the more extreme growth/policy tails. On the flip side, 3m put options on longer-dated bonds (TLT) could help protect fixed income portfolios that have already more duration risk – higher 30-year yields allow investors to better fund put protection strategies (Exhibit 5).

With bonds providing less of a buffer, diversification within equities is becoming more important for risk mitigation. A new diversification opportunity may now be with bond-like equities: correlations between the S&P 500 and both low-volatility and high-dividend stocks have fallen sharply, similar to during the Tech Bubble (Exhibit 6). This has also supported Europe, which has relatively less Tech exposure, although it faces other headwinds from higher gas prices and politics from here. We like a barbell strategy of a well-diversified portfolio of global AI-exposed stocks and high-dividend, low-volatility stocks.

While low-volatility and high-dividend stocks have lagged the broad market since the COVID-19 crisis, they have performed better relative to the market excluding Tech. And in Tech-driven equity drawdowns, like the Deepseek sell-off in Q1 2025 and the momentum unwind in July 2026, they materially outperformed. In fact, the proportion of stocks with a negative beta to the market has increased sharply, very similar to the Tech Bubble (Exhibit 7). As we recently wrote, this is in large part due to the growing Tech and AI dominance in equity benchmarks – after the Tech Bubble burst, low volatility and high dividend yield stocks were also up, with S&P 500 down 46% and the TMT sector down 79% (Exhibit 8). This was in part because they benefited from lower bond yields, but even during the 2022 drawdown without a large decline in bond yields, they managed to outperform materially.

Equity-like bonds such as convertibles can provide upside exposure to equity through their embedded call option while having relatively low duration. As our credit strategy team highlighted, there has been a material pick-up in convertible issuance due to AI funding needs. Both during the late 1990s and around the GFC, convertibles outperformed vs. a 60/40 portfolio (Exhibit 9). They can outperform equities late cycle as they have an element of downside protection and rising volatility acts as a tailwind for convertible valuations because it may increase the value of the embedded option. As we recently illustrated, long-dated equity calls can be a very effective tool to tailor the risk/reward and manage equity exposure late cycle.

Best Trade Ideas Across Assets

1. Stay short SGD/MYR, opened January 24, 2026, at 3.13, with a target at 2.90 and a stop at 3.30, currently trading at 3.20.
2. Stay long TRY, NGN and KZT against the USD, as an equally weighted basket, opened February 18, 2026, at 0%, with a revised total return target at 12%, and a revised stop at 7%, currently trading at 9.7%.
3. Stay long 3y SOFR swap spread, opened April 17, 2026, at -22.6bp, with a revised target at -12bp inclusive of carry; and a revised stop at -17.5bp, trading at -15.7bp.
4. Stay short USD/EGP, opened April 24, 2026, at 0%, with a revised total return target at 12% and a revised stop at 5%, currently trading at 6.7%.
5. Stay short AUD/NZD, via 6m (14 Jan 2027) 1.1650 puts, opened July 14, 2026, at 1.2000, currently trading at 1.2448.
6. Stay long NIFTY Banks vs. short NIFTY Pharma, opened July 15, 2026, at 100 in local currency (INR), with a target at 115 and a stop at 90, currently trading at 95.
7. Stay short PLN/HUF, opened August 5, 2026, at 84.12, with a target at 80.5 and a stop at 87, currently trading at 83.46.
8. Stay short EUR/AUD, opened September 4, 2026, at 1.6123, with a target at 1.5750 and a stop at 1.6520, currently trading at 1.6125.
9. Stay long INR/IDR, opened September 4, 2026, at 187, with a target at 200 and a stop at 178, currently trading at 185.
10. Stay short SGD/CNH, opened September 4, 2026, at 5.29, with a target at 5.00 and a stop at 5.50, currently trading at 5.25.
11. Pay CHF 1y1y against EUR 1y1y, opened September 11, 2026, at 3.02, with a target at 2.65 and a stop at 3.25, currently trading at 3.02.
12. Receive 5Y HUF IRS, opened 17 September, at 5.5%, with a target at 4.9% and a stop at 5.8%, currently trading at 5.6%.
13. 2s10s PLN IRS steepeners, opened 17 September, at 61bp, with a target at 100bp and a stop at 45bp, currently trading at 56bp.
14. Stay long in equal weight BUX Index and ASE Index vs. short PX Index, opened 17 September, at 100 in USD, with a target of 115 and a stop at 90, currently trading at 101.
15. Stay long MXMU Energy vs. Industrials, opened 17 September, at 100 in USD, with a target at 115, and a stop at 90, currently trading at 100.
16. Go short 3m30y straddles on a delta-hedged basis, opened 18 September, at 0bp, with a target at 6bp, and a stop at -4bp.
17. Pay 5s on the 2s5s10s fly JGB, opened 18 September, at -8.6bp, with a target at 2bp, and a stop at -13bp.