Geopolitical risk and concentration risk in the equity index of the Government Pension Fund Global

We refer to the Ministry of Finance's letter of 25 February 2026. In this letter, we answer the questions concerning geopolitical risk and concentration risk in the GPFG’s equity benchmark index. The questions relating to the investment strategy for bonds are answered in a separate letter dated 31 August 2026. In both letters, Norges Bank emphasises the importance of broad diversification in a changing world.

The Ministry of Finance notes that geopolitical risk has increased in recent years and asks Norges Bank to develop relevant scenarios in order to enhance knowledge and understanding of how geopolitical risk may affect the fund's investments. In the first part of this letter, Norges Bank describes its work on scenario analyses in the management of the GPFG. The Ministry further notes that concentration in the fund's equity index has increased over time and requests the Bank's analyses and assessments. This is addressed in the second part of this letter. Finally, the Bank offers some remarks on features common to these two types of risk.

Geopolitical risk

Background

The fund was established in a period characterised by increased international trade and economic integration between countries. This integration has been an important driver of the fund's growth since the first transfer to the fund 30 years ago. The Sverdrup Commission (Official Norwegian Reports NOU 2022:12) pointed out that several of the developments that have been particularly favourable for the fund may not continue. In recent years, war and conflict, protectionism and a more fragmented world economy have created uncertainty. These developments may affect the conditions for the management of the fund. The uncertainty is amplified by the fact that geopolitical events often occur abruptly and can have broad effects across assets and markets.

Geopolitical events may have implications both for the value of the GPFG and for the management of the fund. The value of the fund's investments may be affected through market turbulence, higher risk premiums and weaker earnings at the companies in which the fund is invested. A more fragmented world may also affect the fund's investment opportunities. The Sverdrup Commission noted that a large sovereign fund may attract attention from both allies and hostile actors. The Commission also noted that the GPFG, as a large sovereign fund, may be more exposed to geopolitical risk than other funds, and that the risk of the fund's assets being frozen or confiscated may be higher than for comparable non-sovereign investors. The Commission emphasised that the situation at the time gave no grounds for saying that this is a general phenomenon, but that it should form part of the basis for assessment in future reviews of the investment strategy.

In its report of January 2026, the Expert Council for the GPFG pointed out that increased geopolitical tensions and uncertainty mean that the political risk to which the fund's investments abroad are exposed is rising, and that the fund's size and operations have at the same time contributed to making it more visible, which further amplifies the political risk.[1] The Council pointed out that, in the most extreme case, the fund could become subject to increased taxation, regulatory intervention and even confiscation.

Despite a more turbulent world, the fund's return in recent years has been good. Over the past three years, the total return on the fund's equity portfolio has been more than 70 percent. This is a development we cannot expect to continue.

Scenario analyses

The consequences of potential geopolitical events can be difficult to assess and price. Scenario analyses are therefore useful. Norges Bank has an established process for scenario analyses to strengthen its understanding of how geopolitical risk, among other things, may affect the fund's investments. The scenario analyses are used in stress tests and form part of the Bank's overall risk management. The analyses are also used in preparing for contingencies, including as a basis for internal exercises, and to identify and manage operational risk in the management of the fund.

Stress tests

Norges Bank conducts regular stress tests of the investment portfolio. The purpose is to estimate potential losses on the portfolio under highly adverse circumstances. Each year, a separate report with selected stress tests is published on the Bank's website.[2] These stress tests are also discussed in the annual report.

The stress tests comprise both historical and hypothetical scenarios. The former take earlier crises as their starting point, such as the financial crisis in 2008, and calculate what equivalent events would have meant for the current portfolio. The hypothetical stress tests are forward-looking and designed to capture extreme events that could have a large negative impact on the fund's portfolio. The scenarios used in the stress tests are selected based on an overall assessment of probability and severity. The selection is adjusted from year to year in line with changes in the risk picture. In the stress tests, different outcomes are expressed as changes in cash flows, risk premiums, inflation expectations, etc. These changes are then applied to the portfolio's actual composition to estimate potential losses over the medium term.

In recent years, the hypothetical stress tests have included several geopolitical scenarios. In 2022, the potential effects of a major geopolitical conflict and subsequent decoupling in the world economy were analysed. The assumptions included lower economic growth, a fall in companies' cash flows and higher risk premiums, followed by persistently high inflation expectations and weak real returns on government bonds. In 2023, we published a scenario in which cooperation between two large economic blocs breaks down, and in 2024 a broader scenario in which the world fragments into several distinct blocs. The assumptions included higher tariffs, increased regulation, restrictions on foreign investment and weakened fiscal discipline, which together would result in lower global growth, permanently lower cash flows and increased market volatility. In 2025, the scenario was developed further with more extensive tariffs and countermeasures.

Norges Bank has estimated that the value of the fund could fall by 30–40 percent as a result of geopolitical upheavals of this kind. Despite considerable uncertainty, including in relation to risks analysed in the stress tests described above, markets have recently delivered very strong returns. In Norges Bank's assessment, however, the scenarios described in these stress tests remain relevant in light of recent developments.

Use of scenario analyses in ongoing risk monitoring

Dynamic geopolitical scenario analyses form part of the Bank's ongoing risk monitoring. The scenarios are also used in contingency exercises, both internally within Norges Bank and in joint exercises with the Ministry of Finance. Scenario-based exercises are useful insofar as they can help determine whether there is sufficient flexibility in the mandate and the organisation to manage crises. The scenarios we consider are of such a scale and scope that they would have to be managed in close cooperation with the Ministry of Finance. The details of these scenario analyses are not public. The ongoing work on the dynamic scenarios is organised into four phases: identification of key risks, scenario development, monitoring of various indicators and probability assessments.

The starting point is an annual mapping of geopolitical risk. The risks assessed to have the largest potential consequences for the value of the fund are followed up on an ongoing basis, including through the use of scenarios. The scenarios describe possible developments, with particular emphasis on outcomes with low probability but large potential consequences. The analyses assess how different scenarios may play out across asset classes, sectors and geographies. Key indicators are also identified, making it possible to follow developments on an ongoing basis and capture changes in the risk picture. The various possible courses of development within each scenario are assigned a probability, which is reviewed quarterly and adjusted in line with changes in the risk picture. The probability assessments are based, among other things, on observable market information.

Summary and Norges Bank's assessment

The international environment has become more demanding in recent years, and there is little to suggest that this development will reverse. The investment strategy for the fund is based on the premise that exposure to market risk makes it possible to earn market risk premiums over time. That requires a willingness to bear risk, including in periods of considerable uncertainty. Investments in international markets today entail substantial exposure to geopolitical risk. The fund's distinctive features may also mean that certain geopolitical risks can affect the fund in ways that differ from how they affect other investors. In this context, it is important that the fund's financial objective remains firm and is perceived as such by the outside world.

Norges Bank's stress tests show that the value of the fund could fall significantly in the event of severe geopolitical upheavals. In its report of January 2026, the Expert Council for the GPFG pointed out that geopolitical risk should be managed through continued broad geographical diversification. At the same time, and as illustrated by the Bank's stress tests, diversification will not shield the fund from geopolitical events and upheavals that affect a broad range of markets, asset classes and sectors simultaneously.

Norges Bank has established comprehensive systems for identifying, monitoring and analysing geopolitical risk along many dimensions. The Bank also has comprehensive frameworks for managing related forms of operational risk that may be amplified in periods of geopolitical turbulence, including attempted fraud and cyberattacks. The Bank also monitors, on an ongoing basis, companies in the portfolio that can be linked to situations of war and conflict. Geopolitical events can, however, be predicted only to a limited extent. Even with comprehensive systems for risk monitoring, it is not necessarily possible to reduce exposure to affected markets before a crisis occurs. In many cases, financial losses will already have been incurred before measures can be implemented. The size of the fund also means that the scope for rapid divestment in a crisis is very limited.

The fund's investments consist of ownership stakes and lending placed in markets throughout the world. This means that the fund's assets cannot readily be affected by individual events in one country or one market. At the same time, the management of the fund depends on well-functioning markets, a predictable legal framework and reliable financial infrastructure, including settlement systems, custody solutions and services provided by key market participants. History provides examples of such framework conditions changing as a result of geopolitical events, and of investors' access to their own assets being restricted in some cases. The Bank monitors this type of risk on an ongoing basis. At the same time, several of these factors lie outside the Bank's own control, and monitoring will not in all cases make it possible to avert the consequences.

Under the mandate, Norges Bank has a duty to provide advice to the Ministry of Finance. Should the Bank's assessment indicate that the management of the fund cannot be conducted in a sound manner within the current framework, the Bank could advise the Ministry to amend the mandate.

Concentration in the equity index

Developments in concentration in the equity index

Concentration in the equity index can be measured in several ways. Common measures are the largest companies' share of index weight and of risk contribution, as well as the effective number of companies.[3] Regardless of which measurement method is used, concentration has increased in recent years; see Figure 1 in the appendix. This has been driven by developments in the US equity market and has had a bearing on the degree of diversification in a global market-weighted index. Similar levels of concentration have, however, also occurred previously, globally in the 1980s and around the turn of the millennium, among other times, and in the US market up until the 1960s.[4]

Concentration in the equity index today reflects underlying differences in earnings growth. US companies have over time had higher real growth than companies in other regions, and consensus estimates from economic forecasters suggest that this will continue.[5] At the sector level, the technology sector, and US technology companies in particular, has had higher growth than the economy over several decades. The emergence of artificial intelligence has reinforced this development in recent years. Globalisation and technological change have also made it easier for dominant companies to gain market share.[6] Concentration is thus driven by a combination of higher US growth, the technology sector's increasing share of value creation and the emergence of large companies with global market positions.

US companies' share of total earnings in the global equity index MSCI ACWI, which comprises equities from both developed markets and emerging markets, has risen steadily. Concentration measured by market value is higher than concentration measured by earnings, which indicates that investors are pricing in expectations of continued higher growth for US equities; see Figure 3 in the appendix. This means that concentration in the index reflects not only the current distribution of earnings, but also the market's expectations of future growth.

Possible implications of concentration

Increased concentration may make a market-weighted index less diversified and more vulnerable to developments in individual markets, sectors and companies. In its report of January 2026, the Expert Council for the GPFG pointed out that the current governance model for the fund may give less attention to concentration at the sector level than in funds with a greater element of so-called tactical asset allocation, and that concentration should be a priority area in the risk management of the fund. At the same time, the Expert Council pointed out that a few large companies do not necessarily entail less diversification. Many of the leading technology companies have broad platforms spanning different technologies and markets, both in terms of products and geographically.

Norges Bank has previously pointed out that the most relevant financial risk for a long-term investor is the risk of permanent loss. In the Bank's advice on the geographical distribution of the equity index of August 2019,[7] it was noted that permanent losses can arise as a result of changes in expected earnings. The increased concentration in the equity index has in recent years been reinforced by expectations regarding artificial intelligence. Many of the largest companies in the benchmark index depend on the same factors, including the return on investments in artificial intelligence infrastructure, developments in US technology regulation and access to advanced semiconductors. A shock affecting these factors could therefore affect a large part of the index.

The Bank's stress tests illustrate this. In a hypothetical scenario from 2024, we assumed that companies within artificial intelligence do not meet expectations for future earnings. In 2025, a similar scenario was modelled, in which market developments had in the meantime produced even greater concentration. In the 2025 scenario, we assumed that artificial intelligence does not meet the market's expectations of productivity gains, with lower expected cash flows and a higher equity risk premium. The impact on the value of the fund was estimated at falls of 18 and 35 percent respectively in the two scenarios.

As described above, there have also been earlier periods of high concentration in the equity market. However, there are few such historical periods of high concentration. In the US equity market, concentration reached peaks towards the end of the 1960s and in connection with the dotcom bubble in the early 2000s. In both cases, the peak was followed by around a decade of weak returns. The differences in realised returns over the following ten years between periods of high and low concentration are nonetheless modest; see Table 1 in the appendix. It is therefore difficult to draw conclusions about the significance of concentration for future index returns on the basis of historical data.

The fund's benchmark index

Norges Bank has previously noted that a float-adjusted market-weighted index is the natural starting point for the equity benchmark index, and that any deviations from a float-adjusted market-weighted index should be justified and serve a specific purpose.[8] The equity benchmark index today has a somewhat lower share in North America, and thus somewhat lower concentration, than would follow from a float-adjusted market index.[9] This is due to the geographical adjustment factors the Ministry has set out in the management mandate. In 2020, the Ministry cited[10] a desire for broad geographical diversification, among other things, as justification for the share of the benchmark index in North America continuing to be somewhat lower than would follow from float-adjusted market weights.

The Ministry has in addition established a rebalancing rule between equities and bonds in the benchmark index which ensures that the equity share of the fund remains close to the strategic target of 70 percent. It is important that the framework for the management of the fund contains mechanisms that provide discipline in periods of considerable uncertainty. The rebalancing rule enables the fund systematically to buy equities after large price falls and sell equities after periods of price increases. Predictable, rule-based mechanisms are an important feature of the framework for a fund with the GPFG's distinctive characteristics.

In the letter of 27 August 2020, Norges Bank assessed how a possible cap on individual markets in emerging equity markets could be designed. Caps on sectors and countries were also discussed in the letter of 21 August 2019 on the geographical distribution of the benchmark index. The purpose of such caps is to limit the risk of permanent loss from events that affect an individual market or an individual sector. A high degree of concentration does not, however, in itself necessarily imply that the benchmark index should deviate from market weights. Market weights reflect the market's overall assessment of companies' prospects. Introducing caps on markets, sectors or individual companies may result in an investor missing out on the return from the companies with the best future prospects.[11] This is illustrated by analyses of developments in the benchmark index since 1994, which show that a lower cap on individual companies has consistently produced lower returns than a higher cap or no cap; see Figure 8 and Table 2 in the appendix. A cap further presupposes a discretionary assessment of what constitutes unacceptable concentration. Caps also introduce greater complexity into the index and entail transaction costs that increase the lower the cap is set.

Concentration in the equity index today is to a considerable extent linked to expectations regarding artificial intelligence. The companies concerned are spread across traditional sector boundaries, and companies outside what the index provider defines as the technology sector are also priced on the basis of the expected benefits of artificial intelligence. Sector classifications may also change over time. A sector cap would therefore only partly capture the exposure. Any cap on individual companies would attach to the legal entity, not to the underlying risk exposure. If a company is split up, a cap would potentially no longer apply to the new entities, even though the overall exposure is unchanged.

Summary and Norges Bank's assessment

Concentration in the equity index has increased in recent years, mainly driven by developments in the US equity market and expectations regarding artificial intelligence. The Ministry has discussed the increased concentration in the equity index in several white papers on the Government Pension Fund. At the same time, the Ministry has noted that it is demanding to assess the risk associated with this concentration, and the extent to which it warrants adjustments to the benchmark index. In the 2021 white paper on the Government Pension Fund, the Ministry assessed caps on individual markets in emerging equity markets and found that the drawbacks, including complexity, the need for discretionary considerations and the distance from the broad index, did not justify the measure. In the Bank's view, the same objections also apply to any sector or company cap. Such caps are also less precise tools for managing the type of concentration to which the fund is now exposed. On this basis, the Bank does not advise introducing caps on individual markets, sectors or companies at this time. In the Bank's assessment, the risk associated with the level of concentration in the current benchmark index must be accepted as a feature of a market-weighted index.

The fund's size, long-term horizon and ability to bear risk are distinctive features that may confer advantages which can be used to generate excess return. The limit for deviation from the benchmark index provides scope for exploiting these advantages. In the Bank's assessment, the limit for deviation from the benchmark index is less suited to managing concentration risk, as this measure is not clearly linked to excess return, which would therefore make it difficult to evaluate the active management. In addition, it would reduce the room for value-creating strategies. 

The GPFG differs from several other large funds in that systematic risk is mainly taken in listed markets. The current mandate allows for investments in unlisted real estate. It also allows for investments in unlisted renewable infrastructure, which is a small part of the total market for unlisted infrastructure. With a strategic equity share of 70 percent and a lower share of unlisted investments than comparable funds, the fund is to a greater extent exposed to developments in the listed equity markets. There may therefore be reason to consider whether an increased allocation to unlisted assets over time could give the fund exposure to a broader set of sources of risk and reduce the significance of the concentration in the equity index for the fund’s overall risk profile. The Expert Council for the GPFG has recommended that it be assessed regularly whether the fund should be spread more broadly across asset classes. Norges Bank shares this assessment. At the same time, investments in unlisted markets must be built up gradually, and will only have a material bearing on the fund's overall risk profile if they reach a certain scale. A potential increased allocation to unlisted assets would require thorough review of the benefits and drawbacks, including political and regulatory risk, and broad anchoring.

Concluding remarks

Geopolitical risk and concentration risk are addressed separately in this letter, but they may be interwoven and confront the management of the fund with some of the same fundamental trade-offs. The two types of risk may also reinforce each other. Concentration in the equity index today is to a considerable extent linked to US technology companies. This means that a geopolitical event affecting these companies could have greater significance for the fund. The technology driving the concentration is among the factors involved in the rivalry between large economies. In addition, the services provided by the US technology companies are very important for many businesses and organisations, including Norges Bank.

The Sverdrup Commission pointed out that the fund has a high capacity to bear financial risk, but that there could be an “increasing need to consider, balance and anchor the fund's general capacity to handle various types of non-financial risk, including political risk”. Developments in the geopolitical environment in recent years underline the relevance of this assessment. When a larger share of the fund's return depends on political and regulatory factors, the boundary between financial and non-financial risk may appear less sharp than previously. In the Bank's assessment, this makes it more important than before that the fund is perceived as a financial investor, and that the financial objective of the management of the fund is communicated clearly.

Norges Bank takes it for granted that both geopolitical risk and concentration risk will be present in the management of the fund going forward. Given the current market composition and international conditions, neither geopolitical risk nor concentration risk can be eliminated in a market-weighted equity index. Measures that could reduce one form of risk will in many cases entail new trade-offs between the desired risk reduction and increased complexity, higher transaction costs or unintended exposures.

As discussed in this letter, the scope for predicting geopolitical events or adjusting the portfolio quickly in a crisis is limited. Open communication about the risk entailed by the strategy for the fund, including the publication of stress tests, helps to build understanding that large falls in value may occur. This strengthens the basis for adhering to the long-term strategy, including in periods of turbulence.

Yours faithfully

Ida Wolden Bache
Nicolai Tangen

[1] Expert Council for the Government Pension Fund Global, Report 2025, 23 January 2026, p. 17.

[2] 2022: https://www.nbim.no/contentassets/dca9fe25bacb4cbc97bee89173a35474/stress-testing-2022.pdf

2023: https://www.nbim.no/contentassets/1fa1b9d3c2bd4bd6b3e4833add584be1/stress-testing-2023.pdf

2024: https://www.nbim.no/contentassets/1c26bc90eb274356a60fd77ae5a5bf49/gpfg-stress-testing-2024.pdf

2025: https://www.nbim.no/contentassets/1c26bc90eb274356a60fd77ae5a5bf49/gpfg_stresstesting-2025.pdf

[3] The effective number of companies is calculated from the Herfindahl-Hirschman index (HHI). Lower values imply that market value is distributed across fewer companies.

[4] Pastor et al. (2026), Bye, Soerlie Kvaerner and Werker (2026).

[5] Consensus estimates are a compilation of forecasts from several professional financial analysts or macroeconomic forecasters. At company level, analysts typically prepare estimates for earnings and growth. At the macro level, estimates are prepared for variables such as GDP growth and inflation. Consensus estimates are calculated as an average of the individual forecasts and provide an expression of the market's overall expectations.

[6] See, among others, Autor et al. (2020).

[7] See Norges Bank's letter of 27 August 2019.

[8] See, among others, Norges Bank's letter of 21 August 2019.

[9] See Figure 7 in the appendix.

[10] Report to the Storting No. 32 (2019-2020)

[11] See Pastor et al., 2026.