AZ Sestante Chief Investment Officer Andrea Ciaccio reviews markets for July 2026 — from an abrupt reality check on the funding behind the AI build-out to the washout of the year's most crowded trades and a Federal Reserve that talked hawkishly but still refused to move. Read the full commentary below, or download it as a PDF.
Download the full commentary (PDF)
By Andrea Ciaccio, Chief Investment Officer, AZ Sestante — as at 31 July 2026
Market Review
In July, credit and equity investors began to have second thoughts about AI, questioning both the explosive funding required to build out the technology and whether the rally in the so-called “bottleneck” stocks had run too far. Indications of structural fatigue emerged on the 7th, when Amazon entered the corporate bond market seeking $25 billion to fund capital expenditures but encountered softer institutional demand than earlier in the year. The offering attracted approximately $41 billion in orders, resulting in only 1.6 times coverage, well below the nearly 5 times achieved by hyperscaler bond sales in the first quarter and Amazon’s own 3.4 times on its $37 billion March raising. Two days later, S&P Global Ratings downgraded Oracle’s long-term credit rating to BBB-, leaving the firm just one notch above speculative grade status. The decision followed an aggressive AI infrastructure investment campaign that pushed Oracle’s total debt to $129.5 billion. The company faces additional vulnerabilities from its reliance on complex financing structures and a highly concentrated customer base, with nearly half of its $638 billion backlog reportedly tied to OpenAI. A further one notch downgrade would make the firm the largest “fallen angel” in corporate history, forcing many funds tracking investment grade benchmarks to sell bonds that would no longer qualify under their mandates. The resulting supply shock could be particularly disruptive as Oracle is an unusually large and long-dated issuer for a potential high yield entrant. Its debt portfolio carries an estimated effective duration more than twice that of the approximately 3 to 4 years for the broader US junk bond market. Moreover, assuming it were fully absorbed into the high yield universe, it would instantly amount to approximately 9% of the entire market.
OpenAI’s apparent decision to delay its IPO until 2027 added to the deterioration in the AI-market sentiment. The ChatGPT creator had been preparing for a potential listing in the back end of 2026 and had confidentially filed with US regulators in June, with advisers considering a valuation of as much as $1 trillion. Choosing to wait instead may signal diminished confidence that public investors are willing to validate the lofty valuations assigned to frontier AI labs in private markets. OpenAI’s latest financing valued the business at $852 billion, meaning a listing at the upper end of the contemplated range would require stock market participants to accept a further 17% premium. Scrutiny also intensified around the quality of earnings across the technology sector. On July 22nd, Tesla reported $1.11 billion of GAAP net income for the second quarter. Yet roughly two thirds of that sum stemmed from a $1.01 billion unrealised gain on its $2 billion investment in SpaceX, which amounted to $763 million after tax. This “paper profit” reflected a 50.2% increase in the value of the electric car maker’s investment in just three months. Excluding this item and a further $274 million in tax benefit, Tesla would have generated only about $77 million of net income. On the same day, Alphabet revealed its first quarterly negative free cash flow since going public in 2004, as capital expenditures surged to $44.9 billion. Moreover, the tech behemoth raised its 2026 capex forecast to $195-205 billion, implying that the Q2 deficit was likely to persist as the company accelerates spending on data centres, servers and AI systems.
Tensions continued to build throughout July, ultimately culminating in a clearing event on the 30th, when Ken Griffin’s Citadel, one of the world’s largest high-frequency trading firms, acquired the bulk of the public equity portfolio held by AI-focused hedge fund Situational Awareness at a discount. The fire sale followed a 67% drawdown at the alternative asset manager amid a severe selloff in semiconductor stocks. Situational Awareness was not shut down, but retained its private investments, including a multibillion-dollar stake in Anthropic. The fund is run by Leopold Aschenbrenner, a German-born AI researcher and investor who became prominent after graduating as valedictorian from Columbia University in 2021 at the age of 19, before joining OpenAI’s Superalignment team in 2023. He gained wider attention through his June 2024 essay “Situational Awareness: The Decade Ahead”, which argued that the US needed to prepare for the economic, geopolitical and national-security consequences of rapidly advancing AI capabilities. Aschenbrenner then established the identically named hedge fund and pursued concentrated investments tied to the AI boom. By July 2026, its assets under management had swollen to roughly $45 billion. Its strategy combined long positions in AI-hardware and semiconductor companies with short positions in software stocks, giving the fund substantial exposure to both the winners and the losers of the technological disruption. Banks allowed the fund to borrow roughly four to five times its capital, dramatically amplifying both its gains and losses. That leverage helped propel returns to 439% in H1 2026 and more than 1,000% since inception in September 2024. However, when the trend reversed in July, the fund was forced to liquidate positions into a falling market, eventually reducing its assets to $10 billion at the point of the unwind.
International Equities
US equities were mixed in July. The Dow Jones Industrial (+0.32% in USD terms) broke to new all-time highs, the S&P 500 (-0.13%) was broadly unchanged and the Nasdaq 100 (-6.61%) fell sharply. Weakness in the technology-heavy average was driven by chipmakers, with the Philadelphia Stock Exchange Semiconductor Index tumbling 20.77%, its 10th worst month in history. Conversely, the “Magnificent Seven” group of stocks staged a powerful rebound, outperformance the rest of US large caps by more than 400 basis points. Microsoft, Amazon and Apple led the gains while Google and Nvidia ended the month in the black after recovering from an earlier selloff. On July 1st, Meta announced plans to monetise its AI infrastructure externally, effectively treating idle GPUs like empty airline seats by renting “excess” compute to outside customers. The move comes even as the company expects to spend as much as $145 billion on capital expenditures in 2026. Although the stock closed only marginally lower, the prospect of this new business unit was interpreted as a competitive threat to the so-called “neo-cloud” providers, prompting a crash in CoreWeave (-27.90%) and Nebius (-31.05%). The speculative frenzy in other “hot” areas of the market also came to an abrupt halt. The so-called “RAMageddon” names Micron and Sandisk cratered 28.70% and 46.57% respectively, the GS Non-Profitable Tech and the Morgan Stanley US Momentum Long Basket both recorded their 3rd worst month ever and the S&P 500 High Beta Index underperformed the S&P 500 Low Volatility Index by the most since March 2020. The Russell 2000 (-3.08%) gave back part of its recent gains vis-à-vis the Russell 1000, but smaller companies remained comfortably ahead on a year-to-date basis.
Six sectors finished July in positive territory, led by energy, which rose double digits, followed by financials, healthcare, consumer staples, real estate and communication services. Technology was the worst performing group, relinquishing its leadership since the beginning of the year despite remaining up more than 30% through July. US equities modestly trailed the rest of the world (as exemplified by the MSCI AC World Index ex USA TR Index) amid a rotation from growth to value, with the latter factor recording its 9th best month ever in relative return against the former. Europe was the best performing region, followed by Japan. In the Land of the Rising Sun, Kioxia Holdings, a manufacturer of semiconductor memory products, halved in just a month since becoming the nation’s most valuable company. Emerging markets were dragged lower by Taiwan and South Korea. The KOSPI index crashed a record -43.93% from its peak on June 19th to its (so far) bottom on July 29th, with Samsung Electronics and SK Hynix bearing the brunt of the selloff. The decline was exacerbated by a surge in retail speculation through newly launched single-stock leveraged ETFs tied to the two semiconductor giants. On July 31st the index rebounded +17.91% in a single session, its largest one-day gain ever, as volatility remained elevated. Amid the turmoil, flows returned to the CSI Overseas China Internet, which tracks the performance of the largest Chinese Big Tech firms. All in all, the global benchmark was flat in USD terms but down -1.55% in AUD terms.
Australian Equities
Australian equities rose in July, with the MSCI Australia Net TR beating the MSCI AC World Daily TR by more than 400 basis points and interrupting a four-month losing streak. The S&P/ASX 300 was up +2.13% on the back of the solid gains posted by energy and financials. The former tracked the price of crude higher, as the West Texas Intermediate (WTI) jumped 21.83% for the month. The latter were buoyed by the rebound in banks, with investors snapping up shares in the Big Four despite early signs of softer housing credit. Healthcare outperformed the general index as positive momentum in CSL persisted. Technology was adversely affected by the rotation away from AI-related names, although software names fared better. A weak iron weighed on miners. Finally, the Top 20 trounced mid-caps and smaller companies, which remained under pressure from resource exposures.
International Fixed Income
Five of the six major global central banks that met during the month left interest rates unchanged. The Royal Bank of New Zealand (RNBZ) was the outlier, raising the Official Cash Rate (OCR) by 25 basis points to 2.50% on July 8th, its first increase since May 2023. One week later, the Bank of Canada (BOC) kept its benchmark overnight rate steady at 2.25%, extending its streak of consecutive holds to six. The European Central Bank (ECB) maintained its policy rate at 2.25% on July 23rd, pausing after its previous meeting in June, when it delivered its first 25 basis points increase since September 2023. The Bank of England (BOE) followed suit on July 29th, maintaining its official bank rate at 3.75%. The Bank of Japan (BOJ) held steady at 1.00% on July 31st, opting not to reinforce coordinated intervention by Japan’s Ministry of Finance and the US Treasury aimed at supporting the yen.
Kevin Warsh’s press conference following the July 29th FOMC meeting was the main event of the month. The FED paused for the fifth time in a row, leaving its target rate at 3.50%-3.75%, with three dissenters warning that delaying further hikes could require more aggressive action later. The Chairman reiterated that the central bank’s target is 2% and rejected the notion of a higher “implicit” threshold that policymakers are willing to tolerate. Yet, despite his hawkish rhetoric, he refused to adjust the target rate and did not provide any justification for this decision. He did not offer the market a conventional roadmap for September or the meetings beyond it, nor did he provide a description of the institution’s likely reaction function. Instead, he doubled down on his rejection of forward guidance stating that markets should “play the ball, not the referee” and suggested that the recent rise in Treasury yields showed that investors were already responding more directly to economic conditions. Adding to the uncertainty, he described his upcoming Jackson Hole speech as a “blank sheet of paper”, drawing criticism from economists who argued that his communication had undermined the FED’s credibility.
Yield curves around the world transposed higher, with the Bloomberg Barclays Global Aggregate Index hedged back to AUD dropping 0.94% in July. In the US, the 10- and 30- year yield ended the month at their highest level since October 2023 and June 2004 respectively. The so-called “term premium”, which is the compensation required to buy and hold long dated Treasuries as opposed to simply rolling over short term paper at prevailing rates, pushed to new highs for this cycle at 84 basis points.
Australian Fixed Income
Australian fixed income curbed losses, as the Bloomberg AusBond Composite 0+ Yr declined 0.43% for the month, outperforming global fixed income for the fourth time in the last 5 months. Domestic data were mixed. The housing market showed clearer signs of a broadening downturn in June, with national home values falling 0.4%, the largest monthly decline since December 2022. The weakness was concentrated in Sydney and Melbourne, while Brisbane and Perth continued to record monthly gains and Adelaide was broadly flat. Later in the month, the Australian Bureau of Statistics (ABS) reported that headline CPI moderated to 3.8% over the year to June 2026 and was negative 0.1% QoQ. Moreover, the CPI Trimmed Mean was stable at 3.6% YoY and 0.8% MoM, missing expectations of 3.7% and 0.9% respectively. In contrast, the labour market remained strong in June, with employment rising by 76,300, more than five times estimates, following a revised 44,000 increase in May. The unemployment rate held at 4.4%. At the end of the month, cash rate futures were pricing in only one further hike, with a 60% probability of the move occurring in Q2 2027. The Australian Dollar strengthened across the board, rebounding above 70 cents vis-à-vis the greenback.
Real Assets
Global property rose 2.94% in USD terms and 1.27% in AUD terms in July, with Asia emerging as the best performing region, led by Hong Kong. The UK continued its recovery, supported by renewed investor interest following Prologis’ proposed acquisition of FTSE 100-listed Segro. Australia was the weakest performing region, weighed down by index heavyweight Goodman Group.
Global listed infrastructure gained +0.39% in USD terms but was down -1.24% in AUD terms for the month. Transportation stocks performed best, with North American rails supported by optimism surrounding the proposed merger between Union Pacific and Norfolk Southern. Electric and gas utilities lagged the general index on concerns about affordability.
Alternatives
Returns for Alternatives (-0.32%) were generally negative across strategic mandates in July. Long/Short Equity managers were driven lower by technology as the Goldman Sachs Hedge Fund VIP, which tracks the performance of the most popular long positions, recorded its 2nd worst relative monthly return ever vis-à-vis the Goldman Sachs Hedge Fund Very Important Shorts.
Market Outlook
As the market entered July with equity positioning unusually one-sided, the unwind was correspondingly sharp. However, rather than a generalised loss of confidence, we experienced a genuine washout of crowded trades in momentum and AI-related names. Once positions that had become excessively concentrated repriced, correlations started to normalise and volatility came down. After an initial spike, the VIX, which measures the expected volatility of the S&P 500, ended July almost unchanged from a month before. Moreover, after four consecutive weeks of decline, by mid-August it was hovering at its lowest level of 2026. The market structure has thus become healthier, even as the macroeconomic backdrop is becoming more mixed, with fiscal and corporate support appearing to fade at the margin. Tariff-related reimbursements, which ultimately flowed through to companies’ bottom lines and proved to be a windfall for the corporate sector rather than for consumers, are largely behind us. Likewise, the fiscal impulse from government transfers remains massive but is no longer accelerating at the same pace. Capital expenditure remains exceptionally elevated, but its rate of change is slowing as well. For these reasons, we are open to the possibility that the US economy will cool off from here. However, we are not sounding the alarm yet as a possible “soft patch” is not a recession.
GDP growth in Q2 was not particularly strong although domestic demand proved resilient. The headline figure was distorted by chip imports, which are contributing to robust economic activity in North Asia. The early GDPNow estimate for Q3 produced by the Federal Reserve Bank of Atlanta was recently at +4%, but we expect it to move substantially lower as new data arrives. The labour market has been exhibiting for quite some time what Jerome Powell described as a “low-hiring, low-firing environment”, with companies not conducting large-scale layoffs, but neither hiring aggressively. If employment remains mediocre and inflation behaves over the next two months, following the recent benign readings, it becomes increasingly difficult to make a convincing case for the FED to hike. And yet, the 2- year yield is above the upper end of the central bank's current target range, implying that traders continue to price a meaningful risk of further tightening. At the same time, the 10- year yield is at 4.71%, leaving the difference between the short end and the long end at only about 50 basis points. In a conventional interpretation, such a compressed spread at this stage of the cycle suggests that the monetary policy is not restrictive enough and that the long end needs to move higher.
We posit that the markets have it backwards. The problem is not that the 10- year is too low, but that the 2- year is too high, and that is the result of the FED continuing to run an enormous balance sheet which keeps financial conditions easier than the policy rate alone would suggest. The central bank holds $6.7 trillion in assets, including a securities portfolio with a weighted-average maturity of close to seven years. Since the beginning of 2026 it has also been purchasing additional T-bills through the so-called “Reserve Management Operations” to maintain bank reserves. The institution has effectively withdrawn a very large amount of duration from the private market, that is, it has suppressed one of the central risks underpinning the valuation of financial assets. The result has been the emergence of a “K-shaped” economy in which the “haves” benefit disproportionately from the unimpeded appreciation of equities and record corporate earnings, keeping household consumption resilient and corporate investment elevated. Meanwhile, the labour market gradually loses momentum, with workers capturing an increasingly smaller share of the economic pie. The US economy can therefore continue to run relatively hot, and inflation persists even when the underlying rate of employment growth is slowing.
If the FED were to reduce its balance sheet meaningfully, financial conditions would tighten without requiring a higher policy rate. The transmission would run through asset prices. Equities would likely decline, particularly the parts of the market that have benefited most from abundant liquidity and high valuations. The so-called “reverse wealth effect” would weigh on consumption and prompt a revision of aggressive AI capex plans. In other words, balance sheet reduction would slow the economy through several channels at once, that is, lower equity prices, weaker consumption, tighter credit and eventually slower corporate investment. That would provide room for the FED to cut rates and the 2- year to fall while the 10- year would not decline by anything like the same magnitude, bull steepening the curve. The reason is that the long end is increasingly anchored by structural forces pushing the cost of capital higher. The US government is running a structural fiscal deficit of $2 trillion outside of a recession, Europe is increasing spending to rebuild its defence capabilities and Japan is attempting to stabilise its currency while normalising monetary policy. In addition, the private sector is demanding enormous amounts of capital for AI infrastructure, power generation and other forms of strategic investment. As the world is competing for the same finite pool of investable funds, we have moved from an era characterised by a global “savings glut” to one defined by “capital scarcity”.
This is what, in our opinion, Kevin Warsh meant at the latest FOMC meeting when he cryptically stated that “market prices […] have tightened financial conditions in the intermeeting period”. The (unsaid) implication was that, with the long end of the curve already moving higher, the FED could refrain from hiking. Unfortunately for the FED Chairman, Scott Bessent has other priorities, and one of them appears to be putting a lid on the 10- year. Following the joint US-Japan intervention to support the Japanese Yen (JPY) at the end of July, the US Treasury Secretary called for an expansion of the Foreign and International Monetary Authorities (FIMA) repo facility, allowing Japan to obtain dollars from the FED against Treasury collateral rather than selling its roughly $1.14 trillion holdings into the market. The latest Quarterly Refunding Announcement (QRA) points in the same direction, with T-bills expected to account for 61% and 58% of total borrowing in Q3 and Q4 respectively, a “dovish” financing mix from a duration perspective. Put differently, the US government needs to borrow substantially more, but it is choosing not to put the incremental supply into the long end. More recently, Bessent surprised markets by doubling the size of certain long duration buyback operations to at least $4 billion per clip after the 30- year yield reached 5.34%.
The absolute size of the intervention is too small to fundamentally alter the supply-demand (im)balance of the Treasury market, but it demonstrates the willingness to manage the long end when yields become disruptive. We are sceptical that the FED will aggressively pursue balance sheet reduction, despite the more hawkish rhetoric surrounding the new regime. Hence, if Bessent successfully caps the long end while the central bank maintains a bloated portfolio of assets, the economy could remain relatively hot. The result would be renewed pressure on inflation and, consequently, on the 2- year. The market would continue to price hikes but, if the central bank refused to deliver them, the tension would migrate back to the long end. And if the Treasury then succeeded in containing that part of the curve, the pressure would once again shift to the short end. Eventually, as US rates are being held artificially low relative to the inflation and fiscal backdrop, the adjustment happens in the currency market. A weaker greenback becomes the release valve and the so-called “debasement trade” makes a comeback, as investors hoard real assets to participate in the rise of nominal wealth while real purchasing power falls behind over time. That is why we do not believe the recent rise in long term yields, by itself, is a sufficient reason to reduce our overweight exposure to growth assets.
At the moment the probability of a FED hike is declining as inflation appears manageable in the immediate future, while Bessent is sending a threatening message to the mythical “Bond Vigilantes”. In this context, nominal growth can remain relatively solid and corporate earnings can continue to expand. Moreover, hundreds of billions of dollars can be mobilised for AI infrastructure, extending the investment cycle further. For these reasons, equities remain the preferred asset class and an internationally diversified portfolio appears to be the most appropriate choice. Our positioning continues to favour a more cyclical tilt, with an emphasis on markets outside the US, including emerging markets, and an allocation to global small cap equities. Interest rates remain the principal risk, and whether policymakers ultimately allow the bond market to impose the tightening that the economy may require is the most important question. The equilibrium is precarious, but it still holds. We would rather own equities while it does than try to anticipate the precise level at which the bond market finally breaks them. Accordingly, in fixed income we continue to expect more choppiness ahead despite improved valuations. At this stage bonds provide a sufficiently attractive carry to warrant exposure, but no optionality in terms of capital appreciation. Not yet, at least.
Download this commentary as a PDF
AZ Sestante is a specialist investment consultant focused on designing and managing a range of multi-manager model portfolios via SMAs, MDAs, and fund of funds. Our parent company Azimut is Italy’s largest independent asset manager listed on the Italian stock exchange. E: [email protected] | www.azsestante.com
Important information
The Morningstar Historical Corporate Sustainability Score is a weighted average of the trailing 12 months of Morningstar Portfolio Corporate Sustainability Scores. Historical portfolio scores are not equal-weighted; rather, more-recent portfolios are weighted more heavily than older portfolios. Combining the trailing 12 months of portfolio scores adds consistency while still reflecting portfolio managers' current decisions by weighting the most recent portfolio scores more heavily. ESG pillar scores are displayed as a number between 0 and 100 with most scores range between 0 and 25. It is the asset-weighted average of the company environmental, social, governance risk scores for the covered corporate holdings in a portfolio. The scores measure the degree to which a company's economic value may be at risk driven by environmental, social, and governance factors. The risk represents the unmanaged risk exposure after taking into account a company's management of such risks.
This document has been prepared by AZ Sestante Limited, ABN 94 106 888 662, AFSL 284 442 (AZ Sestante). This document is not an offer of securities or financial products, nor is it financial product advice. As this document has been prepared without taking account of any investors' particular objectives, financial situation or needs, you should consider its appropriateness having regard to your objectives, financial situation and needs before taking any action. Past performance is not a reliable indicator of future results. Although specific information has been prepared from sources believed to be reliable, we offer no guarantees as to its accuracy or completeness. The information stated, opinions expressed, and estimates given constitute best judgement at the time of publication and are subject to change without notice. Consequently, although this document is provided in good faith, it is not intended to create any legal liability on the part of any other entity and does not vary the terms of a relevant disclosure statement. All dollars are Australian unless otherwise specified.