To main content

Government Pension Fund Global – Account of work on climate risk

Letter sent to the Ministry of Finance, 26 November 2019

28 November 2019

1. Introduction

In its letter of 28 June 2019, the Ministry of Finance asked Norges Bank to review and describe its work on climate risk in the Government Pension Fund Global (GPFG). The Bank was asked to:

  • outline the processes and methods for assessing climate risk as one of a number of risk factors for the fund’s asset classes and for the fund as a whole
  • present analyses of climate risk for the equity, bond and real estate portfolios and for individual companies making the greatest contributions to overall climate risk in the fund
  • describe the various tools and how they are used in the work on climate risk in the fund
  • explain how climate risk is integrated into investment decisions, and
  • set out any plans to further develop work on climate risk in the fund within the bounds of the stipulated investment strategy for the fund and the management mandate from the Ministry

We would refer to previous accounts of the Bank’s work on climate risk in the fund. In our letter to the Ministry of 5 February 2015, we reported on our work on integrating financial risk associated with climate change in the portfolio, and on the status of key international initiatives in this area that the Bank was involved in. In our letter of 21 February 2018, we presented an assessment of whether the recommendations of the Task Force on Climate-related Financial Disclosures (TCFD) would have implications for the Bank’s work on climate risk in the fund.

We would also refer to the Bank’s letter to the Ministry of 15 March 2019 setting out its consultation response to Official Norwegian Report NOU 2018:17 “Climate risk and the Norwegian economy”. Part 3 of that letter discussed the Bank’s role as manager of the fund, including the importance of good corporate disclosure and the role of active ownership in supporting companies’ transition to a low-carbon economy.

Norges Bank’s annual reports on responsible investment of the GPFG, most recently for 2018, provide a broad account of the Bank’s work on climate risk as part of its wider work on responsible investment. This letter will build on these previous letters and reports on climate risk.

The Ministry of Finance lays down the mandate for the Bank’s management of the GPFG, including the management objective, investment universe, benchmark index, risk limits and reporting requirements. The Bank manages climate risk within the bounds of this mandate, and active ownership is a key element in our management of this risk.

The management mandate requires responsible investment to be an integral part of the management of the fund. The management objective laid down for the fund is that the Bank seeks to achieve the highest possible return. The mandate does not require the fund’s portfolio to adapt or contribute to the achievement of specific climate targets. The fund’s composition largely mirrors the benchmark index. The mandate does, however, state that a good long-term return is considered dependent on economically, environmentally and socially sustainable development and well-functioning, legitimate and efficient markets. With this in mind, the Bank’s Executive Board has laid down principles for responsible investment management. These are based on internationally recognised principles and standards such as the UN Global Compact, the OECD’s Principles of Corporate Governance and the OECD’s Guidelines for Multinational Enterprises.

The Bank’s work on responsible investment takes its starting point in our role as manager of a large, global fund with a long investment horizon. We promote the principles of well-functioning markets and good corporate governance, we engage in active ownership to promote long-term value creation at the companies we invest in, and we seek to reduce exposure to unacceptable risks. These priorities also form the starting point for the Bank’s work on climate risk in the fund.

2. Climate risk in the GPFG

Climate risk
Climate risk is one of a number of risk factors for the fund. Climate risk is complex, and the understanding of its financial implications for the fund will evolve over time. It consists of both physical risks and transition risks as we move towards a low-carbon economy. These risks have different time horizons. Physical risks might be exposure to extreme weather events such as floods, droughts or heat waves. Transition risks include regulatory changes, technological innovations and evolving consumer preferences. The prices of the assets we buy as an investor, and the degree to which these prices reflect climate risk, affect the fund’s financial risk. A broadly diversified and market-weighted portfolio such as the GPFG will, in principle, have roughly the same financial climate risk as the underlying markets and sectors it is invested in.

Different analyses and methods
We analyse transition risks in the fund partly by measuring the carbon emissions of companies in the portfolio and by performing scenario analyses. Our analysis of carbon emissions involves calculating our investments’ carbon footprint and carbon intensity. Calculation of carbon emissions provides only a snapshot, however, and does not take account of industry structure, company strategy and other factors. Nor does it take account of (indirect) emissions in the supply chain.

One general challenge facing analyses of climate risk is the limited availability of high-quality and relevant data. Numerous initiatives are under way to increase corporate disclosure and investor access to data. Examples include the EU’s Non-Financial Reporting Directive, the TCFD’s recommendations, and initiatives for reporting on environmental, social and governance issues such as the Global Reporting Initiative (GRI), the Sustainability Accounting Standards Board (SASB) and CDP (formerly the Carbon Disclosure Project).

Scenario analyses are used to illustrate different future outcomes for climate risk and better understand climate risk over long time periods. These analyses can shed light on both physical and transition risks in the portfolio, but are based on factors that are often subject to considerable uncertainty, such as climate models, expected technological progress and potential regulatory changes. They provide an illustration of possible outcomes but are not predictions of the future.

Assessment of climate risk in the fund’s asset classes
We have been developing databases of environmental, social and governance information on the companies in our portfolio since 2015. These databases include climate information, such as emissions data reported by companies and third-party estimates where companies do not themselves provide adequate disclosure.

- Listed equities and corporate bonds
We calculate carbon emissions from the companies the fund is invested in, and aggregate them to obtain the fund’s total carbon footprint. Besides information on total carbon emissions, this can provide an insight into risks and opportunities across sectors and form a basis for assessing how adjustments to the portfolio will affect total emissions from investee companies. We published our first estimates of the fund’s carbon footprint in 2015 and have subsequently developed these analyses to cover corporate bonds as well as equities. We follow the recommendations for asset managers from the TCFD when calculating the fund’s carbon footprint. We start from the carbon emissions of each individual company the fund is invested in, measured as tonnes of CO2-equivalents. At portfolio level, we calculate emissions based on our percentage holdings and each company’s market value.

The calculations for 2018 reveal a total carbon footprint for the fund of around 107 million tonnes of CO2-equivalents. These emissions are driven largely by companies that generate power from fossil fuels and by sectors with high energy consumption, such as basic materials and industrials. Within these high-emission sectors, the largest companies account for the largest emissions.

We have also estimated what carbon emissions from the companies in the benchmark index would be without any ethical exclusions under the Ministry of Finance’s guidelines for observation and exclusion. The calculations for 2018 reveal that exclusions reduced carbon emissions from companies in the benchmark index by 14 percent. This is chiefly due to exclusions under the coal criterion.

In addition to total emissions, we calculate carbon intensity – the number of tonnes of CO2-equivalents per million dollars of revenue. Estimates of carbon intensity at company and sector level can be used to compare how much carbon companies and sectors emit in relation to the revenue they generate.

In recent years, the equity portfolio’s carbon intensity has been lower than that of the benchmark index. The difference can largely be put down to the fund’s investments in basic materials, industrials and utilities having a lower carbon intensity than the companies in those sectors that are included in the benchmark index. For corporate bonds, too, carbon intensity was lower for the portfolio than for the benchmark index at the end of 2018. This was chiefly due to the industrial companies in the portfolio having a lower carbon intensity than those included in the benchmark index. The differences in carbon intensity between companies in the fund and those in the benchmark index are the result not of any specific investment strategy but of a series of different investment decisions.

We have launched an internal project to examine how carbon intensity impacts on equity returns. We have constructed theoretical factor portfolios based on carbon intensity and used these to analyse whether and how this affects ordinary pricing models. The data run from 2009 to 2019, which is too short a period from which to draw conclusions about long-term relationships, but the project has given us an insight into which risks such portfolios will be exposed to. To date, we have been unable to document any statistically significant return effects from carbon intensity as a systematic factor.

We are also developing a number of methodological tools for climate scenarios that can give us a better understanding of where and how climate risk might affect individual companies and the portfolio as a whole. The aim of this work is ultimately to understand better how climate risk might affect returns on the equity portfolio. In an internal model for scenario analysis of the equity portfolio, we have considered future cash flows and carbon emissions at company level, and also how possible future regulation in the form of carbon pricing and carbon quotas might impact on different companies, sectors and regions. We have looked mainly at climate scenarios where global temperatures rise by 1.5°C, 2°C and 3°C by 2100. Provisional results confirm that future carbon pricing may present financial challenges for some companies in sectors such as industrials, energy, basic materials, utilities and consumer goods.

In 2018, we teamed up with 19 other global institutional investors in a pilot project under the UN Environment Programme Finance Initiative (UNEP FI) to develop reporting methods for investors under the TCFD recommendations. A report was published in May 2019 with an overview of different methods and tools for scenario analyses and a review of leading analysis providers. The project generated useful knowledge on the challenges facing scenario analyses in different asset classes, and on how the analyses can be translated into financial information. Experience from the project suggests a need for further development before companies and investors can fully implement the TCFD recommendations. It also confirms the need for companies to disclose relevant and material climate information, including asset-specific data, and to do so in a consistent and comparable manner.

As part of the pilot project, Carbon Delta assessed potential climate impacts on the fund’s equity portfolio through to 2032. Carbon Delta’s model calculated the portfolio’s Value at Risk (VaR) [1] by considering the potential impact on companies’ revenue of regulatory risks and technological opportunities under a 2°C scenario and extreme weather events[2].

- Unlisted real estate
The fund’s unlisted real estate investments are directly exposed to climate risk in the form of both physical and transition risks. This exposure is affected primarily by where the properties are located and how well their technical solutions and operation equip them to deal with changes in the climate and associated regulatory developments.

The most important factor for physical climate risks is a property’s location in relation to changes in climatic conditions, such as floods, extreme weather events and heat waves. The greatest exposure is along the US East Coast, which is expected to face an increase in extreme weather events and floods and rising sea levels. Flooding can cause buildings in exposed areas to suffer substantial damage to facades and technical installations, leading to shutdowns. Buildings at risk of flooding often also come with higher insurance premiums and stricter regulatory requirements for flood protection measures. Requirements of this kind were, for example, introduced in New York following the floods caused by Hurricane Sandy in 2012. In newbuilds and major renovations on floodplains, technical installations in storeys expected to be reached by a 100-year flood must be either moved or flood-proofed.

The real estate portfolio’s exposure to transition risks is affected mainly by two factors: regulatory developments in the countries and cities where we operate, and demand in the local market for buildings with green credentials. Our unlisted real estate investments are concentrated in eight cities. One common denominator is that the local authorities have developed long-term plans to cut carbon emissions that will impact on the real estate sector, given that it is often a significant contributor in large cities. This exposes the fund’s real estate investments to regulatory risk in terms of requirements for carbon reductions, which could mean both upside and downside risks for individual investments.

Many of our tenants are global companies that engage in carbon accounting and have set targets to lower their own carbon emissions. This may lead them to look for offices in buildings that are energy-efficient and have low emissions. Research has shown that buildings with green certifications achieve higher sale and rental prices and lower vacancy levels. This means that there is a market risk from declining demand for buildings without green credentials.

3. Integration of climate risk into investment decisions

In general
Within the bounds of its management mandate, the Bank seeks to identify investment opportunities and reduce the fund’s exposure to unacceptable risks. One element in this work is integrating sustainability and climate risk into our investment decisions.

In several parts of the equity portfolio, such as power production, mining and other heavy industry, transport, and oil and gas, developments in climate regulation and new technology are key factors when considering companies’ future earnings. The same considerations apply when investing in corporate bonds. We have furthermore developed internal expertise in environmental technology through our management of the dedicated environment-related mandates. The need for economically, environmentally and socially sustainable development is considered continuously by our portfolio managers as an integral part of investment decisions. This is reflected in the management mandates for both internal and external managers.

When it comes to government bonds, Norges Bank has developed a framework for systematic analysis of the investment risk and operational risk associated with different issuers. Sustainability is one factor included in the assessment of investment risk. This includes consideration of environmental factors, such as exposure to climate change and carbon intensity. These assessments form a basis for which government bonds to invest in, and for the measurement and monitoring of the risks that go with these investments.

Assessment of climate risk is also an integral part of investment decisions when it comes to unlisted real estate. Each investment is preceded by thorough due diligence reviews which include an evaluation of factors related to the property’s location or environmental impact that could expose the fund to physical or transition risks. These factors are also important in the ongoing management of the properties and followed up together with our investment partners and local asset managers.

Environment-related mandates
The management mandate requires the Bank to invest between 30 and 60 billion kroner in dedicated environment-related mandates. At the end of 2018, we had 43.3 billion kroner invested in equities in 77 companies and 13.4 billion kroner in green bonds. The Bank has concentrated its environment-related equity investments on companies in low-emission energy and alternative fuels, clean energy and energy efficiency technology, and technology and services for the management of natural resources. Companies included in the environment-related equity mandates must have at least 20 percent of their business in these areas.

Risk-based divestments due to climate risk
Sustainability considerations may lead us to divest from companies which we believe pose a particularly high long-term risk. In this context, we have considered the long-term risk associated with carbon emissions from companies in the portfolio, with a focus on the highest-risk sectors. Companies that have operations or value chains with particularly high carbon emissions may be exposed to considerable transition risks. This could lead to higher operating costs and/or reduced demand. We also look at business activities that result in deforestation, which is related in turn to climate risk. Between 2012 and 2018, the Bank made risk-based divestments from 142 companies based on an assessment of climate risk. These were mainly small companies that have particularly high carbon emissions (82 companies) or contribute to deforestation (60 companies). These divestments are carried out within the general limits for the management of the fund, which means that the Bank is taking investment decisions that result in deviations from the benchmark index. This gives the fund slightly different exposure to climate risk than the benchmark index.

Exclusion and observation
The ethically motivated guidelines for observation and exclusion of companies include criteria targeting mining companies and power producers that either derive 30 percent or more of their revenue from the production of thermal coal or base 30 percent or more of their operations on thermal coal. A total of 70 companies have been excluded from the fund under these criteria, with a further 13 placed under observation. From 1 September 2019, companies may also be excluded or placed under observation if they produce more than 20 million tonnes of thermal coal per year or have coal-based power generation capacity in excess of 10,000 MW. The exclusion of companies under the coal criteria also results in removal from the benchmark index. Under the current rules, exclusions under the coal criteria make the single largest contribution to reducing carbon emissions from companies the fund is invested in.

4. Tools for working on climate risk in the portfolio

In our active ownership work, we consider aspects of financial risk relating to the environmental and social conduct of companies we are invested in. In our work on climate risk, we prioritise companies with the highest carbon emissions in their operations or value chains. We base our work on available data and relevant research. In recent years, we have strengthened the parts of our organisation working on active ownership and risk assessment related to environmental, social and governance factors.

Understanding of how sustainability and climate risk might impact on financial risks and returns is evolving. We support and initiate research projects that contribute to this process. As examples, we can mention two projects from 2015 and two ongoing projects launched in 2017.

In one of the projects in 2015, we initiated research into the long-term risks that climate change poses for financial investors. We provided financial support for the first scientific conference on stranded assets – assets that may lose value due to climate change. The conference was arranged at the University of Oxford. Later that year, we asked the Smith School of Enterprise and the Environment at the same university to produce a report on various risk factors for companies with coal-based operations. The analysis included a forward-looking assessment of environmental risks for coal companies, including the pricing of carbon emissions and technological changes.

Since 2017, we have been supporting two research projects looking at the financial consequences of climate change. New York University Stern School of Business’s Volatility Institute is conducting research into environmental risks in the financial sector. Led by Nobel laureate Robert Engle, the project is looking at methods for measuring and modelling environmental risks, and how modern risk management techniques can be used to account for climate risk in the composition of investment portfolios. The researchers are also working on improving estimates of the long-term discount rate when investing in climate projects. As part of the project, the institute publishes regularly updated climate risk data on its V-Lab website.

We are also providing funding for Professor Harrison Hong at Columbia University to conduct research into climate change and capital market efficiency. The aim is to encourage more leading finance scholars to look at climate issues. Together with the Review of Financial Studies, the project has selected ten promising research ideas which have been discussed with peers at two conferences. The researchers are looking at how companies and markets form climate change expectations, and how pricing is affected by behaviour and trading decisions.

Work on standard setting
Our work on standard setting means contributing actively to the further development of market standards for sustainable development and well-functioning markets. As part of this work, we support the development of better standards for reporting on sustainability and climate risk. We have argued that the harmonisation of frameworks and consolidation of disclosures are in the interests of both companies and investors. This will provide a better basis for comparisons between companies, countries and regions, and simplify company and portfolio analysis for investors. It will also reduce the overall reporting burden on companies.

In recent years, we have contributed to consultations from CDP as part of its development of reporting standards for environmental issues and climate change. We have provided input for the TCFD in its development of a general framework for reporting climate risk, and we are contributing to the development of standards for reporting on sustainability under the SASB. We have expressed our support for a sector-based approach to reporting on financial climate risk and promoted simplification, improved data quality and increased relevance to investors.

Since 2018, we have supported the Transition Pathway Initiative. Led by institutional investors, this initiative is developing tools for analysing and measuring transition risks in various sectors. The work is being supported by asset managers and academia, including the London School of Economics, and the tool is publicly available.

To help develop standards for responsible real estate management, we are part of a working group under the Better Buildings Partnership in the UK focusing on climate adaptation in the real estate sector. The group aims to facilitate the exchange of information on the measurement and management of physical climate risks.

Expectations of companies
Since 2008, our active ownership efforts have included the formulation of expectations of the companies we invest in. The aim is to set out how we, as an investor, expect portfolio companies to respond to global challenges in their operations. Our approach builds on international standards. Altogether, we have published seven sets of expectations, including on climate change strategy (2009), water management (2010) and ocean sustainability (2018). One common denominator is our expectation that the company boards take overall responsibility for company strategy and address relevant sustainability challenges. The boards should integrate material factors into their risk management and reporting.

In the expectations on climate change strategy, the Bank notes that climate change will eventually impact on most industries and markets. We expect companies to consider climate risk in their operations and plan for relevant climate outcomes. We also highlight the importance of research and development so that companies can take appropriate action on climate change. We refer to measures such as risk adaptation and mitigation, such as better energy and resource efficiency, use of less carbon-intensive raw materials, optimisation of logistics and distribution, protection of landscapes, operational adjustments and other measures to make their business more robust to climate change. In addition, we encourage reporting on climate risk and transparency on contact with regulators concerning climate legislation.

We update our expectations regularly. In 2018, the expectations on climate change strategy were updated to clarify our position on banks’ role in financing coal activities. We expect those that continue to finance coal to be open about their lending policies and the assessments they perform before approving new loans. We also request disclosure on other lending to activities based on fossil fuels, such as mining, infrastructure and power. At the same time, we ask for information on loans for renewable energy and other climate-related opportunities and development projects. This provides us with a basis for assessing banks’ exposure to transition risks. In addition, we call for reporting on physical risks and express an expectation that companies disclose information on assets and facilities, including location and technical data.

Company reporting
We encourage companies to move from words to numbers so that we can better evaluate their operations and the financial risks and opportunities associated with their environmental and social conduct. We have been assessing companies’ climate risk disclosures since 2010. We monitor companies’ reporting on climate risk in the following sectors: automotive, construction and building materials, banking, insurance and financial services, real estate, consumer goods, travel and leisure, industrial goods and services, chemicals, food and beverage production, basic materials, technology, utilities, and oil and gas. In 2019, we will analyse around 3,950 company reports, including 1,500 for disclosure on climate change and 250 for disclosure on deforestation. Companies’ reporting is assessed against indicators for corporate governance, strategy, risk management and performance reporting.

We reach out to companies with poor or limited disclosure. Of the companies we contacted about weak reporting on climate change in 2017, 36 percent began to report on this area in 2018. Our general observation is that there is considerable variation in levels of climate disclosure between both companies and sectors. We have seen some improvement in companies’ reporting on climate change in recent years.

As we are a shareholder in more than 9,000 companies, voting is an important means of exercising our ownership rights. Shareholders are increasingly raising sustainability issues and environmental risks with companies by putting forward proposals for action or change at general meetings. These include proposals for climate disclosure. In recent years, we have backed a number of shareholder proposals for companies to plan for and report on different climate outcomes for their operations where we believe that the proposal furthers the fund’s long-term financial interests.

Dialogue with companies
As a large, long-term shareholder, we engage regularly with the companies in our portfolio. In 2018, we engaged in dialogue with 272 companies on climate change and deforestation. The focus was on companies in the industrial, consumer goods, energy, basic materials, oil and gas, and financial sectors. We expressed our expectation that companies should plan for relevant climate outcomes and assess climate risk in their operations. For example, we spoke to banks in the portfolio about climate-related disclosure and the introduction of the TCFD framework. The aim of this dialogue was to promote more relevant reporting and gain a better understanding of how banks are using the recommendations. We also engaged with banks in Asia and Latin America on financing of companies that contribute to deforestation and how they formulate their lending policies. We urge companies to pay attention to climate issues and have sufficient flexibility to adapt to the low-carbon transition.

We initiated dialogue with companies in additional sectors in 2018 and 2019, including cement, shipping, automotive, cocoa and clothing. All generate substantial carbon emissions and also face other big challenges in terms of the environment and human rights. In the dialogue we raise these companies’ management of climate risk, including their plans to cut carbon emissions and manage the transition to a low-carbon economy. We have also supported initiatives that bring companies in the same industry together to find joint solutions and standards for sustainable business conduct. The need for standardisation and more universal approaches is considerable. One example of these initiatives is the UNEP FI pilot project on TCFD reporting mentioned above.

Tools in the unlisted real estate portfolio
We work together with investment partners and asset managers to integrate environmental measures into the business plans for the properties in our unlisted real estate portfolio. Better measurement and access to high-quality data are a priority. We have published guidance for our partners and asset managers setting out what we consider to be sound principles for sustainable real estate management.

We are developing a platform for gathering data on properties in the portfolio, including information on energy and water consumption, waste management and environmental certification. This information will be used to measure the environmental impact of our assets, assess possible environmental measures, and produce climate accounts for our unlisted real estate investments. In addition, we monitor the share of our unlisted real estate investments that are in areas considered by the local authorities as being at particular risk of flooding, based on whether or not they have been flooded at least once in the past century. We have also taken steps to protect relevant buildings against flood damage, such as installing temporary flood barriers, moving technical installations to higher storeys and increasing our insurance cover.

Environmental certification is set to be ever more important for properties’ competitiveness. This involves an independent third party assessing a property against a set of criteria, such as energy and water consumption, use of renewable energy, and proximity to public transport. The certificate confirms that the property meets the chosen environmental standards. Our long-term goal is for all office and retail properties in the portfolio to be certified. Each year, we assess all of our unlisted real estate investments against the Global Real Estate Sustainability Benchmark (GRESB). This process includes information on carbon intensity and the use of renewable energy.

5. Plans to further develop work on climate risk

The Ministry has asked Norges Bank to set out any plans to further develop work on climate risk in the fund within the bounds of the stipulated investment strategy for the fund and the management mandate from the Ministry. The Bank’s work on climate risk has been under development since 2006. We have taken a broad approach to the management of climate risk as a financial risk factor for the fund. Assessment of climate risk is integrated into our risk management, our investment decisions and our active ownership.

Information on companies’ carbon emissions and how their operations might be affected by climate risk (both physical and transition risks) is critical for our analysis of climate risk in the portfolio. Access to relevant, high-quality data is crucial for this work. As an investor in more than 9,000 companies in more than 70 countries, we need to be able to quantify and analyse the risks facing the fund. There are numerous useful initiatives currently under way to improve and standardise company reporting, but there are still variations in the frequency and quality of corporate disclosure. We will continue to support initiatives to develop global standards for this reporting.

We will also continue to work on developing analytical tools for climate scenarios. We will expand and improve our internal analytical models for different climate outcomes by integrating more detailed company-specific data, such as financial data, emissions data, concrete targets for carbon reductions, and exposure to regional carbon-pricing mechanisms. In addition, we will look at how we can develop methods for analysing physical climate risks facing companies in the equity portfolio.

As a shareholder, we expect the boards of companies in the portfolio to take overall responsibility for the company’s strategy and manage sustainability and climate challenges. We will continue to engage with companies in high-risk sectors based on our expectations. In this dialogue, we will be clear about our expectations for climate risk management and carbon reductions. We will also strengthen our work on following up companies’ reporting on climate risk.

When it comes to real estate, work on assessing physical risks will remain crucial. We plan to expand our flood risk analyses based on historical data with forward-looking scenario analyses. We will also continue to support a joint initiative to develop methods to analyse the expected decarbonisation of the real estate market. The results of this work will be relevant for assessing transition risks for the fund’s real estate portfolio.


Yours faithfully


Øystein Olsen                                                                                                       Trond Grande


[1] Value at Risk is defined as the potential loss on a portfolio with a given probability over a given time period.

[2] Report available at: