ReadingResearchRadarInvestment framework
Sign in / Sign up
← Back to reading
NewsMaterial3 min readDisconfirmAI Editorial

Singapore GIC: AI Capital Expenditure and Government Deficits Compete for the Same Savings Pool; The Reasons for Rate Increases Are as Important as Their Magnitude

GIC’s Macro & Capital Markets team posits that three forces—geopolitics, AI, and fiscal risk—are converging in bond markets to push rates higher. Its core judgment is that the reasons for rate increases are as important as their magnitude, with the drivers determining asset winners: growth driven by productive AI investment favors equities and the US dollar, while term premium rises driven by fiscal risk favor gold and weaken the dollar. The article also presents a three-tier classification framework for AI investments (enablers, monetisers, adopters) and three criteria for identifying durable value.

2026-11-04 14:33 GMT+8

In this outlook released in November 2026, Singapore’s Government of Singapore Investment Corporation (GIC) Macro & Capital Markets team proposes a judgment with direct operational significance for asset allocation: interest rates are likely to stabilize at a higher level, but why they rise matters as much as how far they go. A rise driven by stronger growth has entirely different implications for equities, the US dollar, and gold than one rooted in fiscal concerns.

This judgment rests on a specific mechanism. GIC argues that three forces are currently reshaping the investment environment: geopolitical shifts, AI technological change, and fiscal risk. These do not operate independently but converge in bond markets. Governments are expanding defense and industrial spending due to security priorities, while AI hyperscalers are pushing capital expenditure to nearly 2% of US GDP due to compute demands; both are competing for the same limited pool of savings. Citing data, GIC notes that net debt issuance by hyperscalers over the past year has exceeded the government bond issuance volume of any single European country. As price-insensitive pension and insurance funds shrink and more price-sensitive global investors fill the gap, cross-market correlations structurally rise, making stress in one nation’s bond market more easily transmitted to others.

Sign in / Sign up
ReadingResearchRadarInvestment framework

Within this framework, GIC categorizes the drivers of rising rates into 'good' and 'bad'. If rising rates reflect higher nominal growth from AI capital expenditure, equities may perform well and the US dollar will be supported. If rising rates reflect concerns about fiscal sustainability and rising term premia, the US dollar may weaken while gold benefits. GIC judges that current yield increases are primarily driven by higher nominal growth, which is broadly supportive for equities and the dollar; however, the risk lies in the same yield rise shifting to reflect fiscal premiums and capital competition.

Regarding AI investment itself, GIC provides a two-layer framework. The first layer answers 'where to look': classifying investment targets into enablers (firms building computing hardware, high-speed networks, and energy systems), monetisers (firms creating AI-powered products and services), and adopters (firms embedding AI into existing operations to enhance productivity), with leaders in both public and private markets across each category. The second layer answers 'what to look for': within each tier, GIC focuses on three criteria—moat (proprietary data, governance over critical workflows, and products hard for AI to displace), management (leaders who tie AI to actual business outcomes rather than just cost-cutting), and momentum (early success generating better data, which improves AI performance in a self-reinforcing cycle).

GIC explicitly marks the limitations of this framework. The three forces evolve rapidly, and short- and medium-term effects may diverge, necessitating finer-grained judgments across time horizons, sectors, and countries. The impact of AI on productivity has a 'wide cone of outcomes,' with uncertainty spanning the scale of impact, value distribution, and ultimate beneficiaries. The geopolitical shift toward multipolarity is gradual rather than abrupt. Fiscal tightening is politically unpopular; the IMF estimates that fiscal tightening in 18 advanced economies since the COVID-19 pandemic is the lowest since 1990. GIC concludes with a constructive view on the long-term productivity potential of AI, but warns that a world of higher rates, higher volatility, and higher inflation requires prudence, with core principles being diversification, agility, and granularity.

Sources

  • gic.com.sgGIC↗
Continue researching →