Scarcity, Rivalry and the New Logic of Asset Allocation

An evolving market landscape: from a “soft” world to a “hard” world

The investment regime that generated returns for four decades is over. From the mid-1980s through to the pandemic, investors operated in a “soft world”, a rule-based world of frictionless globalisation dominated by soft power. That regime of falling inflation and declining interest rates generated extraordinary returns in financial assets, particularly long-duration growth equities and bonds. It rewarded capital-light business models and financial engineering.

We now live in a “harder” world defined by geopolitical rivalry and intensifying competition between countries, with Artificial Intelligence (AI) at its centre. Security and sovereignty have become the central preoccupation for governments. Capital intensity is back, and the competition for physical and financial resources has no precedent in the modern era. The investment implications are profound and durable.

This world of scarcity is defined by simultaneous pressure on three categories of resources:

Physical resources under strain. Demand from the energy transition, AI infrastructure and national defence is colliding with decades of underinvestment in extraction and processing capacity. Critical minerals face structural supply gaps that cannot be resolved quickly. Supply chains for these materials are concentrated in a handful of geographies, making access a matter of geostrategy as much as economics.

Technological resources as a chokepoint. AI has created a new form of scarcity: access to high-performance compute. Hyperscalers are spending close to $765 billion on capital investment in 2026 alone, 80% more than a year ago[1]. The ripple effects run through the entire AI supply chain, energy infrastructure and grid capacity. This is not a US story: as countries seek to secure sovereignty over AI, ecosystems are decoupling along geopolitical lines, making demand for compute infrastructure structural within each sphere of influence and global in scale.

Financial resources: competition for capital. Governments and the private sector are drawing on the same pool of global savings. US federal debt has risen from 55% to 124%[2] of GDP since the turn of the century, and major European sovereigns show a similar trajectory. German and Japanese 30-year bond yields, near zero just a few years ago, now approach 4%. For the first time in a generation, sovereign borrowers are competing with a private capex supercycle for global capital. The result is a structurally higher cost of capital.

This is not a cyclical rotation. It is a structural break, and portfolios need to adapt.

Equity allocation: where you invest matters less than what you own

Regional allocation is no longer the primary framework for equity outperformance. The divergence between winners and losers within sectors is widening, and stock selection has become an essential driver of returns.

The underlying dynamic is the shift from multiple expansion to earnings growth. In a world of higher interest rates and rising cost of capital, companies must earn their valuation through concrete earnings per share (EPS) delivery, not through inflation. The outperformance of hardware and semiconductors over software, which faces structural disruption from agentic AI, illustrates how quickly this dispersion can materialise.

Our equity investment themes are anchored in capex generation and its impact on EPS growth, organised around three pillars.

• AI infrastructure. The AI supply chain spans a remarkably broad range of industries from physical inputs such as critical minerals to the hardware layer covering semiconductor design, chip fabrication, packaging and memory, through infrastructure layer of data centres, model developers and cloud platforms. The structural case is intact, but the pace of investment is outrunning cash flows, and the market is entering a phase of digestion. Hyperscaler capex commitments through 2027 exceed $1 trillion,[3] financed increasingly by debt. Returns on this capital are not yet proven. In this context, the difference between the right name and the wrong name within the same theme matters as much as the theme itself. Dispersion within AI-exposed equities is at its widest since the emergence of large language models.
• Power generation and energy security. The AI buildout cannot exist without electricity, and the grid in most advanced economies was not designed to absorb demand at this scale. We see a multi-year investment cycle across the full power value chain, spanning grid equipment and automation, renewables and storage.
• Defence and national security. The defence theme rests on a structural and synchronised increase in military spending across major economies. Europe is engaged in a significant rearmament cycle, Japan and South Korea are committing to higher defence budgets, and the US is redirecting investment toward more cost-efficient platforms. This is not a short-term budget impulse but a durable response to a more contested geopolitical environment. The investment opportunity spans the full defence supply chain, from large diversified contractors and naval shipbuilders to specialised component manufacturers and defence IT providers.

Across these three pillars, we are positive on technology, materials, industrials, and utilities. These are the sectors where capex intensity is highest, earnings revisions are most positive, and the structural demand drivers are least dependent on the macroeconomic cycle. But this also implies a more cyclical sector allocation and potentially higher volatility. The concentration of performance within each theme itself is a risk: getting the sector right but the company wrong is increasingly costly.

The risk of a speculative excess exists, but the evidence does not yet support it. Total equity issuance including IPOs and secondary offerings remains below 1% of GDP[4], compared to over 2% at the peak of the 1990s bubble. Valuations, while elevated in some sub-segments, are anchored in a concrete earnings cycle rather than terminal value speculation. That said, the Bank of International Settlements in its Annual Economic Report published in June has drawn an explicit parallel with past investment booms, from railways to dotcom, where genuine technological progress attracted more capital than returns could ultimately justify. Markets will need time to digest the scale of commitments already made. The more relevant risk is not a bubble but a rotation: the market is transitioning from rewarding growth and dominance to rewarding returns on invested capital. Companies that cannot demonstrate those returns will face multiple compression regardless of sector.

Fixed income allocation in a higher-rate world

Interest rates are structurally higher. The combination of elevated sovereign debt competing with private sector capex for global savings, and the end of the deflationary tailwinds that drove rates lower for forty years point to an equilibrium cost of capital that is materially above the levels of the 2010s.

The most recent hawkish impulse, the market expectations of a new tightening stance by the Fed, the ECB’s rate hike, the Bank of Japan at its highest policy rate since 1995, may all prove temporary. The key variable is oil: the US-Iran memorandum of understanding has already brought prices below $80 per barrel[5] and is pulling breakeven inflation rates lower. If this persists through H2, the inflationary shock that prompted the hawkish pivot could dissipate. We are not buyers of a new tightening cycle, but neither are we forecasting a return to zero.

Duration positioning

We remain cautious on US duration. The fiscal trajectory, with debt above 124% of GDP[6] and a new administration with limited appetite for fiscal restraint, keeps the long end of the US curve under pressure. We prefer European duration, particularly German Bunds: the European economy is structurally weaker, more exposed to global trade fragmentation, and less supported by domestic demand. If oil prices stay down, the ECB’s hawkishness will prove short-lived.

Credit: yields, not spreads

Despite spreads being historically tight, credit has continued to deliver meaningful carry. We are buyers of absolute yield, not of spread compression. In a higher-cost-of-capital environment, credit differentiation matters far more than it did in the zero-rate era: companies that relied on cheap refinancing to sustain marginal business models will face genuine stress as debt rolls at higher coupon levels. We are constructive on investment-grade quality in sectors with structural earnings support. In high yield, we focus on issuers with genuine competitive advantage and manageable refinancing needs, and we avoid the index.

Emerging market debt

We hold a positive view. The carry premium in hard-currency EM sovereign debt is meaningful and backed by structural tailwinds. Commodity-rich economies benefit directly from the resource scarcity thesis. Several major EM central banks are further advanced in the monetary normalisation cycle than their developed-market peers, creating a positive duration backdrop in local currency. The geopolitical reconfiguration of supply chains is directing investment flows into third-party manufacturing hubs across Asia, Latin America and selected parts of Africa, improving fiscal balances and reducing external vulnerability. We approach EM selectively, with highest conviction in commodity-exporting economies and in manufacturing hubs benefiting from supply-chain duplication.

Portfolio construction: Complexity as opportunity

The 60/40 portfolio was designed for a world where bonds hedged equity risk through their negative correlation with risk assets. That correlation held during demand-shock recessions, where falling growth pushed rates lower and bond prices higher. It broke down from 2022 onwards as supply-side shocks generated stagflationary pressure and bonds and equities fell simultaneously. The hedge failed precisely when it was needed.

For H2 2026, as the supply-shock inflation impulse fades, sovereign bonds are gradually recovering their hedging properties, particularly in Europe. They should be viewed as positive contributors to portfolio return through carry, with a hedging role that is real but partial. The challenge is to find additional sources of genuine diversification, assets whose return drivers are structurally different from the equity-bond complex.

Commodities as a scarcity hedge. Physical commodities, including industrial and precious metals, provide direct exposure to the resource constraint thesis at the heart of our strategy. Their return drivers are less correlated with equity market performance and are positively correlated with the supply-side inflationary environment where bonds fail as a hedge. After two decades of underinvestment, the conditions for a structural revaluation are in place.

Private assets: access to the capex supercycle without listed-market volatility. A significant portion of the most compelling investments in AI infrastructure, power grids, critical minerals and defence manufacturing are not yet accessible through listed markets. Private equity and infrastructure funds provide access to companies at the centre of the capex buildout, with smoother return profiles than the volatile listed-market expression of the same themes. The investment case in the “hard world” rests on operational value creation and genuine sector exposure, not on leverage and multiple expansion, which drove private equity returns in the zero-rate era.

Liquid alternative strategies. The wide dispersion of returns across sectors and the divergence in monetary policy trajectories create conditions that are historically favourable for global macro and market neutral equity strategies.