The first group of major sovereign wealth funds and public pension funds to report calendar year 2025 results shows that performance dispersion was primarily structural with high-equity savings funds such as Norges Bank Investment Management (NBIM), the New Zealand Superannuation Fund (NZSuper) and Korea Investment Corporation (KIC) clustered in the mid-teens. In contrast, diversified defined-benefit models such as Canada’s La Caisse and OMERS and the Swedish AP funds delivered lower returns. Stabilisation-oriented funds such as the State Oil Fund of Azerbaijan (SOFAZ) reported much higher headline outcomes, heavily influenced by revaluation effects. Where benchmarks were disclosed, excess returns were generally modest, with a small number of exceptions with a pattern indicating that allocation design and currency exposure mattered more than security selection.

Equity Beta and Geographic Exposure

Institutions with high global equity allocations benefited most directly from market conditions. NBIM returned 15.1% with more than 70% of assets in equities, as we reported in January. NZSuper and KIC delivered 14.3% and 13.9% respectively as both operated reference-portfolio structures with relatively high global listed exposure. Alaska Permanent Fund Corporation (APFC), which is also heavily exposed to North American equities, reported 12.5%. Meanwhile, Taiwan’s Bureau of Labor Funds (BLF) achieved 16.0%, which we saw was supported by strong public equity exposure including domestic tech holdings.

These funds share a common feature: high participation in listed equity markets. Their performance dispersion is narrow relative to the broader universe with benchmark-relative outcomes further reinforcing this point: KIC reported excess returns measured in tens of basis points; NBIM underperformed modestly; and NZ Super outperformed by roughly 1.4 percentage points. In each case, the scale of alpha was limited compared with the magnitude of market beta.

Geographic structure compounded these effects. Funds with large North American allocations were more sensitive to US equity dynamics and to US dollar movements. Nordic funds such as NBIM and AP7, both with substantial North American exposure, explicitly cited currency translation as a material factor. In contrast, funds with broader regional diversification or greater domestic orientation displayed different sensitivity patterns. Khazanah, with a high domestic allocation relative to global savings funds, returned 5.2%, which we reported refleted both Malaysian market conditions and its investment orientation which aligns with national economic development objectives.

KIC provides a useful reference point within this group. It reported a 13.9% return in 2025, with traditional assets accounting for over three quarters of total assets and generating a 15.1% return. Benchmark outperformance was measured at 24 basis points. The scale of excess return is modest relative to the headline number, reinforcing the broader pattern: in 2025, strong equity markets did most of the work. Value add came from bond positioning and selective equity exposure rather than from large deviations in asset allocation. KIC’s result sits squarely within the mid-teens cluster and supports the view that dispersion across savings funds was limited once benchmark structure is taken into account.

Currency Policy as a Return Driver

Several state-owned investors identified foreign exchange as a key determinant of reported performance. OMERS reported that US dollar depreciation detracted from total returns, although active hedging mitigated part of the impact. Meanwhile, La Caisse similarly highlighted adverse effects from US dollar weakness, partly offset by currency management overlays.

In unhedged structures, currency effects were more visible. AP7, which does not hedge foreign exchange exposure in its equity portfolio, experienced translation drag as the Swedish krona strengthened and deliverdd a 4.2% return, as we reported last month. NBIM’s reporting currency exposure likewise amplified the impact of exchange-rate movements on headline results. These examples show that two funds with similar asset allocations can produce different reported outcomes depending on currency policy.

FX policy is important when comparing funds with similar equity weights but different hedging frameworks. NZ Super, with substantial global equity exposure and limited currency hedging, delivered a return close to that of NBIM but with different currency dynamics. OMERS, with active hedging decisions, quantified the basis-point impact of currency management. In 2025, FX policy was a measurable contributor to dispersion.

SOFAZ reported a 27.2% return, significantly above the rest of the early reporters. However, as we reported in January, the underlying investment portfolio return was mid-single digit, with much of the increase in net assets attributable to foreign exchange and gold revaluation. Gold represented a substantial share of assets at year-end, and rising prices materially influenced reported outcomes.

This pattern differs from that of savings or pension funds. Stabilisation funds are often structured to manage fiscal transfers and macroeconomic buffers. Their balance sheets can be heavily influenced by commodity prices and currency movements. Comparing SOFAZ’s headline return with that of NBIM or NZ Super without adjusting for revaluation mechanics would obscure mandate differences.

Pension Architectures: Diversification, Policy Portfolios and Liability Hedging

The dispersion within public pension funds in 2025 reflects two distinct architectural approaches.

The first group consists of diversified defined-benefit compounders such as CDPQ, OMERS and the larger Canadian and Nordic multi-asset plans. These institutions balance public equities with material allocations to private equity, infrastructure, credit and real estate. In 2025, this structure moderated total-fund returns relative to high-equity savings funds. La Caisse returned 9.3% against a 10.9% benchmark. OMERS delivered 6.0%, with public equities in double digits but private equity negative. The Swedish buffer funds reported mid-single-digit outcomes, with AP7 at 4.2%, AP4 at 6.3% and AP2 at 4.6%. KEVA reported 5.8%. In each case, strong listed equity markets were partly offset by slower-moving private market valuations and stabilising fixed income allocations.

Canada’s biggest PPF, the federally administered CPP Investments which reports a fiscal year to end-March, reinforces this pattern. At the end of December 2025, net assets stood at C$780.7 billion, with a quarterly return of 0.5% and a nine-month fiscal year-to-date return of 7.0%. Public equities were the primary contributor to results, while foreign exchange movements reduced returns in Canadian-dollar terms. The 10-year annualised net return of 8.4% remains comfortably above the Chief Actuary’s long-term real return assumption of just over 4% above inflation. As with La Caisse and OMERS, CPP’s diversified allocation across public markets, private equity, credit and real assets moderates upside participation in equity-led years while supporting funding resilience over time. Its 2025 profile therefore aligns with the broader Canadian defined-benefit model rather than with the high-equity savings funds.

These outcomes are consistent with portfolio design. Diversified defined benefit funds are built to stabilise funding ratios across manu economic cycles rather than to maximise yield in single equity-led years. In a year such as 2025, when listed markets were the main driver of returns, such funds’ structural diversification limited headline performance when compared to to high-equity sovereign wealth funds, particularly those with stabilisation mandates.

A further illustration of mandate-driven structure is provided by South Korea’s National Pension Service (NPS). As a public pension fund, NPS combines a large overseas equity allocation with a still-significant domestic bond base. Its portfolio design places it between high-equity savings funds such as NBIM and diversified Canadian defined-benefit plans such as La Caisse and OMERS. In a year when listed markets were the primary return driver, NPS’s growing global equity exposure would have supported double-digit outcomes, while its bond allocation moderated volatility. Its structure therefore reinforces the broader pattern visible in 2025: performance differences across pension funds were largely determined by equity weight and currency exposure rather than by concentrated active bets.

Japan’s Government Pension Investment Fund (GPIF) illustrates how a statutory policy portfolio constrains dispersion in outcomes. GPIF maintains its long-standing allocation of roughly 25% each to domestic equities, foreign equities, domestic bonds and foreign bonds, with alternatives still a small share of assets. In CY2025, GPIF returned 12.3% (5.8% in the first three quarters of FY2025 to end-December) with assets of approximately JPY293 trillion. Equity portfolios performed strongly over the period, while domestic bonds were negative and foreign bonds positive. In a year when equity-led savings funds posted mid-teens results, GPIF’s outcome reflects its policy mix rather than its tactica positioning because its mandate requires significant bond exposure to stabilise Japan’s public pension reserve.

The second group consists of liability-driven pension structures, exemplified by ATP. ATP operates a dual-portfolio model separating  its larger liability-hedging portfolio from a smaller risk-seeking investment portfolio. Its reported return close to 20% in 2025 reflects the behaviour of its hedging portfolio in a year of significant rate and market movement. Unlike diversified DB compounders, ATP’s total return can be massively influenced by gains in derivatives and interest-rate hedges that are designed to match guaranteed pension liabilities. Asu such, its performance therefore responds to liability dynamics as much as to market beta.

Putting ATP alongside La Caisse or OMERS without acknowledging this structural distinction would not be comparing like with like and the 2025 results show that within the pension universe, dispersion was driven not only by equity exposure and private market allocations, but also by the extent to which pension liabilities are actively hedged through derivatives and rate-sensitive instruments.

Together, these two pension models reinforce the broader conclusion from the early data: differences in return outcomes in 2025 largely reflect mandates and balance sheet composition rather than deviations in investment skills of fund employees and managers.

Value Add and Benchmark Discipline

Across the early reporters, benchmark-relative performance was generally restrained. NZ Super outperformed its benchmark by roughly 1.4 percentage points, making it one of the most successful value-adders among the cohort. KIC reported modest excess returns. PUBLICA in Chile slightly outperformed its benchmark. NBIM, La Caisse and OMERS underperformed in 2025, although over longer periods the Canadian funds remain ahead.

The most notable benchmark outperformance came from the Future Fund, which returned 12.4% against a materially lower benchmark as we reported last month. Its allocation across equities, infrastructure, alternatives and tactical overlays contributed to excess returns. Unlike reference portfolio-driven savings funds, the Future Fund’s structure allows more a greater deviation from allocation targets and is focused on returns referenced to CPI.

In most other cases, outperformance was modest and came from careful bond selection, interest rate management, tactical positioning and efficient implementation rather than from large or concentrated equity positions. The gap between funds’ total returns was much wider than the gap between their results and their respective benchmarks.

The early CY25 data therefore show that institutional design translated market trends and conditions into different results. High-equity savings funds such as NBIM, NZ Super and KIC captured strong market beta, while diversified defined benefit models such as La Caisse, OMERS and the AP funds captured less of the upside but maintained mandate consistency and stabilisation funds such as SOFAZ reflected commodity and currency revaluation. As further large allocators report, more granular patterns will emerge that will give a better basis for reflecting on how mandates and allocations impact returns and value add performance.