Banks increase absolute CO2 financing, but overall transition risk exposures drop

• Euro area investors also seem to buy most green bonds domestically, while social and sustainability bonds are largely bought outside the eurozone

• The data also show that banks increased financing of carbon emissions in 2021, although this was largely due to a pick-up in economic activity following the outbreak of the pandemic in 2020

• In relative terms, banks reduced their climate-related risk exposures in 2021, a trend set to continue in the coming years

• Finally, the report shows that physical-related climate risk can be effectively mitigated by taking climate adaptive measures

The ECB published the second release of a paper on climate-change related statistical indicators last month, following its initial launch last year. The update includes some statistical improvement, although lack of data remains the key impediment. Nevertheless, the data is important as it provides more insights into issuance and holdings of sustainable (green, social and sustainability) debt securities by euro area residents. Furthermore, carbon emission indicators help to assess the exposure of the financial sector to climate-change related risks via its loan as well as investment portfolios, while also evaluating financial sector’s involvement in financing the energy transmission. Finally, the dataset includes information about the financial sector’s exposure to physical climate risks. In the end, these data is used to enhance transparency on climate risks for financial stability, supporting banking supervision and, to a lesser extent, monetary policy. Overall, it fits within the proactive stance that the ECB has taken regarding climate change.

Issuance and holdings of sustainable debt securities on steady growth path

The data on issuance of sustainable debt securities by euro area entities shows that this market has been steadily growing, having reached a level of more than EUR 1,200bn by end of Q3 2023. Having said that, this only accounts for around 6% of total debt issuance, indicating that the sustainable bond market is still a niche market. The split by debt type shows that most bonds are in green format, followed by social bonds and sustainability bonds (which are a blend of ‘green’ and ‘social’). Interestingly as well, the ECB also looked at the proportion of sustainable bonds verified by second-party opinion (SPO) providers, a feature that adds to the strength of a sustainable bond given that an external party checks whether the bond is aligned with, for instance, the ICMA Green or Social Bond Principles. According to the dataset, 85% of sustainable debt issued in the euro area had obtained a second-party opinion, with green bonds having the largest share. Similarly, most sustainability-linked bonds (80%) also benefitted from an SPO.

The ECB’s dataset also includes information of holdings of sustainable debt instruments in the euro area. Although the data mirrors to a large extent that of issuance, it contains some interesting features. One is that euro area holdings of sustainable debt exceeds that of the region’s issuance of these instruments, implying that the euro area is a net buyer of sustainable debt securities. This could be a supportive argument for why sustainable debt instruments have historically traded at a premium over regular instruments (the so-called greenium). In any case, the euro area is an important buyer of sustainable funding for the rest of the world. Meanwhile, the data suggests (based on SPO figures of sustainable debt instruments) that euro area investors mainly buy green bonds as well as sustainability-linked bonds domestically (i.e., from euro area entities), while they buy a relatively large share of social and sustainability bonds abroad.

A breakdown by country and sector shows that Germany, France, and the Netherlands are the largest issuers as well as investors in green bonds, with Luxemburg also scoring well in terms of green debt holdings (reflecting that many investment firms are in the country). On a sector level, governments and monetary financial institutions are the key issuers of green bonds, while investment funds, insurance companies and central banks are the largest investors. The latter likely reflects the sizeable bond purchases under the Eurosystem’s asset purchase programmes (and the central bank’s aim to green its corporate bond portfolio). Looking forward, the share of the central banks is likely to gradually diminish now that the ECB has stopped reinvestments of its APP holdings, while it has scaled down PEPP reinvestments.

Carbon emission financing calculated

The second part of the data relates to carbon emissions, focussing on the direct emissions financed by the financial sector as well as the exposure of the financial sector to counterparties that are engaged in emission-intensive activities. Below we give a brief wrap-up on how the ECB calculates the key indicators:

FE is an indicator of absolute emissions (how much of the company’s total emissions can be attributed to the financed institution financing it). On the other hand, CI and WACI are mostly intensity indicators, which means that these refer to not absolute emissions, but rather how these emissions compare to the company’s revenues. Specifically looking at CI, this indicator represents a financial institution’s share of emissions (that is, proportional to the investment into the company), relative to its share of a company’s revenues. On the other hand, WACI focuses on how much weight a company’s emission intensity has in the financial institution’s portfolio. Hence, while CI takes into account only the share of emission intensity, WACI puts that into perspective by weighting that share across the financial institution’s portfolio value. The same is true for CFP, which is also takes into account the portfolio value. However, contrary to WACI, CFP uses as numerator the company’s absolute emissions.

Overall, FE and CI are indicators of how much a financial institution’s contributes to a company’s emissions, representing therefore indicators of how the financial sector contributes to the financing of high emitting economic activities. On the other hand, WACI and CFP are better indicators of a financial institution’s exposure to transition risks, as they take into account the financial institution’s portfolio value.

Financed emissions rose in 2021, likely related to lower headwinds from the pandemic

The graph on the next page depicts the scope 1 emissions financed by euro area banks (left) and the sectoral breakdown of the emissions during 2018 and 2021. The data is based on a single-entity level and not on a group level. It shows that the downward trend in 2019-2020 was broken in 2021, when banks increased financing of carbon emissions. However, the Covid-19 pandemic has likely impacted the data, as economic activity was strongly supressed in 2020, driving down emissions. Indeed, as shown here, global emissions increased by nearly 5% in 2021 in comparison to the previous year (for the EU this was 6.5%). As result, the uptick in 2021 should be merely seen as payback for the unusual strong downturn in 2020. Correcting for this, the downward trend probably continued in 2021, assuming 2020 will turn out to be the outlier.

The reduction in financed carbon emission during 2019-2021 was 25 million tons of CO2, or 17% of financed emissions in 2018. A sectoral breakdown reveals that the primary production sector accounted for the largest drop so far, although it accounts only for 13% in the total. Unsurprisingly, manufacturing (34%), energy (18%), and transport (17%) are sectors that emit most CO2. All three sectors saw only a 0.1% reduction in financed emission between 2018 and 2021.

The rise in financed emission is not visible in the intensity indicators, such as the CFP and WACI, as these also consider the revenues of the companies that banks lend to as well as the value of a bank’s investment portfolio. Indeed, in 2021, a rise in corporate revenues and/or the value of bank investment portfolios compensated for the increase in financed emissions, resulting in intensity indicators continuing on their downward pathways.

A further decomposition of the changes in the WACI confirms the above. The charts below show that the rise in financed emissions via bank loan portfolios was more than compensated for by the rise in corporate revenues. On the investment side, there was only a marginal increase related to more emissions, while the increase in the value of investment portfolios resulted in a further decline in the indicator in 2021. This can largely be explained by a pick up in economic activity in 2021. Overall, it suggests that banks have become less exposed to climate-related risks in their loan books as well as their investment portfolios in 2021, despite the fact that financing of absolute levels of carbon emissions actually rose.

This is also a good illustration of the large role that Covid-19 played in the data exercise. Looking forward, not only should the impact of the pandemic on the data peter out, but it is also likely that companies improved their energy efficiency in the past two years, following the sharp increase in energy prices in 2022. This will probably have resulted in a drop in corporate carbon emissions. At the same time, it is likely that banks have also become more selective in lending policies, preferring to lend more to less carbon-intensive companies, which up to 2021 was not a real factor at play according to the report. As such, we expect the FE, WACI and CFP indicators to have continued on the downward trend in 2022 and 2023.

Nevertheless, as several other ECB studies (see for example here and here) have indicated, transition risks for banks are still very elevated. For example, a study by the ECB showed that 90% of the supervised banks by the ECB face elevated transition risks. Interestingly as well, a recent paper by the European Commission attributed the slight decrease in transition risks from banks towards a shift from exposure to less regulated financial institutions (see here). More specifically, the study shows that the exposure to transition risk of less regulated financial institutions has more than tripled from 2014 to 2023, indicating that transition risks might be shifting to less regulated parts of the financial system.

Adaptation measures effective in reducing physical climate risks

The third part of the data exercise is devoted to the impact of physical climate risks on banks’ loan and investment portfolios. It takes a bottom-up approach starting with the location of a debtor (in case of loans) or an issuer (in case of securities). It then assesses the risks related to several climate-change induced natural hazards (such as coastal flooding, wildfires, landslides, river flooding, subsidence, and windstorms), followed by an estimate of the total exposures at risk. Finally, so-called ‘damage-functions’ are used to translate hazard intensities into monetary losses. One of the most insightful results is that the quantification of the impact that protection measures can have on expected losses. This is especially true for flood risks. The charts below show that, especially in case of coastal flood risks, potential damage can be reduced by up to 100% when flood defence systems have been put in place. In case of the Netherlands, this reduces the potential exposure at risk (or PEAR) from EUR 65bn to around EUR 3bn. This stresses the need to invest in climate adaptation measures, next to climate-change mitigating measures. The result also fits recent media reports that water shortages are the country’s biggest climate risk.