Global Opportunities
Portfolio

Global Market Summary — July 2026

North America: Headline indices concealed a forceful internal rotation. The S&P 500 was essentially flat while the Nasdaq Composite fell 3.2%, yet the equal-weight S&P gained approximately 1.1%. Semiconductors suffered their worst month in more than two decades as investors rotated toward energy, financials and other value exposures. The Federal Reserve held rates at 3.50%–3.75% with three dissenters favouring an increase, while higher long-term yields and stronger oil prices reinforced the move away from long-duration growth.

AI Infrastructure: July delivered the first genuine de-rating of the infrastructure trade, but the market repriced valuation, ownership and financing rather than end demand. More than US$1 trillion of semiconductor value was erased as investors questioned the timing of returns on record capital expenditure and the durability of scarcity-driven margins. Hyperscaler spending continued to rise, but higher capex was increasingly treated as a free-cash-flow cost rather than an automatic demand signal. Power, grid and electrical-equipment exposures held up better than merchant compute and semiconductor layers.

Europe/UK: European equities advanced, with the STOXX Europe 600 up 1.3% and UK equities among the strongest developed markets, but sector dispersion remained extreme. Energy, banks and insurers led, while technology fell 7.2% and utilities declined despite higher wholesale power prices. Hawkish ECB and Bank of England holds, combined with a 35-basis-point rise in the 10-year Bund and a 29-basis-point increase in the 10-year gilt, created a direct valuation headwind for regulated and long-duration assets. Grid and electrical-equipment companies retained the strongest earnings visibility.

Asia-Pacific/Japan: Regional outcomes diverged more dramatically than anywhere else. The Hang Seng rose 13.1%, the Nikkei 225 fell 8.1% and the Kospi lost approximately 22% as Japanese and Korean semiconductor leadership unwound. Capital rotated into offshore Chinese technology and consumer equities on stimulus expectations and enthusiasm for China’s domestic semiconductor stack. Australia provided a quieter source of resilience, with the ASX 200 gaining 2.0% for a fourth consecutive positive month.

Macro and currency: July combined hawkish central-bank holds with higher long yields and renewed energy inflation. The Fed, ECB, BoE and BoJ all retained restrictive or tightening biases, while oil and European power prices rose on geopolitical and supply risks. Currency translation was material for AUD investors: AUD/USD gained 1.5%, reducing USD-linked returns, while AUD/JPY fell 1.6%, cushioning Japanese equity losses.

Global read: The common thread across all four sleeves was not fading AI-infrastructure demand but a repricing of who captures the economics and how those future cash flows are discounted. Long-duration semiconductor and merchant-compute exposures de-rated, while energy, financials, grid equipment, power infrastructure and selected Chinese platforms gained relative support. The portfolio implication is to favour active sleeve allocation, constraint-layer beneficiaries and selective risk overlays rather than uniform global beta.

Global Special Analysis Four Regional Solutions to the AI Power Constraint

North America interconnection and speed: Power delivery, rather than capital or generation availability, became the binding constraint. FERC’s June intervention marked a federal shift toward regulating large load interconnection directly, while gas and behind the meter generation gained share because they can be delivered faster and at lower connection cost. The key beneficiaries are grid equipment, transformers, switchgear, flexible generation and distributed power platforms.

AI Infrastructure the capex revenue gap: The central risk is no longer whether demand exists, but whether monetisation can catch up with spending before investors lose patience. Hyperscaler capex is rising far faster than directly attributable AI revenue, and debt issuance is increasing as companies bridge the gap between investment and free cash flow. The picks and shovels layer remains best supported by contracted spending, while the application layer still depends on unproven monetisation.

Asia Pacific/Japan abundance versus scarcity: China is solving the AI power problem from a position of generation and equipment abundance, while Japan is solving it from structural scarcity. China’s advantages include shorter transformer lead times and a dominant renewable equipment supply chain, but curtailment and coal dependence remain important caveats. Japan, by contrast, requires nuclear restarts, transmission investment and better alignment between generation locations and data centre clusters.

Europe/UK nuclear sovereignty: The regional response is increasingly centred on baseload sovereignty rather than speed alone. The UK’s structural power cost disadvantage is already affecting investment decisions, while SMR and nuclear projects are moving into active policy and capital formation channels. Brussels is pursuing a more regulation led path built around efficiency, disclosure and carbon neutral data centre standards.

Portfolio implication: The same AI demand shock is producing four distinct investment regimes. North America rewards speed to power; the AI sleeve rewards contracted infrastructure revenue; China rewards equipment scale and system abundance; Japan rewards nuclear, grid and automation; and Europe/UK reward sovereignty, efficiency and specialist engineering. The Global Opportunities structure is designed to capture that dispersion rather than assume a uniform global buildout.

Global Volatility Regime — July 2026

North America: Index volatility remained deceptively cheap. The VIX ended July at 15.99, below its 16.45 June close, despite a 3.4% S&P peak-to-trough decline and a 7.0% Nasdaq drawdown. The more important signal was the explosion in sector and single-stock dispersion. Broad index puts therefore provided limited protection for the risk that materialised, while semiconductor and high-beta single-name hedges carried much greater relevance.

AI Infrastructure: The semiconductor complex lost more than 20% from its June peak as merchant-compute, memory and chip exposures repriced sharply. Volatility was driven by monetisation, vertical integration and financing concerns rather than a collapse in demand. This favours targeted collars, put spreads and earnings-event protection on crowded compute exposures, while preserving upside participation in power, grid and electrical-equipment beneficiaries.

Europe/UK: Positive regional indices masked severe sector divergence between energy and financial winners and technology and utility losers. Rate volatility was the primary transmission channel as Bund and gilt yields rose sharply, while energy prices added asymmetric pressure to utilities whose hedging limited immediate earnings participation. Broad European index protection was therefore less efficient than selective downside structures on utilities, technology and other long-duration holdings.

Asia-Pacific/Japan: Realised volatility reached tail-event conditions. The Nikkei experienced a 12.8% peak-to-trough drawdown, Korea triggered a market-wide circuit breaker before an 18% rebound, and USD/JPY reversed sharply after reported intervention. Japanese equities, JGB yields and currency volatility are now tightly connected, requiring explicit protection around BoJ meetings, the 162–165 USD/JPY intervention zone and concentrated semiconductor exposures.

Global implementation: July exposed the weakness of uniform index hedging in a dispersion-led market. The preferred framework is multi-layered: retain selective index tail protection for systemic shocks; use single-name and sector collars or put spreads where concentration risk is highest; harvest elevated event premium only where balance-sheet risk is controlled; and integrate AUD/USD and AUD/JPY overlays when currency translation can materially change reported returns. The objective is targeted convexity against the risks actually owned, rather than indiscriminate long volatility.