For most chief executives, volatility is a condition to be managed. For Lim Chow Kiat, it is the environment in which Singapore’s financial reserves must remain useful across generations. GIC’s mandate is not to beat a public index in a convenient year or to maximise the value of a disclosed pool of assets. It is to preserve and enhance the international purchasing power of capital entrusted to it by the Singapore government over the long term.
That mandate has become harder to execute because the assumptions underpinning global diversification are changing together. Trade fragmentation is altering supply chains and asset prices. Artificial intelligence is concentrating extraordinary capital expenditure in a narrow group of technologies and companies. Climate change affects physical assets, regulation and energy systems. Higher public debt and geopolitical rivalry make correlations less stable precisely when investors most need protection.
Lim’s response is to make GIC more granular and more agile: break structural themes into investable segments, preserve liquidity, deepen specialist knowledge and strengthen links between portfolio strategy and execution. Leadership changes announced in July 2026 will give deputy group chief investment officers broader cross-domain responsibilities from October, including closer oversight of total-portfolio research, strategy and implementation. The design reflects a central tension. GIC needs the scale and patience of a sovereign investor, but it must make decisions with the precision of a specialist.
A long horizon does not remove the present
GIC’s latest disclosed performance covers the twenty years to March 2025. The portfolio produced an annualised nominal return of 5.7 per cent in US-dollar terms and an annualised real return of 3.8 per cent after global inflation. Five- and ten-year nominal returns were 6.1 per cent and 5.0 per cent respectively. GIC treats the rolling twenty-year real return as its primary performance measure because it aligns with the purpose of protecting international purchasing power.
The horizon is appropriate, but it can be misunderstood. Long-term investing does not mean ignoring entry prices, liquidity or shorter-cycle risks. It means accepting temporary volatility when the underlying value remains intact while avoiding permanent impairment. A twenty-year metric also changes slowly because each new year replaces a distant year in the calculation. Performance can rise or fall because of both current outcomes and the period leaving the window.
Lim must therefore run the institution on several clocks. Investment teams need to respond to immediate dislocations. Asset departments must build capabilities that take years to mature. The total portfolio must remain robust across economic regimes. The government and public, meanwhile, need confidence that national reserves are being managed with discipline even though GIC does not disclose assets under management or annual returns in the same way as a listed fund manager.
The organisation’s comparative advantage is not freedom from accountability. It is the ability to hold positions through cycles without being forced to meet redemptions from external clients. Lim’s obligation is to convert that patience into better decisions, not simply longer holding periods.
Resilience becomes an allocation discipline
GIC describes resilience through diversification across asset classes, geographies, sectors and time. That broad spread is necessary but no longer sufficient. Global shocks can make apparently different assets respond to the same drivers. Higher discount rates can hurt both equities and long-duration bonds; a power shortage can affect data centres, manufacturing and utilities; geopolitical restrictions can reach technology, trade finance and logistics simultaneously.
Lim’s emphasis on granularity addresses those hidden connections. Instead of treating AI as a single theme, GIC divides the value chain among early technologies, enablers, monetisers and adopters. Each segment carries different economics. Model developers face intense research costs and uncertain competitive moats. Semiconductor and energy infrastructure may benefit from demand but requires heavy capital. Software and service companies must prove that AI improves willingness to pay. Adopters may capture productivity without owning the underlying technology.
The same principle applies to climate. Energy transition pathways differ across countries, and adaptation may offer more immediate investable opportunities than a uniform assumption about decarbonisation. Infrastructure, buildings, insurance and supply chains all face local physical and policy conditions. A global allocation must be assembled from many bottom-up judgments.
Granularity can improve risk selection, but it also increases organisational complexity. More specialist strategies create more interfaces, more data and a greater need to aggregate exposures. GIC must understand not only whether an investment team is right about a company, but how that position interacts with the rest of the portfolio. Cross-domain leadership is intended to make those connections visible.
The AI portfolio is also an operating transformation
GIC is investing across the AI value chain while adopting the technology internally. An AI Council established in 2023 has guided work on efficiency, innovation and agentic capabilities. The institution has been developing tools to support investment discussion and decision-making, drawing on its own data. Lim has also highlighted uses in trade processing and operational efficiency.
The internal adoption matters because a global investor processes enormous volumes of market, company and portfolio information. AI can assist research, identify patterns and automate routine workflows. It may allow investment professionals to test more scenarios or monitor holdings more continuously. It cannot remove the need for judgment about incentives, governance, valuation and regime change.
GIC also faces a version of the same concentration risk confronting public markets. Capital has rushed towards model companies, chips, data centres and power infrastructure. Some demand is durable; some prices assume years of flawless growth. A sovereign investor can participate at several points, from venture rounds to infrastructure and public equities, but broad exposure does not guarantee diversification if every position depends on the same pace of AI adoption.
Recent activity illustrates both scale and selectivity. In 2026, GIC announced investments and partnerships involving enterprise AI, music catalogues and large infrastructure assets. These transactions span very different cash-flow profiles. The question is not whether they all carry an innovation label, but whether each offers a risk-adjusted return appropriate to its structure and place in the total portfolio.
Private markets require a harder liquidity argument
GIC’s ability to invest directly and through funds in private equity, infrastructure and real estate is a strategic strength. Private assets can provide access to operational improvement, long-duration cash flows and companies before public listing. They can also obscure changing valuations, lock up capital and demand follow-on funding at inconvenient moments.
Liquidity is therefore part of the investment thesis rather than a residual balance. Lim has stressed the need to preserve flexibility so GIC can act when trends or prices change. That requires enough liquid assets to meet the government’s needs and fund commitments without selling into stress. It also requires a realistic assessment of how quickly nominally private assets could be monetised and at what discount.
The environment of cheap financing that supported much private-market activity has ended. Higher borrowing costs expose business models built on leverage, while slower exits extend holding periods. Managers may report stable valuations even when comparable public assets have moved sharply. GIC’s scale gives it negotiating power and access, but it does not exempt the institution from the mathematics of purchase price and cash yield.
A more granular approach can help distinguish scarce assets with durable economics from fashionable exposures. It must be paired with willingness to decline transactions. Sovereign capital is attractive to sponsors and companies precisely because it is patient and credible. Those qualities create opportunity only if GIC resists becoming the capital provider of last resort at terms that protect others more than Singapore.
Japan and China show the value of local duration
Lim has been reinforcing GIC’s long-standing presence in Asia’s largest economies. At a Japan partners event in February 2026, he described Japan as both a source of diversification and an opportunity as corporate transformation accelerates. The market combines governance reform, deep industrial capability and important positions in semiconductor materials, energy technology and automation.
GIC’s more than four decades in Japan demonstrate why local relationships matter to long-term capital. Access is not simply a function of cheque size. It depends on trust, continuity and the ability to work with companies through changes that may not fit a conventional fund horizon. The expansion of the Japan office supports more bottom-up sourcing across asset classes.
China presents a different but equally important set of trade-offs. Growth has slowed, property deleveraging continues and strategic rivalry shapes technology and capital flows. Yet the economy retains large pools of talent, manufacturing capacity and consumer demand. GIC’s Beijing presence allows it to assess opportunities close to their operating context rather than treating China solely as a macro exposure.
Local duration does not remove political risk. It can improve the quality of information and partnerships, but regulatory decisions can still alter returns quickly. Lim must balance familiarity with independence, ensuring that long relationships do not weaken the willingness to reprice risk or reduce exposure when fundamentals change.
Sustainability moves into the investment teams
GIC marked its forty-fifth anniversary in 2026 by pushing sustainability expertise more deeply into asset departments. The evolution moves from a cross-asset sustainable investment fund and a central sustainability office towards teams that combine sector investment knowledge with climate and sustainability analysis.
The logic is practical. Physical risk in a property portfolio, transition risk in an industrial company and regulatory risk in a utility require different expertise. Embedded specialists can evaluate how sustainability affects cash flows, capital expenditure and competitive position rather than treating it as a reporting overlay. They may also identify opportunities in adaptation, energy efficiency and infrastructure that a central screen would miss.
Decentralisation carries a governance risk. Different departments can apply inconsistent assumptions or emphasise opportunities while underweighting portfolio-wide exposure. GIC still needs common standards, scenario analysis and escalation mechanisms. The total portfolio must capture correlations among climate-sensitive assets even if individual teams make the investment decisions.
The stronger test is whether sustainability analysis improves returns and reduces permanent loss. Labels are less useful than evidence about asset resilience, customer demand and regulatory direction. Lim’s bottom-up approach will be credible if it changes position sizing, prices risk accurately and occasionally leads GIC to walk away.
Leadership architecture becomes part of performance
GIC has announced several senior appointments as it refreshes institutional capability. Changes taking effect in October 2026 will strengthen the role of deputy group chief investment officers and connect total-portfolio research more closely with implementation. New members have also joined the Group Executive Committee, and the international advisory network has been expanded.
For an institution of GIC’s size, organisation design can affect returns as much as a single investment decision. Specialist departments need authority to act, while the centre must allocate risk and capital across them. Too much central control slows decisions and suppresses expertise. Too little allows duplication, hidden concentration and competition for balance sheet.
Lim’s own experience spans fixed income, global investment leadership and the group chief investment officer role before he became chief executive in 2017. That background supports an integrated view of the portfolio. His challenge is to build a system that does not depend on one leader’s ability to see every connection. Succession, common data and clear decision rights are essential to a mandate measured in decades.
What Lim must prove next
GIC enters its next phase with considerable advantages: stable ownership, global reach, patient capital and a mandate that rewards long-term thinking. The institution has also produced a positive real return over its latest twenty-year window despite crises, inflation and changing market regimes. Those strengths make complacency more dangerous, not less.
Over the coming years, Lim must show that granularity improves the total portfolio rather than creating a collection of attractive stories. AI exposure must be mapped across technology, infrastructure, power and adopters. Private assets must justify illiquidity at new financing costs. Sustainability expertise must influence prices and decisions. Leadership changes must accelerate high-quality action without weakening accountability.
The public will not be able to assess every position, nor should a sovereign investor disclose information that harms execution. It can still judge the consistency of the framework, the long-term results and the institution’s explanation of risk. Trust depends on knowing that opacity is used to protect returns, not avoid scrutiny.
Lim’s central proposition is that GIC should prepare rather than predict. Preparation is harder than it sounds. It demands liquidity before a crisis, selectivity during a boom and the courage to invest when the prevailing narrative is hostile. If GIC can preserve those behaviours at greater scale and complexity, its long horizon will remain an advantage. If it cannot, patience may merely delay recognition of mistakes.
