Bonds Are Bonds Again — But the Index Still Can’t Choose

The Australian bond market has just handed investors something it has not offered in more than a decade: a genuine margin of safety in duration. As at 14 September, the 10-year ACGB was trading at 5.36% — its highest level since 2011 — after the front end repriced roughly 50bps in a month on a market-implied ~75% probability of an RBA hike on 29 September.

That repricing is uncomfortable if you simply own the index. It is an opportunity if you can choose.

Starting yields have changed the arithmetic

Over a one-year horizon, a bond’s total return is approximately its starting yield less modified duration multiplied by the change in yield. Set that to zero and you get the yield breakeven — how far yields can rise before price losses consume the income earned along the way.

Through the 2010s that margin of error barely existed. A 10-year ACGB yielding 1–2% with eight to nine years of duration could absorb only 15–25bps of further yield rises before its one-year total return turned negative — substantial duration risk for very little compensation. Today, at a 10-year yield of ~5.4% and duration of ~8 years, the breakeven is roughly 70bps.

Higher yields do not eliminate duration risk. They increase the amount of adverse rate movement you can withstand before that risk overwhelms carry. The starting yield has become a source of resilience.

That is the core of the “bonds are bonds again” argument, and we agree with it. But it is a statement about starting yields — not, on its own, a case for buying the index.

Passive owns duration; it does not choose it

An index-tracking bond fund or ETF gives you whatever duration the issuance calendar hands it, at whatever yield prevails, weighted towards the most indebted borrowers. It was fully exposed when the 10-year yielded 1% and it is fully exposed today at 5.36% — the same risk, on radically different compensation. It had no mechanism to shorten into the 2021–22 repricing, and none to lean in now that investors are finally paid for the risk.

The constraint is sharper here than offshore. The domestic index is narrow — sovereign and semi-government paper plus a handful of large financials — while much of what generates return in AUD credit sits outside it: off-benchmark investment-grade issuers, the bank Tier 2 wave created by the AT1 phase-out, and senior secured syndicated loans.

Those constraints are mechanical, not a question of skill. Index rules decide the universe, screening out unrated loans and smaller off-benchmark lines on issue size and rating before value is considered; rebalancing decides the timing, at whatever price prevails; and a downgrade forces a sale into a weak market irrespective of the credit view. Low fees are worth having; they are not worth paying for beta you cannot time.

The MDA carries none of that. We can hold to maturity and take capital back at par, decline a new issue or take the full allocation of one we like, sit in short floating-rate paper while we wait, and act on our own timing rather than a rebalance calendar. The mandate sets the boundaries; 50% investment-grade bonds and 50% bank syndicated loans, 20 holdings, per-issuer and per-security caps — and within them the judgement is ours.

How the MDA was positioned — and what we are doing now

The IAM Income MDA has run a deliberately short, floating-rate-biased book: modified duration of roughly 1.0–1.5 years. That duration was cut deliberately from 3.0 years, and early — ahead of the renewed Middle East escalation and the move in crude back above US$85/bbl that rebuilt the inflation risk premium across the curve. An index fund had no means of making that decision, and it is the main reason the subsequent back-up in yields has cost this portfolio far less than it has cost long-duration bond funds.

From that starting point we are now selectively adding duration in the AUD market. Selectively is the operative word:

• Where the curve has moved most. The 10–15 year segment bore the largest share of the repricing, so the carry cushion is deepest there. Better to take extension risk at that point than lengthen the book uniformly.
• Through quality, not quantity. A smaller amount of duration in high-grade corporate and semi-government paper at cycle-high yields, rather than a great deal of it further down the credit curve.
• Via primary, where we are paid to. New-issue concessions and the bank Tier 2 issuance wave are the most efficient way to extend, and our capital markets relationships give direct access to that pipeline.

A recent example: selling a 2055 floater to buy a 2036 fixed line

A recent switch makes the point: we sold the Dexus 2055 subordinated floating-rate note (DXSAU Float 12/10/55) and redeployed into Aurizon’s senior fixed-rate September 2036 line (AZJAU 6.84 09/16/36). Two things changed, both intentionally.

• Interest rate duration went up — by design. The Dexus note is a floater, resetting off BBSW, so it carried almost no interest rate duration and gave the portfolio no participation in a curve that had just repriced ~50bps. The Aurizon line is fixed, ten years out, bought at a yield of 6.84% — a level unavailable in this part of the curve for most of the past decade.
• Credit risk went down at the same time. The Dexus holding was subordinated, dated 2055, with long spread duration and extension risk if not called. Aurizon is senior unsecured paper from an established rail freight and regulated network infrastructure business. We moved up the capital structure and improved the rating from BBB to BBB+ S&P ‘equivalent’.

Credit health and income, not duration, remains the core driver of returns here — but when the cushion moves from 20bps to 70bps, engaging with duration is an active decision that makes more sense.

The other half of active: what you choose not to own

Security selection cuts both ways. The 25 August voluntary administration of Bathla’s Universal Property Group saw Teneo appointed across 542 entities, with preliminary known creditor claims of approximately $3.4bn. The IAM Income MDA held no direct exposure to Bathla or the UPG group.

Our one indirect point of contact is a position we remain comfortable owning, and the episode has if anything reinforced the case for it: the La Trobe Financial 1st Lien Term Loan B. La Trobe’s disclosed Bathla exposure was $38.1m — under 0.15% of a A$23bn book — is secured by first registered mortgages at a weighted average LVR of 62.5%. The development exposure is ~95% complete, fully funded to completion and presales cover 89% of the current loan balance. La Trobe expects to recover the exposure in full. Critically, that lending sits in its managed funds rather than in our borrower, the asset management group, and our facility holds a first-priority lien at 2.0x net leverage against a business valued at A$3.0bn in April 2026. That is precisely the credit profile we underwrote.

Nor is that accidental: every syndicated loan is Investment Committee-approved, senior secured and covenant-tested, capped at 10% per issuer and 5% per security across 20 holdings.

What followed matters just as much. Centuria Bass and CVS Lane suspended applications and redemptions, and MA Financial capped monthly redemptions at 1% of FUM despite no Bathla exposure at all — roughly $5.1bn of assets affected. Liquidity contagion does not require credit losses, only a pooled structure and nervous co-investors.

In a pooled vehicle you own your neighbour’s redemption behaviour as well as your own credit risk. In an MDA, you own your bonds.

The MDA answers that directly: assets registered to each investor, no pooling, custody through Perpetual Corporate Trust, full transparency over every holding and fee, and a 30-day redemption pathway independent of what other investors do.

The bottom line

Higher starting yields have restored the defensive value of fixed income. But the volatility that produced them has also produced dispersion — between segments of the curve, between issuers, and between managers who underwrote carefully and those who did not. Passive captures the average; active decides which side of it to sit on. In a market repricing this quickly, that is worth far more than a few basis points of fee saving.