Stranded fossil-fuel assets translate to major losses for investors in advanced economies

The transition to a low-carbon global economy necessitates rapid and extensive structural changes, posing significant economic adjustment challenges. A key challenge for both the real economy and financial markets is the rapid phase-out of fossil fuel production, requiring the write-down of substantial functioning capital assets and reserves. While numerous studies have analyzed the early retirement of fossil fuel facilities under various scenarios, these studies have not adequately linked this retirement to financial ownership. Consequently, stress tests of the financial system often begin with synthetic shocks to financial assets rather than the underlying real assets. The distribution of financial ownership and exposure to loss risk remains poorly understood. Asset stranding, the collapse of future profit expectations from invested capital due to disruptive policy or technological change, directly impacts investor expectations of enterprise value and market prices. This results in wealth losses for asset owners and can propagate through interconnected financial networks. When these effects destabilize financial markets, asset stranding becomes a significant social concern, impacting pensions and government finances. While capital stock obsolescence is common in dynamic capitalist economies, the scale of industrial change needed for a 2°C or 1.5°C warming goal is unprecedented, creating substantial transition risks. This research comprehensively maps the global financial geography of stranded oil and gas asset risk focusing on equity ownership. It traces potential losses from extraction sites to ultimate owners (governments and individual shareholders) through corporate headquarters and shareholders (including banks and fund managers). The study links fossil fuel stranded assets and transition risk studies at the asset level to understand the equity mispricing transmission channel, distinguishing geographic and functional characteristics of organizations along the ownership path. This detailed analysis reveals that exposure to wealth losses is geographically more evenly distributed than the distribution of oil and gas production assets suggests, implying a greater stake and exposure for private investors in wealthy countries than previously understood.

Over 100 studies have analyzed scenario-contingent early retirement of fossil-fuel supply facilities. However, these studies have not been effectively linked to financial ownership. As a result, existing academic and regulatory studies that conduct stress tests of the financial system start from synthetic shocks to financial assets, not the underlying real assets. The distribution of financial ownership and exposure to loss risk remains insufficiently understood. Previous studies have also focused on producer countries, neglecting the broader picture of financial ownership and the international transfer of risk.

The research operationalizes asset stranding as the impact of changing expectations on the present value of discounted future profit streams. It calculates profits based on expectations per asset using data from Rystad's Ucube dataset, which details 43,439 oil and gas production assets. Whether an asset is expected to supply demand depends on its production cost and reserve profile relative to the expected market-clearing oil price. A fall in investor expectations for total oil and gas demand renders some assets unprofitable. To quantify profit losses, the study uses a baseline scenario (IEA's WEO 2019 current policies scenario, consistent with 3.5°C median warming) and a policy scenario (incorporating EU and East Asia's net-zero policies by 2050/2060, leading to 2°C median warming). The expectations shift (realignment) is assumed to occur in 2022, and three additional realignments are considered to assess sensitivity. Oil and gas demand and prices are generated using the E3ME-FTT-GENIE integrated assessment modelling framework, which couples macroeconomic modelling, energy technology modelling, and climate system modelling. The study traces the ownership chain in four stages: 1) loss at the oil/gas field (by country); 2) loss at headquarters (by country); 3) corporate owner loss (by country and sector); and 4) ultimate owner loss (by country and institutional affiliation). Data on company financial and ownership are sourced from Bureau van Dijk's ORBIS database. A network model propagates financial losses through the equity ownership network. Missing company data are imputed using statistical models. The model accounts for technical insolvencies, where losses exceed equity, and analyzes the amplification of losses in financial markets through cascades of stock market losses and the impact on financial institutions' balance sheets. Sensitivity analyses explore the robustness of the findings across different realignments and network properties.

The study estimates that under a medium realignment scenario, the present value of stranded assets in the upstream oil and gas sector exceeds US$1.4 trillion. This loss is disproportionately borne by private investors in OECD countries. A significant portion (more than 15%) of the global stranded asset risk is transferred internationally to OECD-based investors, highlighting the considerable stake that rich countries have in managing the energy transition. Geographically, losses are initially concentrated in the United States and Russia but significantly shift to OECD countries as the ownership chain is traced. The financial sector, particularly in OECD countries, bears a substantial portion of the loss. At the ultimate owner level, losses are largely borne by governments and individual shareholders, including pension funds. Listed companies own the majority of stranded assets, with an additional 19% of losses emerging through the ownership chain. The financial system's interconnectedness amplifies losses. Potential losses affecting financial companies could reach US$681 billion, with funds being significantly exposed. Technical insolvencies add to credit risk. The United States and the United Kingdom are the most affected countries, both in terms of physical losses and financial exposure. The study’s findings are robust across different realignment scenarios and sensitivity checks. Comparing lost revenues (to cover wages and supplier payments) to lost profits reveals that revenue losses are on average four times larger than profit losses. However, losses relative to GDP are largest in oil-exporting developing countries.

The findings challenge the conventional focus on producer countries as the primary bearers of stranded asset risk. The study demonstrates that a significant portion of the risk is borne by investors in OECD countries, highlighting the potential for perverse incentives to slow the energy transition. This raises concerns about the political influence of financially exposed groups lobbying for government bailouts, a moral hazard that could delay necessary expectation realignments and prolong fossil fuel investments. The research emphasizes the limitations of using domestic sectoral exposure as an indicator of financial risks from asset stranding. International linkages significantly increase the risk of financial instability. Stress tests and scenario exercises should incorporate variable risk distributions within sectors, not just across them. The significant exposure of pension funds is a major concern, given their relatively less stringent regulation and lack of a full understanding of their role in systemic contagion. International cooperation is crucial in managing the energy transition and mitigating its destabilizing social and economic consequences.

This study reveals the global financial geography of stranded fossil fuel assets and its implications for investors in advanced economies. The significant transfer of risk from producer countries to OECD investors highlights the need for policy intervention, particularly to manage the moral hazard of potential bailouts. The findings underscore the importance of considering the complex international financial network in assessing and mitigating the risks of the energy transition, improving stress testing methodologies and encouraging international cooperation to manage the phase-out of fossil fuels.

Data limitations exist regarding the precise allocation of losses at stages 3 and 4 due to incomplete information on ultimate owners in the public domain. The study focuses on equity transmission channels, neglecting potential amplification through debt channels. The assumption of a 6% discount rate could influence the magnitude of estimated losses. Imputation of missing company data might introduce some uncertainty, although sensitivity analyses suggest the robustness of the results at aggregated levels. The model uses a specific integrated assessment framework; changes in IAM assumptions could change the overall outcome.