A listener recently asked: “Would you be able to cover long-dated fixed-rate bonds? I feel like we’re getting close to considering locking in some duration in Australia.”
Duration measures how sensitive a bond’s price is to changes in interest rates. The longer the bond’s maturity, the more its price will generally move when rates change.
The case for buying longer-dated bonds is that, if we assume yields have peaked, their prices could rise when interest rates fall. I imagine that’s the thinking behind the listener’s question.
But I think the more useful question is: does the potential return justify taking the extra risk?
What a 10-year bond yield tells you
Australian 10-year government bonds are yielding around 5.20% to 5.25%, their highest starting yield in roughly 15 years. With the RBA cash rate at 4.60%, that’s a term premium of about 0.60%. At first glance, it looks like a reasonable time to extend duration.
After working through the numbers, I don’t think it is. The term premium hasn’t disappeared. The alternative to taking duration risk has become more attractive.
The textbook case for holding longer-dated bonds is the term premium: the extra yield you receive for accepting the risk that interest rates will change over the life of the bond. That case was compelling for a long time because there weren’t many alternatives. A retail investor without an adviser could hold a term deposit or take duration risk through a long-dated bond fund. Those were probably the two main options. But there are more options available today that I think offer better risk-adjected returns.
Credit risk and duration risk are different
It’s useful to separate fixed-interest risk into two components because they behave differently and the market prices them differently.
The first is credit risk: the risk that the issuer will not be able to repay you, and the extra return the market demands for taking that risk. Credit spreads have two parts. There’s issuer-specific risk, where a smaller or lower-rated bank pays more than a major bank for a comparable investment. And there’s market-wide risk, where spreads widen because investors become more concerned about credit generally, regardless of any one issuer’s fundamentals.
The second is duration risk: how sensitive an investment’s price is to changes in the risk-free rate, regardless of who issued it. A long-dated, fixed-rate investment has more duration. A floating-rate note resets its coupon periodically, so its price is much less sensitive to rate changes. It can carry meaningful credit risk while having little duration risk.
A bond’s value and risk depend on how long it has until maturity, whether its interest rate is fixed or floating, and how its credit spread might change. These factors combine to determine its overall interest rate risk.
Why your starting yield often shapes your return
For investors who buy and hold bonds, the starting yield has historically been the best single indicator of their return over the following several years. It’s hard to earn more than the yield you start with.
Consider investors who bought government bonds between 2019 and 2021, when 10-year yields fell to around 1% p.a. and, briefly, below it. There was little room for yields to fall further, but plenty of room for them to rise. That’s what happened: the Bloomberg AusBond Index fell about 10.5% in the 2022 financial year, its worst year on record. The income from a starting yield of around 1% wasn’t enough to offset the capital losses.
Today’s (23/10/2026) 10-year government bond starting yield is around 5.20% to 5.25% is the highest in roughly 15 years. That puts investors in a much better position than buyers in 2019 to 2021. But a high starting yield does not automatically make long-dated bonds the right choice. It also means that, if rates unexpectedly rise further, the potential loss is smaller than it was when yields were around 1%.
How much are you really being paid to take duration risk?
The term premium is the extra yield you receive for holding a 10-year bond, over and above the average cash rate expected over those same 10 years. It’s tempting to estimate it by subtracting today’s cash rate from the 10-year Australian government bond yield. With the yield at about 5.2% and the cash rate at 4.60%, that gives 0.60%. But this assumes the cash rate stays at 4.60% for the next decade, which is unlikely.
What matters is the average cash rate the market expects over the next 10 years, not today’s rate. Markets are currently pricing in a strong chance of another RBA rate rise, possibly in a few weeks’ time. However, fixed interest expert, Christopher Joye, founder of Coolabah Capital, estimates the term premium on Commonwealth government bonds is currently about 0.9%, compared with an average of 0.5% since 1999. Under these assumptions, that implies the market expects the cash rate to average about 4.3% over the decade (5.2% less 0.9%), which is below today’s 4.60%.
Even if we accept the 0.9% estimate, and acknowledge that it is attractive by historic standards, is it enough compensation for taking interest rate risk? Many factors could change interest rate expectations, as I discussed last month here.
The term premium now has a stronger competitor
Historically, investors took on duration risk not because the premium was especially attractive, but because it was one of the few ways to earn more from defensive investments. If a retail investor wanted to earn more than a term deposit interest rate, a long-dated bond fund was one of the few options. Even a modest term premium could make sense because there was little else to do with that part of the portfolio.
But that has changed. In the past few years, retail investors have gained access to a broad array of ETFs that hold investment-grade bank subordinated debt. These are low-duration investments. Betashares’ BSUB and VanEck’s SUBD are two examples.
BSUB currently has an all-in yield of about 5.96%, a running yield closer to 6.3%, a modified duration of just 0.20 years and an average credit rating of A- (the running yield is higher than the all-in yield because some of the bonds are currently trading above par value).
SUBD is similar but also includes regional and foreign banks rated BBB+ to BBB, alongside the major banks (weighted average credit rating is still attractive at A+). It offers a slightly higher yield, with slightly more credit risk.
Neither fund has meaningful sensitivity to the RBA’s next move, and future yields will move in like with any RBA rate moves.
Many of the bonds held by these subordinated debt ETFs currently trade above face value. As they mature or are redeemed, investors generally receive the face value, which may be less than its current market price. That expected capital loss is already reflected in the fund’s yield to worst. It does not disappear if the fund keeps holding bonds or replaces them with new ones. Because these are ETFs, their holdings change over time, so the fund’s actual return will depend on the yields, prices and terms of the bonds it holds, as well as fees and credit events.
BSUB and SUBD are useful comparisons when weighing up duration risk alternatives. If an alternative offers a higher return without the same exposure to interest rate movements, and not materially more credit risk, then it probably offers a more attractive risk-adjusted return. Some readers will correctly point out the sovereign debt is safer than corporate debt and that is true, of course. In my view, A-rated Australian bank debt is not materially riskier in practice. It has also been very capital stable, while offering a higher yield and almost no interest rate risk.
I should also mention bank hybrids. They have been available on the ASX for decades, with fully franked distributions as one of their main attractions. Bond interest, by contrast, doesn’t come with franking credits. ASIC has since banned banks from issuing hybrids to retail investors because they deemed them too complex. So, the hybrid market is now winding down, with only about 17 bank hybrid securities remaining. For that reason, we prefer to avoid the hybrid market.
The risks that could change the comparison
I should also acknowledge a potential risk. Spreads on major-bank Tier 2 debt are currently relatively tight, at around 110 to 145 basis points above swap rates. Investors are being compensated for an ordinary credit cycle but may not be adequately compensated for a serious banking-sector shock, if one occurs.
Credit spreads have also widened overseas in recent weeks, particularly in France, where concerns about government finances have pushed up borrowing costs. If that spread to Australia, the price of bank subordinated debt could fall, even though the income would continue to be paid. It’s a risk worth watching rather than one we’re overly concerned about at the moment, and investors who concerned are can move up the credit scale into senior bank debt ETFs such as QPON, or into government and semi-government bonds (AGVT), in exchange for a lower yield.
There’s another risk worth making clear. Bonds don’t always hedge shares in the way the textbook suggest they should. In 2022, government bonds and equities fell together, and their correlation reached its highest level in more than 20 years. That’s because the shock was inflation and rising interest rates, rather than weak economic growth. Duration can help protect a portfolio against a growth shock but offers little protection against an inflation shock. Inflation is closer to the risk we face today.
The reader who prompted this blog post mentioned PGBF (PIMCO’s Global Bond Active ETF), as paying more than 7%. The fund holds a diversified, currency-hedged portfolio of global bonds, with an average credit rating of AA and duration of about 6.7 years. The fact sheet reports an average coupon of about 4% and an estimated yield to maturity of 7.26%. The yield to maturity is higher because the bonds trade below their face value. It assumes the bonds are held to maturity and pay as contracted, allowing for the potential capital gain as they move towards face value.
Of course, PGBF is an ETF, not a single bond with a fixed maturity date. Its holdings change over time, so investors cannot simply hold the ETF until all its bonds mature and expect to earn the stated yield to maturity. The fund’s actual return will depend on how bond prices, interest rates and other factors change. The 7.26% figure is an estimate, not necessarily a likely return. That’s why the running yield and the fund’s actual distributions are the two most important figures to focus on when assessing potential “income” return.
Why we’re keeping duration low for now
For now, we hold the defensive part of client portfolios in investment-grade credit with very little duration, rather than extend into long-dated government or global bonds.
The term premium has not disappeared, but it may be smaller than the headline spread suggests. It also must compete with an alternative that were not available to most retail investors a decade ago: an investment that currently offers a comparable or higher yield with much less interest rate risk.
This may not be a permanent position, but if nothing else, I hope this blog highlights why it’s important to understand an asset class before choosing an ETF/investment option within it.
Who this discussion is for
This blog is for investors who have already decided how much of their portfolio is to be held in growth and defensive assets.
I have explained how we think about that decision here. In general, investors with a long-term horizon who accept volatility as the price of higher returns should hold most, if not all, of their portfolio in growth assets. In retirement, however, it may make sense to keep 1 to 3 years of living expenses in cash, such as in offset accounts, or in other defensive assets.
