PhD Chapter - OECD report on promoting green finance and macroprudential

Central banks’ mandates do not explicitly embed objectives related to climate change

Risks related to climate change were not treated with the same sense of urgency when central bank mandates were initially defined in the countries of Emerging Asia as they are today. Nevertheless, some rules and principles that define central banks’ remit, and state the limits of their responsibility to address future challenges such as climate change, are embedded in the mandates of several central banks in the Association of Southeast Asian Nations (ASEAN) (Table 2.1). The examples of the central banks of Indonesia, Malaysia, Myanmar, the Philippines and Singapore are very relevant, as their mandates embed support for the government’s economic policy, which includes sustainable growth. Cambodia’s central bank is also tasked, albeit in a less explicit manner, with conducting its monetary policy in a way that facilitates sustainable economic development, in line with the country’s overall economic and financial policy.

These underlying rules determine rather general obligations and limits on how each Emerging Asian central bank must contribute to the urgent need of tackling climate change. Central banks in the region should take decisive action wherever their financial-stability mandate overlaps with climate change.

Physical and transition risks related to climate change

The increasing severity and frequency of natural disasters stemming from climate change is likely to have an adverse impact on the financial sector, and indeed on financial stability. Risks to the financial system and to financial stability can be grouped into two broad categories. The first encompasses physical risks, while the second includes risks that stem from the transition to a low-carbon economy. Physical risk results from the direct impact of climate change on people and assets. When it materialises, physical risk could have three types of consequences for the financial sector. First, it could lead to an erosion of the value of the assets and collateral that have been pledged in credit transactions, when they are located in disaster-prone areas. The second type of impact is the increase in damage that must be covered by the insurance and re-insurance sector. Third, physical risk stemming from climate change could lead to a deterioration of local economic activity, which could then impact the solvency of borrowers.

In addition, the objective of mitigating climate change also exposes the financial sector to risks that are related to the transition towards a more sustainable economy. These risks may be broadly defined as the uncertain financial impacts on economic agents, both positive and negative, that will result from the implementation of a low-carbon economic model. Transition risks take several forms, namely: risks posed by policies aimed at decreasing greenhouse gas (GHG) emissions to meet the 2 degree target by the end of the century (e.g. carbon prices); legal risks arising as a function of climate litigation (e.g. in the context of climate damages); and technology risks that relate to the uncertainty in technological development and deployment. These risks are characterised by a very high level of uncertainty about the trajectory of the transition, notably the speed of reduction of greenhouse gas emissions, which will have a restructuring effect on the economy. Transition risk could have a material effect on the assets held by banking institutions, and on the investment side for insurance corporations. Table 2.2 summarises the major categories of financial impact from physical and transition risks.

Ultimately, there are a number of channels through which physical and transition risks could impact the financial system and financial stability. The impact of these risks will depend on the scale and scope of the actions that are implemented to tackle climate change, as well as on the speed of their implementation. In this regard, the Network of Central Banks and Supervisors for Greening the Financial System underlines in its progress report published in October 2018 (NGFS, 2018[3]) that the risks associated with climate change are a source of financial risk. It calls for central banks and supervisors to ensure that the financial system is resilient to these risks. Among the first recommendations issued by this network, in April 2019, was for microprudential supervision and the monitoring of financial stability to take account of the risks associated with climate change (NGFS, 2019[4]).

Owing to the challenges inherent in their quantification, there is currently little quantitative evidence of the impact of climate-related risks on the financial sector. However, the insurance stress test carried out by the Bank of England’s Prudential Regulation Authority in 2019 also integrated three types of climate-related scenarios (PRA, 2019[5]). The results of these stress tests provide an indication of the likely impact that the risks posed both by transition and by the physical impact of climate change may have on the investments of general- and life-insurance companies. In the case of physical risks, the impact ranges from -5% to -30%, depending on the sector, and assuming an orderly transition to a low-carbon economy by 2050. In the worst-case scenario, which assumes no transition, and a temperature increase of 4°C by 2100, the impact would be even more severe, ranging from -10% to -60% (Figure 2.1).

Figure 2.1. Estimation of losses related to physical risks for UK insurance companies, under different climate transition scenarios

Impact on investments in various economic sectors

Note: Scenario A describes a rapid and disorderly policy action with shock parameters set to hit in 2022. Scenario B describes an orderly transition, which assumes carbon neutrality in 2050. Scenario C assumes no transition and a temperature increase of 4°C by 2100. All three scenarios reference temperature targets that reflect different underlying greenhouse-gas emission pathways, and which are assumed to impact firms at different points in time (2022, 2050 and 2100).

Source: PRA (2019[5]), Life Insurance Stress Test 2019: Scenario Specification, Guidelines and Instructions, Prudential Regulation Authority, Bank of England, London, .

 StatLink https://stat.link/nsi1zy

Some of these climate-related risks have started to materialise in several Emerging Asian countries. For instance, droughts have affected business operations in the Indian energy sector, while emissions regulations that apply to the chemicals sector in China have translated into lower operating rates at several chemical producers in the country (Box 2.1).

The measurement of risks related to climate change poses various methodological challenges

The measurement of climate risk is still a developing field in the area of quantitative research. Even though climate risk intersects with the different categories of risk to which banks are exposed, such as credit, market, operational, and sovereign risks (Table 2.3), current models fail to capture climate risk in its entirety. For instance, climate-related risks can lead to credit risk as they can cause deteriorations in both borrowers’ ability to repay their debts, and in banks’ recovery rates. There is also a prospect of market risk, in that a sharp correction in valuations of assets such as equities and commodities may occur in the scenario of an abrupt transition to a low-carbon economy. Similarly, legal and reputational risk are the two main categories of operational risk that banks face due to climate-related uncertainty. It is necessary, therefore, to consider changing risk models in order to integrate climate-related issues. Moreover, it is also necessary to look at making climate risk a fully-fledged element of banks’ risk-management strategies.

Capturing physical risk in banking risk models is a methodological challenge. A natural disaster can cause a borrower to fail. However, credit risk models are ill-equipped to anticipate such strongly-correlated events. In the event of a localised natural disaster, such as a flood or an earthquake, the correlation between default events is primarily a geographical one. The correlation is more difficult to capture in the case of a non-localised natural disaster, or a localised disaster with broad effects, such as a pandemic or heat wave. For a comprehensive approach, one would need to identify the idiosyncratic vulnerabilities of each counterpart to climate risk. However, the effects of physical risk are complex to anticipate, each type of event giving rise to a specific scenario. In the case of physical risk, the coverage rate of a population or territory by flood, storm, or earthquake-type insurance is, therefore, a factor that has a direct impact on credit risk. If losses are insured, more frequent and severe weather events first affect insurance and re-insurance companies. Then, indirectly, they affect their customers through higher premiums. If losses are not insured, the burden falls on households, businesses and, ultimately, on government budgets.

In a joint report, the Bank for International Settlements and the Banque de France have compared climate risk to a “green swan” (Bolton et al., 2020[10]). Indeed, and as touched upon already, climate change has specific characteristics that make its impacts difficult to model. First of all, the use of historical data is of little use in measuring future risks, as physical risks will worsen with global warming, and transition risks remain low for the time being, as policy actions are still at an early stage at the global level. Second, the changes induced by climate change will be far-reaching, with non-linear, correlated, and potentially irreversible impacts. Third, the different time horizons at which the various effects of climate change will materialise are uncertain, whereas the likelihood of these changes occurring is high. Finally, the extent of long-term changes will depend on short-term policy actions.

As a result, a value-at-risk-type representation will therefore only imperfectly capture climate risk, which is typically positioned beyond the 99% confidence interval. Given its unprecedented nature, climate risk (and in particular transition risk) seems to be more compatible with a forward-looking approach based on scenarios, than with statistical models, which are necessarily based on historical data. Physical and transition risks represent two distinct types of risk, which must be handled differently, because they have different characteristics and channels of transmission to the banking sector.

The climate stress test undertaken by the central bank of the Netherlands, which was the first exercise of its kind, provides an initial methodological reference framework upon which to build. It is also more holistic compared to other stress tests, in that it tries to better assess the offsetting benefits of the transition. The climate stress test of the Dutch central bank was built around three shock scenarios. Under the technological shock scenario, unanticipated technological breakthroughs make it possible to double the share of renewable energy in the energy mix. Under the regulatory shock scenario, a set of policies aimed at reducing greenhouse gas emissions is implemented abruptly, resulting in a sharp increase in the price of carbon. The third scenario is that of a confidence shock, in which uncertainty about the government’s climate change policies causes a sudden drop in confidence among consumers, producers and investors (Vermeulen et al., 2018[11]).

Integration of risks related to climate change into macroprudential stress tests

An effective framework for monitoring systemic risks is an essential element of the macroprudential toolkit. Among the wide range of risk indicators, moreover, macroprudential stress tests play a key role. For years, national supervisors and the International Monetary Fund have conducted stress tests on the banking sector, or indeed the entire financial sector in a given country, to quantify the systemic impact of deteriorating macroeconomic conditions. Such stress tests also take into account the interactions between a deteriorating situation in the financial sector and the real economy. To account better for systemic climate risk, it is necessary to integrate the impact of climate change into these macro-stress tests. Some exercises have already been carried out. For example, researchers from the Netherlands central bank assessed the impact of a selection of transition scenarios on the country’s financial sector, concluding that the impact would be significant (Vermeulen et al., 2018[11]).

Climate stress tests are typically based on macroeconomic models that are capable of integrating climate change so as to determine its impact on macroeconomic variables. Given the complexity of the links between climate change, climate impacts, and socio-economic conditions, such an exercise requires sophisticated modelling. Calibration is also proving particularly difficult, given the lack of historical data on the impact of climate change. Overall, then, climate stress tests face major challenges. First, they need to take account of both physical and transition risks. So far, however, the focus has mostly been on the latter. Second, the definition of scenarios is another major challenge. Scenarios should be plausible, but also sufficiently severe, in order to be meaningful for all financial institutions within a jurisdiction, for the sake of comparability, or at the international level, to ensure a level-playing field. They should, moreover, focus on relevant risk factors, namely the channels through which climate risks will affect both the counterparties of financial institutions and the financial institutions themselves. Another difficulty relates to the gap between the short-term horizon of the typical stress test exercise and the medium to long-term horizon of climate risks. Longer-term horizons would also benefit from taking into consideration policies to mitigate such impacts, in order to better inform financial stability and fiscal policy initiatives.

Several major jurisdictions have already announced climate-related stress tests, and several others are considering undertaking them (Table 2.6). In Emerging Asia, methodologies for stress testing climate change-related risks are still at an early stage. The Monetary Authority of Singapore has already started to stress test for climate risks, but only in the insurance sector. In its 2018 industry-wide stress test exercise, it subjected insurers to a scenario involving extreme flooding, requiring them to consider the impact on their balance sheets of higher claims from damage to insured property. In addition, it is working towards incorporating a broader range of climate-related risks in thematic scenarios for its future industry-wide stress tests (MAS, 2020[19]). In April 2021, meanwhile, China’s central bank announced that it would work with other financial regulators to develop a methodology for climate-related stress tests, and would conduct such exercises in due course.

Given the complexity of climate stress tests, macroprudential policy makers could envisage alternative short-term solutions, such as scenario analysis. This is a less comprehensive exercise that could be used to assess the sensitivity to selected parameters of banks’ profit and loss accounts, and the items on their balance sheets. The level of these parameters is not derived from a macroeconomic model. The Task Force on Climate-related Financial Disclosures issued several recommendations in its 2017 report (TCFD, 2017[2]) on how to incorporate scenario analysis into strategic planning processes. The first recommendation is to further develop scenarios of transition that would keep global warming to two degrees Celsius or lower, that can be applied to specific industries and geographies, along with supporting outputs, tools, and user interfaces. The second recommendation concerns the development of broadly accepted methodologies, data sets, and tools, to allow organisations to undertake scenario-based evaluations of physical risks. Another recommendation is to make these data sets and tools publicly available in order to facilitate their use by organisations, reduce organisational transaction costs, minimise gaps between jurisdictions in terms of technical expertise, enhance the comparability of climate-related risk assessments by organisations, and help to ensure comparability for investors.

Existing macroprudential policy instruments could be amended to support green finance

When considering how to use existing macroprudential policy instruments to address the risks of climate change, the most relevant instruments could be those that target credit growth directly, or indeed those that target the sectoral allocation of credit. Examples of the former include capital buffers for risk-weighted assets. Examples of the latter include large-exposure rules that apply to potentially encumbered assets. Moreover, the Basel III counter-cyclical capital buffer could be particularly useful for promoting financial stability while transitioning from a high-carbon to a low-carbon economy, assuming there is an analogy between “financial bubbles” and “carbon bubbles”. Instruments targeting specific categories of loans, such as loan-to-value and debt service-to-income caps, could also be considered. However, their impact would, by definition, be limited to a specific sector. For example, it would be limited to the real estate sector in the case of loan-to-value and debt service-to-income caps. In the same vein, specific requirements targeting leverage ratios could be considered as a macroprudential policy response to risks related to climate change. However, the effectiveness of limiting bank indebtedness by regulating credit growth should be compared to that of implementing specific capital buffers.

Capital requirements could indeed be used to address climate-related risks. For example, a climate adjustment factor could be used to modify the risk weights of assets depending on how they affect the transition to a low-carbon economy. This could be achieved either by applying lower risk weights on green assets, or by increasing the risk weights applicable to so-called brown assets, whose carbon intensity is relatively high. One challenge to the integration of climate-related elements into capital requirements is the lack of sufficient evidence of a risk differential associated with green products. A more in-depth assessment of risk differentials associated with sustainability and climate risks will therefore be necessary before such capital adjustments become widespread practice (OECD, 2020[21]).

Another type of requirement that could be used is the leverage ratio. Introduced as part of the Basel III framework, this tool aims to limit each bank’s overall leverage. The leverage ratio could be supplemented with a sectoral leverage ratio requirement that imposes stricter rules for assets with high-carbon intensity. Further research would be needed to evaluate the effectiveness of such a policy tool as compared to minimum capital requirements. This research would also need to take into account the difficulties of implementation.

Liquidity requirements represent another type of requirement that could be amended in order to promote the transition to a low-carbon economy. Under Basel III, banks are subject to a short-term liquidity ratio, which requires banks to hold a certain level of short-term assets. They are likewise subject to a long-term structural liquidity ratio, which requires that long-term assets be financed by instruments with a maturity above one year. As has already been discussed with regard to capital requirements, liquidity requirements as they currently stand could hamper the financing of green activities, by making long-term financing more expensive. Regulators could consider differentiating these liquidity requirements to account for climate change, in order to give preferential treatment to green assets over brown assets. From a risk perspective, more in-depth research is necessary in order to understand whether green assets do indeed pose lower liquidity risks compared to other types of asset on banks’ balance sheets. From an economic policy perspective, meanwhile, applying differential treatment for green versus brown assets could facilitate the financing of green activities, and slow that of brown activities. Putting aside the question of the taxonomy that is needed to identify green and brown activities, evidence of the effectiveness of such a policy is still lacking. More research is therefore required in this area, before policy makers could implement any binding requirements.

Exposure limits and credit ceilings constitute the last group of tools that could help to promote and manage the transition to greater sustainability. Rules on large exposures typically set limits, usually a certain percentage of capital that individual loans cannot exceed. Concentration limits, meanwhile, usually set a given percentage of capital that the total amount of large loans cannot exceed. The aim of such limits is to force banks to diversify their loan portfolios in order to better withstand the bankruptcy of a large individual company, or a group of large companies. Concentration limits could be applied to overall levels of investment in carbon-intensive assets, which would be highly sensitive to a sharp transition to a low-carbon economy. As regards credit ceilings, limiting the expansion of bank lending to certain industries, and investments in certain specific asset classes, could also reduce financial flows to sectors or companies that exceed given carbon emission targets.

Specialised research has devoted attention to the need for macroprudential frameworks to evolve in order to improve the management of risks linked to climate change. Still, no consensus has emerged on the best way forward. Several studies have explored ways to amend central bank mandates so that they would include instruments that could support the transition to a low-carbon economy. For instance, Batten, Sowerbutts and Tanaka (2016[22]) explore the impact of climate change on the monetary and financial stability objectives of central banks, and identify several ways in which climate change, and policies on carbon emissions, could affect them. First, a weather-related natural disaster could trigger financial instability and a macroeconomic downturn if it were to cause severe damage to the balance sheets of economic agents. Second, climate change could have longer-term implications by affecting an economy’s potential growth rate. Third, a sudden and unexpected tightening of policies on carbon emissions could give rise to transition risks, in the sense that a disorderly re-pricing of carbon-intensive assets may generate a negative supply shock. Finally, increased reliance on bio-energy could increase the volatility of food and energy prices, thus making it more challenging for central banks to keep inflation close to their targets (Batten, Sowerbutts and Tanaka, 2016[22]). Batten, Sowerbutts and Tanaka (2016[22]) also argue transition risk could be mitigated through transparent, predictable and forward-looking policies on carbon emissions that encourage an early redirection of private investment towards low-carbon technologies.

Elsewhere, Dikau and Volz (2018[23]) list additional ways in which the macroprudential policy framework could be amended or enhanced to support climate-friendly investments. For instance, climate-related stress tests could help to calibrate green macroprudential policy instruments, and to facilitate the incorporation of the vulnerabilities that the tests identify into capital buffers, risk weights, and caps. In addition, capital requirements could be calibrated to differentiate asset classes based on sustainability criteria. Capital surcharges could be applied for institutions with large exposures to carbon-intensive assets, altering the identification of systemically-important financial institutions to take account of their systemic impact on the climate. In anticipation of sudden and negative price developments that may transpire in the future, higher risk weights could be assigned to carbon-intensive assets. Furthermore, loan-to-value and loan-to-income caps, or restrictions on large exposures, could be deployed in order to limit the amount of credit that banks extend to certain sectors, counterparties or geographical areas (Dikau and Volz, 2018[23]). Some of these proposals are in line with Schoenmaker and Van Tilburg (2016[24]), who identify capital instruments such as increases in risk weights, as well as restrictions on large exposures, as the most promising prudential instruments for carbon-intensive assets.

In a similar vein, D’Orazio and Popoyan (2019[25]) focus on the financial regulatory instruments that can be implemented within the existing regulatory framework. The authors argue that many existing policy interventions and proposals, such as the net stable funding ratio and the liquidity coverage ratio, run the risk of creating destabilising effects for the financial sector. Therefore, the authors suggest a set of alternative strategies to “greening” the existing Basel III requirements (Table 2.7). For example, the capital requirements toolkit could be expanded with a counter-cyclical, or negative, capital buffer over the course of the “carbon-intensive” credit cycle, favouring financial stability and mitigating excessive credit growth to brown projects. Second, leverage and liquidity requirements could be amended through the implementation of a sectoral leverage requirement. This could help to channel lending towards a specific green sector. Moreover, liquidity regulations could also discourage short-termism in financial intermediation. Minimum or maximum credit floors or ceilings, and large-exposure limits to constrain banks’ funding of brown sectors, are additional measures that could be considered.

While the implementation of green macroprudential policies could yield numerous benefits, certain factors may also constrain their effectiveness. As Bolton et al. (2020[10]) discuss the European Union’s experience with its SME Supporting Factor policy, which reduces capital requirements for lending to small and medium-sized enterprises. This EU policy uses similar levers to those that have been reviewed as potential amendments to macroprudential policy frameworks with regard to climate change. However, it has not been successful in triggering major changes in banks’ lending to SMEs. Furthermore, the implementation of a green macroprudential policy could potentially lead to trade-offs between the short- and long-term stability of the financial system. This is due to the mismatch between the long-term desirability of rapid and extreme measures to accelerate the transition to a low-carbon economy, and their potential to have a destabilising effect in the shorter term. As also argued by Bolton et al. (2020[10]), the main challenge in the short run is not the cost of green projects, but their number, which the authors still consider insufficient to drive real change.

Examples of policy initiatives to promote the transition to a low-carbon economy

In Emerging Asia, several countries have taken decisive action towards developing green policies (Table 2.8). For instance, India’s central bank imposes lending quotas in order to make sure that a minimum proportion of bank lending flows to environmentally-friendly sectors. It has issued guidelines on priority sectors for lending, in order to encourage and support environmentally-friendly lending by commercial banks, which in turn helps to achieve the United Nations Sustainable Development Goals. Eight priority sectors have been identified, one of which is renewable energy. The guidelines require domestic commercial banks, plus foreign banks with 20 branches or more, to make sure that loans to the eight priority sectors constitute at least 40% of their adjusted net bank credit, or credit equivalent of off-balance sheet exposures, whichever is higher (RBI, 2020[26]). Elsewhere, Malaysia’s central bank published an impact assessment framework for value-based intermediation financing and investment in 2020, with sectoral guidelines on palm oil, renewable energy, and energy efficiency. Also in 2020, it began a pilot implementation of its Climate Change and Principles-Based Taxonomy (Table 2.8). Moreover, China’s central bank and banking regulatory commission have developed a series of green credit guidelines, requiring banks to establish a monitoring and evaluation system for green credit (Park and Kim, 2020[27]). This progress notwithstanding, Emerging Asian policy makers have yet to implement specific green policies in the macroprudential area.

A reflection on the role of prudential regulations in promoting the transition to a sustainable economy is also already well underway in the European Union. The EU’s banking authority recommended in 2019 that legislators integrate sustainability considerations in directives and regulations that apply to the banking sector, in particular where they relate to governance and risk management. Additionally, it emphasised that the calibration of prudential requirements, which are primarily based on historical data, may not be a sufficient means to assess future changes, which may be without precedent. It pointed to the need for an enhanced, dynamic and forward-looking perspective in areas such as the calibration of prudential requirements, and approaches to modelling. As it seeks to incorporate environmental, social and corporate governance (ESG) factors in the prudential regulation applicable to banks, the aim is to ensure that there is no bias towards unsustainable financing (EBA, 2019[29]).

With specific regard to the transition to a low-carbon economy, central banks and other financial regulators around the world have implemented a range of policies. Campiglio et al. (2018[30]) provide a summary of these interventions (Table 2.9). For example, in 2011 Brazil’s central bank extended its requirements for the internal capital-adequacy assessment process that stems from the second pillar of the Basel II accords, by requiring commercial banks to take into account their exposure to environmental risks. More specifically, commercial banks are requested to incorporate these issues into their lending strategies, to carry out environmental stress tests, and to publish annual reports outlining their methods for assessing risk and exposure to social and environmental damage. These practices could serve as a benchmark for macroprudential policy makers in Emerging Asia.

Finally, the need to improve the comparability and quality of data with which to assess climate-related risks and opportunities is apparent. Non-financial reporting directives in a range of countries could serve as an example. The need to improve data comparability and quality ranks high on the G20 Sustainable Finance Working Group’s agenda, with the Financial Stability Board conducting work on this topic. The OECD has also raised the need for better quality reporting on climate-related factors, highlight that effective market pricing of climate transition is hampered by insufficient data, including financially material metrics and analytical tools to measure and manage climate transition risks, and lack of policy clarity regarding carbon pricing and support for renewables.

The development of a green taxonomy is a prerequisite for effective green macroprudential policy

Implementing green macroprudential regulation implies being able to distinguish with certainty between green and brown projects, and also those that are neutral with regard to climate change. The development of a stable, clear and standardised taxonomy in as many countries as possible is, therefore, of critical importance. Such a taxonomy will make it possible to apply common transparency rules on all financial products, resulting in an obligation for all companies to report the proportion of green activities that make up their financial portfolios. Such a taxonomy could be based on discriminating thresholds for factors such as carbon emissions, and could take into account the no-harm principle in order to define which projects count as green. However, ongoing debates on nuclear energy (JRC, 2021[31]) are testament to the complexity of implementing this taxonomy, which is nevertheless of primordial importance for identifying and supporting green investments. A binary classification of projects into low-carbon and high-carbon may, however, not be possible nor beneficial from the perspective of financial market participants. It may therefore be acceptable to assume that certain projects will lay in-between these two categories.

Additionally, stronger benchmarks and taxonomies to evaluate environmental, social and corporate governance (ESG) performance in the financial sector may also facilitate green financing and reduce reliance on ESG rating and research providers (OECD, 2020[21]). Various core approaches to ESG investing exist, including negative and positive screening, tilting investment portfolios aligned with ESG scores, as well as ESG impact and integration practices. Combined with different investment strategies, these approaches could alter asset selection in portfolios. Furthermore, the lack of standardised reporting practices and transparency, combined with the difficulty to translate qualitative information into quantitative facts, hinders the proper integration of climate-related factors into investment decisions (Boffo and Patalano, 2020[32]). It is also important to align industry-based (i.e. rating providers) environmental scores with composite ESG scores in order to meet investors’ environmental expectations. At the same time, environmental and ESG scores need to be carefully interpreted. High-ESG portfolios may not necessarily reflect a strong environmental performance or low-carbon activities. In turn, investors will need to conduct more thorough due diligence to get a better understanding of whether the scores of the rating provider properly incorporate and weigh such factors (Boffo, Marshall and Patalano, 2020[33]).

Some ASEAN member states have provided guidance on eligible project categories for green finance. Moreover, ASEAN’s framework of specific standards for green, social, or sustainability-related bonds are commonly used by member states to label a new bond or sukuk. These standards are based on the International Capital Market Association’s principles and guidelines (ASEAN, 2020[34]). In 2019, meanwhile, the Malaysian central bank issued a discussion paper on climate change and its impact on the financial system. It serves as a guide for financial institutions practices in identifying and classifying the economic activities that could contribute to climate change objectives (BNM, 2019[35]).

In 2015, the central bank of China published an endorsed project catalogue for green bonds, which is applicable to financial institutions and listed companies that want to issue them. Since then, several green standards and classification methods have been issued by various ministries and commissions in China, as well as by regional authorities, including, the China Securities Regulatory Commission (CSRC), the National Development and Reform Commission (NDRC), and the China Banking and Insurance Regulatory Commission. The taxonomies issued by the various bodies are to a large extent similar, although slight differences can be identified. In May 2020, the central bank, along with the NDRC and the CSRC, submitted a new edition of their project-support directory for green bonds, in a bid to harmonise the different taxonomies used by different government agencies (World Bank, 2020[36]).

At the regional level, central banks have worked together to develop the ASEAN Central Banks’ Agenda on Sustainable Banking. The agenda provides guidance to participants on ways to safeguard financial stability while at the same time supporting the transition to a low-carbon economy. Among other recommendations, central banks may collectively explore the development of a common, principles-based, ASEAN-wide taxonomy, along with specific green lending principles, in order to channel capital towards environmentally friendly activities. The common agenda on sustainable banking also promotes the establishment of a data collection framework. The goals of this framework are to ensure the proper monitoring of risk exposures, and to facilitate the assessment of the financial sector’s vulnerability to climate and environment-related risks (ASEAN, 2020[37]).

These initiatives notwithstanding, there is currently no common regional taxonomy or classification system for green finance in Emerging Asia. The European Union’s Taxonomy Regulation, which has been applicable in member states since June 2020, could provide a useful reference point for policy makers in Emerging Asia. It is a classification system that establishes a list of environmentally sustainable economic activities. Furthermore, it provides appropriate definitions to companies, investors, and policy makers on which economic activities can be considered environmentally sustainable. The overarching goal of the EU’s green classification system is to create security for investors, protect private investors from “greenwashing”, help companies plan the transition to a low-carbon economy, mitigate market fragmentation, and eventually help shift investments to where they are most needed. Box 2.3 contains more details about the guiding principles of the EU taxonomy for environmentally sustainable economic activities.