• After a volatile July and start to August, global equity markets ended the month little changed in sterling terms
• Despite the uncertain macro backdrop, AI data points remain constructive, including investment from the larger cloud providers and early examples of AI use cases
• Increased volatility is typical of early phases of a new technology cycle, and we expect AI progress to continue through 2025
  

Market review

Global equity markets ended August in positive territory, the MSCI All Country World Net Total Return Index returning +0.3%, while the S&P 500 and DJ Euro Stoxx 600 indices returned +0.2% and +1.6% respectively (all returns in sterling terms).

Solid headline index returns belied unusual market volatility early in the month, following weak US labour market data and significant yen strength after the Bank of Japan (BoJ) hiked its policy rate by 0.25% and announced plans to halve its purchases of Japanese government bonds. This prompted a sharp unwind of the yen carry trade (where investors borrow ‘cheaply’ in yen – as Japanese interest rates are lower – to buy other assets/currencies). Markets rebounded rapidly; however benign US macroeconomic data suggested further economic stability rather than a sharp slowdown, while a more dovish BoJ stance helped contain yen strength.

The early August volatility followed the Bureau of Labor Statistics report that the US economy added only 114,000 jobs in July, below forecasts of 179,000. In addition, the US Consumer Price Index (CPI) annual inflation rate moderated to +2.9% year on year (y/y) in July (below forecasts of +3%), down from +3% y/y in June, and the weakest monthly reading since March 2021. The combination of a softening labour market and tempering inflation trends solidified expectations for a September interest rate cut.

The minutes of the Federal Reserve’s (Fed) late July policy meeting showed broad agreement between committee members that “it would likely be appropriate to ease policy at the next meeting” in September, and came ahead of softer employment and inflation data in August. Towards month end, Fed Chair Jerome Powell delivered a speech at Jackson Hole that signalled US interest rate cuts were imminent. Although “the economy continues to grow at a solid pace”, Powell noted the Fed does not “seek or welcome further labour market cooling”, which explained why “the time has come for [monetary] policy to adjust”.

Technology review

The technology sector underperformed the broader equity market in August, with the Dow Jones Global Technology Net Total Return Index (W1TECN) returning -1.1%. Large-cap technology stocks outperformed their small and mid-cap peers; the Russell 1000 Technology Index (large cap) and Russell 2000 Technology Index (small cap) returning -1.1% and -4.7% respectively. The Philadelphia Semiconductor Index (SOX) returned -3.5%, while the NASDAQ Internet Index and Bloomberg Americas Software Index returned +0.7% and -1.1% respectively.

Despite a less certain macro backdrop, earnings reports have remained robust, albeit with cyclical weakness in some areas offset by continued AI strength. This is underpinned by robust capital spending on key infrastructure builds by the hyperscalers (the largest cloud service providers).

In the semiconductor sector, NVIDIA once again exceeded market expectations as its core data centre GPU (graphics processing units) business grew a remarkable +154% y/y.

In the semiconductor sector, NVIDIA once again exceeded market expectations as its core data centre GPU (graphics processing units) business grew a remarkable +154% y/y. More importantly, CEO Jensen Huang also allayed investor concerns about further delays to the company’s next-generation Blackwell chip, with the company expecting “several billion dollars of Blackwell revenue” towards the end of the year, while also seeing increased demand for existing Hopper GPUs in the interim. In addition, NVIDIA raised its expectation of revenue from sovereign AI customers to low double-digit billions, up from high single-digit billions, as countries make strategic investments to control their own AI destinies.

Other semiconductor names posted strong results driven by AI infrastructure spending. Taiwanese server maker Quanta Computer delivered a solid beat of previous guidance with better than expected margins, and the company expects full-year AI server sales to increase by triple digits y/y. Server rail kit provider King Slide Works beat analyst expectations as demand for high-end AI server rails ballooned, seeing the company grow revenue 30% quarter on quarter (q/q) with gross margins close to 70%. Monolithic Power Systems’ Enterprise Data segment, which produces chips to power advanced GPUs, grew +290% y/y and the company also saw strength and a return to growth in more cyclical areas such as Storage and Computing. However, memory-related semiconductor firms including Disco, Micron Technology and Samsung Electronics were weak on memory pricing and customer inventory concerns. Intel* was especially weak following a dividend cut and a significant restructuring plan.

In semiconductor capital equipment, Advantest delivered strong operating profit to beat consensus expectations by 50%, as sales of system-on-chip (SoC) testers for AI applications reached new highs, while high-bandwidth memory (HBM) tester strength remained robust. Kokusai Electric also beat on sales and operating profit but failed to raise its full-year guidance.

In the internet sector, Meta Platforms (Facebook) delivered a strong set of results, growing 22% y/y with a 23% y/y increase in advertising revenue driven both by both price and impression growth; Q3 revenue guidance of +16% y/y implies continued sequential momentum. The company raised the low end of its capital expenditure (capex) guide from $35-40bn to $37-40bn, largely driven by AI infrastructure spend, and CEO Mark Zuckerberg expects the company’s generative AI (GenAI) assistant will become the most widely used in the market by year end. Trade Desk posted above consensus results, as connected TV outpaced all other formats as recent partnerships with companies such as Netflix, Disney* and Roku* begin to bear fruit.

Amazon saw growth in its cloud computing business, AWS, reaccelerate to +19% y/y and reach 35.5% operating margins, but backlog growth of +19% was somewhat disappointing. Amazon’s retail sales were only in line with expectations and profitability below, raising questions about the path of margin expansion. In contrast, Shopify delivered strong results as revenue grew +25% y/y (excluding the sale of its logistics business) and group operating margins reached 14.6%, c2% above consensus expectations. The Q3 guidance for low to mid-20% growth was also ahead of expectations.

Cybersecurity results remained a bright spot. Cloudflare revenue growth of +30% y/y was strong as the company benefitted from more customer additions, particularly customers >$100k. CyberArk Software also beat across all metrics and raised its full-year expectations to +24.5% y/y at the midpoint. Management’s tone was constructive, with deal velocity accelerating across all verticals and geographies despite a more challenging economic backdrop.

The broader software sector remains more challenging, in part because of uncertainty about the role of GenAI in reshaping the application landscape and perhaps some budget reallocation. We remain cautious on many software stocks. That said, HubSpot reported in line with expectations as various issues seen last quarter appear to be stabilising. The company’s multi-hub motion is gaining momentum and has AI features in customers’ hands at scale today. Elastic also reported a solid quarter, but reduced guidance due to deal slippage amid a sales reorganisation. More positively, search accelerated, driven by GenAI applications, and consumption patterns remain robust.

In hardware, Apple delivered a well-received quarter slightly ahead of expectations, growing +5% y/y on iPad and Services strength despite iPhone revenue being down -1% y/y. Our focus remains on the recent iPhone launch (9 September) and the rollout/take-up of Apple Intelligence. Apple has partnered with OpenAI to deliver the new feature set, but whether this drives a significant refresh cycle in Apple hardware remains unknown.

Elsewhere, Pure Storage delivered a steady quarter and reiterated its full-year guidance. However, the company is seeing some elongation of sales cycles, specifically for larger deals in certain product lines, and there were limited updates on the company’s hyperscaler opportunity. Public safety company Axon Enterprise reported another strong quarter, above consensus, driven by all product lines, growing +35% y/y – its 10^th consecutive quarter growing above 25%. The company’s AI product, Draft One, has already generated more than $100m of pipeline, the fastest of any of the company’s software products to date.

Outlook

After the surprisingly soft jobs report at the start of August, macro data has subsequently been solid. The inflation picture continues to improve which has changed the balance of risks for many central banks (including the Fed). As a result, focus has shifted from managing the risk of higher/sticky inflation to supporting economic growth and the labour market which has prompted many to shift into interest rate-cutting mode. We expect the speed and size of interest rate cuts to be determined by incoming macroeconomic data, with any signs of more pronounced labour market weakness likely to be met with steeper cuts.

Absent a US recession, rate-cutting cycles have been supportive of positive equity market returns. However, we are cognisant of challenging seasonality (encapsulating a Q3 pre-announcement season) and heightened geopolitical risk, particularly in light of a close-run US presidential election.

Turning to technology, we remain highly constructive largely because of the powerful secular tailwind from GenAI which we consider the next GPT (general purpose technology), as well as a rare example of discontinuous technology progress. While we understand fears that AI excitement will fade, we believe the opposite is true and that the longer-term potential of this highly disruptive technology will become much more visible as 2025 progresses.

In our view, GenAI will reshape every industry and we believe an understanding of technology-enabled change (and deep AI knowledge) will be critical to navigating the considerable disruption ahead and unlocking the broader AI opportunity.

In contrast with stock weakness, AI-related news flow has been overwhelmingly positive. Alongside strong results, Walmart* announced it “used Generative AI to improve our product catalog…to accurately create or improve over 850 million pieces of data and the catalog. Without the use of generative AI, this work would have required nearly 100 times the current headcount to complete and the same amount of time”.

Amazon CEO Andy Jassy also discussed how Amazon’s use of GenAI to help with Java upgrades “has saved us the equivalent of 4,500 developer-years of work” and $260m in annualised efficiency gains”.

An AI-powered natural language query engine built to sit in front of the International Monetary Fund’s massive body of economic data has delivered a 50% increase in research productivity and a 35% increase in research accuracy. Amazon CEO Andy Jassy also discussed how Amazon’s use of GenAI to help with Java upgrades “has saved us the equivalent of 4,500 developer-years of work” and $260m in annualised efficiency gains”.

These remarkable statistics point to the considerable productivity benefits we expect AI to deliver once it is more widely adopted. Another high-level way to think about the opportunity is that, given AI is a substitute (and potential replacement) for human knowledge work, AI changes the addressable market for technology from ‘technology spend’ (c3-6% of overall enterprise revenues) to a large portion of total ‘operational expenditure’ (c30-80% of revenues). We believe this transformation explains why leading technology companies are investing as aggressively as they are today.

Recent AI commentary suggests demand remains strong, corroborated by monthly sales data from the Asia supply chain. That said, material upside to expectations could be constrained ahead of NVIDIA’s Blackwell ramp up (starting in Q4), but we believe the focus should be on Q4 outlook and guidance on Q3 earnings calls. We expect strong demand associated with Blackwell to drive further strength in AI infrastructure spending through H1 2025 which we expect to reignite interest in AI stocks. Additionally, significant developments in large language model capabilities are expected with Open AI’s GPT 5.0 anticipated in the next six months.

Near term, market gyration could persist for several more weeks, largely explained by normal seasonal factors, exacerbated by pre-election jitters, and increased volatility typical of early phases of a new technology cycle (in this instance triggered by NVIDIA’s Blackwell delay). While the macro backdrop is somewhat softer, we still believe it remains supportive of sustained strong Al investment. Recent off-quarter earnings and company meetings (many of the team have attended technology conferences this month) also support this view. As such, we see this period as an opportunity to accumulate Al exposure and refocus portfolios on the Al winners of 2025 and beyond.

* not held