Strategy commentary cont'd: Government bond markets sold off sharply across all major regions as investors grappled with the combination of stronger energy prices, elevated inflation and shifting central bank expectations. The geopolitical backdrop repeatedly pushed yields higher and flattened curves during the month, although the market reaction to each fresh escalation became progressively more muted as investors grew accustomed to the pattern.

In the US, a weak non-farm payrolls print early in July and subsequently softer CPI and PPI data, with headline CPI unexpectedly declining in June to leave annual inflation at 3.5%, pushed expectations for the next Fed rate increase further out.

The underlying inflation pulse nevertheless remains uncomfortably strong: six-month annualised core PCE is running at 3.8%, a clear acceleration on the 3.3% twelve-month rate and nearly double the Fed's target, with three of the last four monthly prints holding at that pace and inflation now above the 2% goal for more than five consecutive years. Our Taylor Rule modelling implies the policy rate should be a further 75 to 100 basis points above the current 3.50–3.75% range.

Against this backdrop, Fed Chair Kevin Warsh left his target interest rate range unchanged in a 9-3 decision, with three FOMC members dissenting in favour of a 25bp hike. Warsh concurrently abolished forward guidance by declining to offer any hints on the future rate path. The result was a sharp bear-steepening of the US curve: the 2-year yield, the purest read on near-term policy, actually fell 4bps to 4.24% on the day, while the 10-year jumped to 4.68% and the 30-year spiked as much as 14bps to 5.23%—its highest level since 2007—before finishing the month around 5.27%.

While much of the commentariat framed the sell-off as a credibility warning for the new chairman, we would offer a different interpretation: Warsh appears to be deliberately weaponising the bond market to do his tightening for him. In America, the long end—not the cash rate—is the key transmission mechanism, with households fixing 30-year mortgages (now 6.66%, the highest in a year) and corporates terming out debt across the long curve. By sounding hawkish, maximising policy uncertainty and eschewing guidance, Warsh has engineered materially higher nominal and real yields across the curve—an outcome he explicitly claimed authorship of, observing that while the Fed had done little between meetings, the markets "have done quite a bit", describing the move as "a change for the better", and adding that "we're just getting started".

Crucially, the tightening has been delivered with inflation expectations intact: 10-year breakeven inflation rates sit at just 2.28%, with the sell-off decomposing almost entirely into higher real yields (around 2.4% at the 10-year tenor, a genuinely restrictive level) and rising term premia—the Fed's Kim-Wright estimate of the 10-year term premium reached 0.73% in July, while the New York Fed's ACM measure turned positive this year for the first time since 2023. This is real tightening with credibility intact. With the political calendar arguably constraining action ahead of the November mid-terms, futures price roughly 37bps of tightening by the December meeting, which looms as the first realistic window for the Fed to move.

Notwithstanding that late steepening, yields finished July substantially higher. The 10-year US Treasury yield increased 27bps to 4.73%. German 10-year Bund yields climbed 35bps to 3.21%, their highest level in more than a decade, while 10-year UK Gilt yields rose 29bps to 5.05%. French OAT yields increased 35bps to 4.00% and Italian BTP yields rose 39bps to 4.02%. Japanese 10-year government bond yields increased another 11bps to 2.79%, leaving them 124bps higher than a year earlier. French sovereign spreads were essentially unchanged against Bunds at 79bps, while the BTP/Bund spread widened 4bps to 81bps.

Central bank guidance varied across jurisdictions. The ECB left rates unchanged, as expected, although President Lagarde revealed that some governors had considered whether another hike was warranted, which reinforced expectations for a September increase. The Bank of Japan also maintained a hawkish tone, having lifted its policy rate to 1.00% in June, whereas the Bank of England was more dovish, with Governor Bailey pushing back against suggestions that the Bank was necessarily moving towards another hike.

Japan attracted particular attention late in the month after the yen weakened to a 40-year low of ¥163.7 against the US dollar. Japan intervened on 30 July, followed by a coordinated yen-buying operation with the US Treasury on the final trading day—the first joint US-Japan intervention to support the yen since 1998. The move reversed the currency's slide, with the yen finishing July as the strongest G7 currency: USD/JPY declined 3.2% to 157.40. The euro also appreciated against the US dollar, increasing 0.9% to 1.153.

Strategy commentary cont'd: The same uncertainty was reflected in global equity markets, where performance diverged sharply by region and sector. A reassessment of stretched AI-related valuations weighed heavily on technology stocks, with the Philadelphia Semiconductor Index declining 20.6%—its worst month since 2002 and a fall of some 25% from its June high—while the Nasdaq 100 returned -6.6%. The broader S&P 500 was effectively flat, returning -0.1%, as stronger performance outside technology offset much of the weakness in the largest growth stocks. Indeed, the rotation beneath the surface was striking: the S&P 500 Equal Weight Index gained 1.0% and the Dow Jones Industrials rose 0.6%, with the equal-weight index outperforming the Nasdaq 100 by 7.5 percentage points—its best month of relative performance on record.

Japanese and South Korean equities were also hit hard, with the Nikkei 225 returning -8.1% and the KOSPI declining 22%, its steepest monthly fall since the global financial crisis. The Korean drawdown exceeded 30% from the June peak intramonth, triggering seven circuit breakers, before a record 18% single-day rebound on the final trading day; Samsung Electronics fell 21% and SK Hynix 35% over the month. European markets were considerably more resilient. The Euro Stoxx 50 returned 0.6%, the Euro Stoxx Banks Index gained 5.6% and the FTSE 100 returned 3.6%. Gold increased a comparatively modest 1.0% to US$4,046/oz, while bitcoin rallied 7.3% to US$62,899.

Credit markets also softened, although the moves remained relatively orderly given the magnitude of the rates and equity volatility. In synthetic indices, US CDX IG widened 2bps to 53bps and CDX HY widened 7bps to 312bps. European iTraxx Main widened 2bps to 54bps, Xover widened 16bps to 261bps and senior financials moved 1bp wider to 56bps.

Cash credit again showed some regional divergence. US investment-grade corporate spreads widened 4bps to 78bps, reflecting in part the heavy weight of large technology hyperscalers in the US index and continued issuance from the sector. By contrast, euro aggregate corporate spreads tightened 1bp to 78bps, supported by much more limited summer issuance and strong investor cash balances. Sterling corporate spreads widened 2bps to 88bps.

The sharp rise in government yields meant duration was the dominant driver of corporate bond returns. The Global Aggregate Corporate Index hedged to US dollars declined 1.30% in July, whereas its duration-hedged equivalent gained 0.51%, illustrating the significant drag from higher risk-free rates.

US investment-grade primary issuance remained very active despite the increase in market volatility, with US$147bn of supply during July taking year-to-date issuance to roughly 35% above the equivalent period last year. The most notable transaction was Amazon's US$25bn eight-tranche deal. While the size was comparable with the hyperscaler transactions seen in June, demand was substantially softer: books were approximately 1.6 times covered compared with around three times for similar transactions the previous month. This was an early indication that investor appetite for the extraordinary volume of AI-related debt supply is no longer unconditional.

US Financials also returned to the market following earnings. We participated selectively in transactions from MUFG, Goldman Sachs, JPMorgan and Morgan Stanley where pricing generally offered modest concessions to our estimates of fair value. Several of these bonds subsequently performed well, particularly where broader secondary-market conditions provided an additional tailwind.

European issuance slowed materially as the summer period took hold, with only €34bn of investment-grade supply. The scarcity of new bonds left investors with substantial cash to deploy, helping financial transactions achieve average subscription levels of around four times. Deals from Societe Generale, Lloyds and Deutsche Bank were particularly well supported and subsequently performed positively.

TenneT Germany was the standout European corporate transaction. The newly established issuer, following its separation from the wider TenneT Group, attracted strong demand for its €3.5bn multi-tranche transaction, with the bonds rallying between 6bps and 10bps after pricing. By contrast, AT&T's euro and sterling transaction struggled despite offering relatively sizeable new-issue concessions, reflecting weaker investor sentiment towards the sector.

Strategy commentary cont'd: Australian primary markets also remained active despite the renewed geopolitical volatility. Approximately A$15.5bn of investment-grade credit was issued during July, around 50% more than the circa A$10.3bn issued in July 2025. This took year-to-date supply to approximately A$115.4bn compared with A$92.8bn at the same point last year.

Major-bank issuance resumed during the month, with NAB and CBA raising a combined A$5.25bn through 5-year and 3-year senior FRN transactions respectively. The deals attracted approximately A$9.4bn of combined demand, equivalent to average subscription of around 1.8 times.

Covered-bond issuance was another feature of the Australian market. Canadian banks CIBC and RBC issued a combined A$4.5bn of 5-year Kangaroo covered bonds against approximately A$7.1bn of demand. Nationwide also issued A$550m of 5-year senior preferred FRNs, attracting 3.4 times subscription.

Corporate supply resumed through CDC, which issued A$700m across 6-year and 10-year fixed-rate senior bonds. Both tranches attracted very strong support, with approximately A$4.4bn of demand equating to subscription of around 6.3 times. Semi-government syndicated supply totalled another A$6.0bn through transactions from QTC, SAFA and NSWTC.

Australian government bonds sold off alongside global rates markets. The 10-year Australian government bond yield increased 21bps to 4.93%. Credit was more mixed. Five-year major-bank senior spreads tightened marginally by 0.4bp to 62.7bps, while major-bank subordinated spreads were essentially unchanged at 119.9bps. Major-bank hybrid spreads, however, widened sharply by 22bps to 161bps, while the Australian iTraxx Index widened around 3bps to 71bps.

Domestic fixed-rate benchmarks were weighed down by the increase in yields, with the AusBond Composite Index declining 0.43% and the AusBond Credit Index falling 0.06%. Floating-rate credit remained more resilient, with the AusBond Credit FRN Index gaining 0.43%. The Australian dollar appreciated 1.5% against the US dollar to US$0.7019, while Australian equities performed strongly, with both the ASX200 price and total-return indices increasing 2.3%.

New Zealand markets similarly experienced a sharp increase in bond yields, with the 10-year New Zealand government bond yield rising 31bps to 4.48%. The New Zealand dollar appreciated 3.5% against the US dollar to US$0.5879. Equities were modestly positive, with the NZX50 price index gaining 0.5% and its total-return index rising 0.6%.

Overall, July was a materially more challenging environment for duration and technology-related risk. The combination of renewed geopolitical tensions, higher energy prices and a global sell-off in sovereign bonds placed pressure on longer-duration assets, while the sharp reassessment of AI valuations exposed increasing differentiation within equity and credit markets. Against this backdrop, we remained selective in primary credit, favouring transactions with clear valuation support and reducing exposure to heavily supplied areas where investor demand and subsequent secondary-market performance appeared less dependable.

Please note that past performance is not a reliable indicator of future performance. Investors should read the relevant Product Disclosure Statement and Target Market Determination before making any investment decision and consider obtaining advice from an independent financial adviser to determine whether an investment is appropriate for their objectives, financial situation and needs.

The tech boom is boosting US inflation

Tech retail prices in the US have swung from falling over recent decades to posting large gains, with the turnaround under way well before AI started to take off. Tech has switched from consistently subtracting 0.25pp from annual core inflation to adding 0.25pp. This boost is likely overstated because of measurement problems with software, but there has been upward pressure on inflation from every type of tech good and service. The Fed might argue for keeping rates steady because the tech boom should ultimately be deflationary, but the Fed still has a problem with high non-tech inflation, which is also tracking well above the Fed's 2% target.

Strategy commentary cont'd: Core inflation in the US – as measured by the core PCE deflator – has picked up from a low early last year of 2.6% to 3.4%, which is well above the Fed's 2% target.

Some of this pick-up reflects the lagged impact of tariffs on goods prices, although this effect should soon fade in the monthly figures.

The tech boom, which has been turbocharged by immense investment in AI, should ultimately be a deflationary influence on the US economy depending on how AI boosts productivity.

In the short term, though, the boom has directly contributed to higher inflation, with tech switching from being a longstanding constraint on inflation to adding to price pressures.

Calculating the direct impact of tech on core inflation, tech can be approximated by grouping a range of goods and services, including software, streaming, internet access, computers, tablets, phones, and AV equipment.

On this basis, tech accounts for about 6% of all consumer spending, excluding food and energy, with the share relatively stable except for a spike during the pandemic.

In aggregate, estimated tech prices are up almost 5% over the past year, which compares with average annual declines of about 3.5% during the 2000s and 2010s.

While the AI boom has accelerated the recent surge in tech prices, prices were turning around during the pandemic, well before the public release of ChatGPT in late 2022.

The large turnaround in prices has caused tech to switch from subtracting an average of 0.2pp from annual core inflation over the 2000s and 2010s to adding 0.3pp.

More recently, this means that tech prices have accounted for about one-third – or 0.3pp – of the 0.8pp increase in core inflation from last year's low of 2.6% to 3.4%.

More broadly, it also means that if tech prices were falling like they did in past decades, core inflation would be around 2.9% instead of 3.4%.

Strategy commentary cont'd:

However, it is worth noting that the increase in tech prices – and hence their contribution to core inflation – is likely overstated. This is mainly because the Bureau of Labor Statistics, which supplies the underlying data to the Bureau of Economic Analysis, has difficulty adjusting for quality improvements when calculating hedonic estimates of price changes.

Another issue relates to accidentally capturing huge AI-driven price gains for flash drives and related media that are not included in consumer spending.

The Fed has analysed these measurement problems for the price of software and accessories, which, on our calculation, has increased by 15% over the past year, accounting for 0.2pp of tech's estimated 0.3pp contribution to core inflation.

The Fed concluded that, "a quarter to well over a half of the contribution of the computer software and accessories category to core PCE inflation may be imputed to measurement error."

If this is correct, then tech might have only added 0.1-0.2pp to core inflation over the past year, such that "true" core inflation might round to 3.3% instead of the published rate of 3.4%.

Even so, the Fed's focus on software, while understandable, overlooks the fact that there has been strong upward pressure on all tech prices. This is clear from comparing the current change in prices for each series with growth prevailing prior to the pandemic.

Strategy commentary cont'd:

While the pressure on tech prices should abate when the current boom inevitably peaks – after all, investors seem unlikely to achieve an adequate return on the massive investment in AI – it should be noted that there is still upward pressure on non-tech inflation.

On our calculation, estimated non-tech core inflation has picked up from a low of 2.8% last year to 3.3%. This should ease when the impact of tariffs on goods prices finally fades over the rest of this year, but it still seems that, regardless of the exact direct contribution to inflation from the tech sector, the Fed faces a difficult job in sustainably returning inflation to the 2% target.