Too big to strand? Bond versus bank financing in the transition to a low carbon economy

Winta Beyene, Manthos Delis, Kathrin de Greiff, Steven Ongena 04 December 2021



In the ongoing debate on the need for a transition towards a decarbonised economy and the actions that should be undertaken by central banks, financial authorities, and governments, the role of bank and bond market financing is of paramount importance. Financiers could play an important role in terms of channelling funds away from fossil fuels and polluting types of activities and investing in greener activities (Caselli et al. 2021). However, fossil fuels still dominate energy investments, and particularly, banks still show unwavering interest in fossil fuel projects (e.g. RAN 2020, Pinchot and Christianson 2019, Delis et al. 2018). A major concern in the transition to low-carbon energy provision, therefore, is to steer investments away from fossil fuels.

In light of this, there is a real risk that large investments in fossil fuel companies will decrease in value and result in bad loans when climate policies finally tighten (Löyttyniemi 2021). Stranded asset risk – the risk related to the re-evaluation of carbon-intensive assets as a result of this transition away from a carbon economy – needs to be reflected in the fossil fuel firms’ cost of debt to compensate for the increased risk of default. 

In a recent paper (Beyene et al. 2021), we examine the potentially different roles of market-based versus bank-based credit in the allocation of resources to fossil fuels. We do so by investigating fossil fuel firms’ cost of corporate bond financing versus syndicated bank loan financing, and the consequent composition of these two debt types along these fossil fuel firms’ risk of seeing part of their assets stranded. Following the observation that bank financing on average has not decreased with stricter climate policies, we investigate the question of whether stranded assets risk is increasingly concentrated in a few large exposures for some large banks? 

Our dataset consists of corporate bonds and syndicated bank loans issued from 2007 to 2017 by firms that have had access to both markets during that period. Fossil fuel firms’ risk of stranded assets is proxied with the variable ‘Climate Policy Exposure’, which is constructed as the product of a country’s climate policy stringency and the relative amount of reserves a firm has in this country. The relative reserves of firms we hand-collect from firms balance sheets, and to measure a country’s climate policy stringency we use mainly the Climate Change Policy Index (CCPI) by Germanwatch (Burck et al. 2016). Large energy companies are going to have reserves in different countries, and these reserves are going to be exposed to differential climate policy stringency, which is what Climate Policy Exposure captures. While the finance literature on the topic of carbon emissions-related risks has largely been focused on firm-level emissions, focusing on fossil fuel firms’ holdings of fossil fuel reserves, and the risk stemming from this, is closer to the root of the problem. Much of the global stock of carbon emissions can be traced to a small set of largely fossil fuel companies (Ilhan et al. 2020).

The analysis is carried out in four parts. First, we look at the pricing of stranded assets risk of fossil fuel firms by the corporate bond market and by banks. We find that newly issued corporate bonds in the fossil fuel industry have higher yields than syndicated bank loans, and with increasing climate policy exposure, bond markets earn a higher premium relative to the syndicated bank loan-implied credit spread. Second, we show that fossil fuel firms shift from issuing bonds to obtaining bank loans as their stranded asset risk exposure increases. Third, we show that bond-to-bank substitution is unlikely to arise from differences between banks that underwrite corporate bonds and banks that underwrite syndicated bank loans, and ultimately from a resulting difference in the quality of borrower. For this, we collect information on lead manager banks, combine the loan and bond subsamples, and construct a dataset in which the same banks are observed to engage in corporate bonds and in syndicated bank loans as lead manager in order to control for the underwriter. Fourth, we look at whether bank characteristics related to bank size may influence banks’ reaction to stranded asset risk impulses in terms of lending and risk-taking.  We find that across all syndicated loans, large banks acting as lead managers charge a lower all-in spread drawn than small banks do, and consequently there is a migration towards the very largest lead manager banks along fossil fuel firms’ Climate Policy Exposure.

Figure 1  Credit allocation towards fossil fuel

Figure 1 is an illustration of some parameters of fossil fuel debt and summarises our findings.  We assume that an increase in Climate Policy Exposure implies an increase in the expected loss. Hence, to cover the expected loss on a debt, the lender needs to apply a higher interest rate.  We conclude from our findings, however, that for at least large banks, the expected gains from an increased investment today may in some ways still compensate for the expected loss due to the risk of stranded assets. Therefore, while the corporate bond market requires rBond, which accounts for firms’ risk of stranded assets to some extent, banks require only rLoan. Consequently, this differential in the pricing of the risk of stranded assets implies that banks continue to finance the fossil fuel projects that the corporate bond market would not, as visualised by the red area.  The same figure can be applied to illustrate the migration of stranded asset risk towards big banks within the banking sector. 

The level of fossil fuel financing from the world’s largest banks remained in 2020 higher than in 2016, the year immediately following the adoption of the Paris Agreement (RAN 2020). Our findings add to the limited literature on the impact of stranded asset risk on firms’ (bank) funding cost and provide empirical evidence for this narrative that there is a migration of fossil fuel stranded assets risk away from markets and towards (large) banks. On the subject of stranded assets risk and debt, our paper provides the following new insights:

  • Market discipline, on its own, seems to be more effective in driving bondholders, rather than banks, to price the negative externalities associated with the risk of stranded assets. 
  • The ability of large banks to hold large exposures to firms with stranded asset risks may prevent the steering of investments away from fossil fuels.  
  • A substitution mechanism between bond and bank financing, or even within the banking industry, could potentially mitigate the capital constraints on fossil fuel firms imposed by the bond market or some more ‘environmentally friendly’ banks. 


Burck, J, L Hermwille, and C Bals (2016), “CCPI background and methodology”, Germanwatch and Climate Action Network Europe.

Beyene, W, M Delis, K DeGreiff, and S Ongena (2021), “Too-big-to-strand? Bond versus bank financing in the transition to a low carbon economy”, CEPR Discussion Paper 16692.

Caselli, F, A Ludwig and R van der Ploeg (eds) (2021), No Brainers and Low-Hanging Fruit in National Climate Policy, CEPR Press.

Delis, M, K DeGreiff, and S Ongena (2018), “The carbon bubble and the pricing of bank loans”,, 27 May.

Ilhan, E, Z Sautner, and G Vilkov (2020). “Carbon tail risk”, The Review of Financial Studies 34(3): 1540-1571. 

Löyttyniemi, T (2021), “Integrating climate change into the financial stability framework”,, 08 July.

RAN (2020), Banking on Climate Change, Fossil Fuel Finance Report 2020.

Pinchot, A and G Christianson (2019), “How are banks doing on sustainable finance commitments? Not good enough“, World Resource Institute, 3 October.



Topics:  Environment Financial markets

Tags:  climate change, fossil fuels, CO2 emissions, asset risks, low carbon, banks, bond market financing

PhD student in Finance, University of Zurich and Swiss Finance Institute

Professor of Financial Economics, Montpellier Business School

PhD candidate in Banking and Finance, University of Zurich and Swiss Finance Institute

Professor in Banking, University of Zurich, the Swiss Finance Institute, KU Leuven and NTNU; Research Fellow, CEPR


CEPR Policy Research