Bank climate commitments and green lending in emerging markets

The transition to a low-carbon economy will require trillions of dollars in annual investment over the coming decades. Recognising this challenge, Article 2.1 of the Paris Agreement calls for “making finance flows consistent with a pathway toward low greenhouse gas emissions and climate-resilient development“, with policymakers emphasising that private capital, particularly bank lending, must complement public funding. While 500 financial institutions (representing over $100 trillion in balance sheets) have already developed voluntary transition plans (GFANZ 2025), it remains unclear whether such commitments reflect meaningful changes in bank lending, let alone real decarbonisation. This potential gap between pledges and practice is especially concerning in emerging markets, where the potential for energy efficiency gains is largest and climate vulnerability most acute.

Existing evidence on voluntary climate commitments is mostly discouraging. European banks with extensive environmental rhetoric often maintain substantial exposure to carbon-intensive industries (Giannetti et al. 2023), and even when climate-committed banks reduce credit to high-carbon companies, borrower environmental performance is unlikely to improve (Kacperczyk and Peydró 2022). More recent studies reinforce this scepticism: banks joining the Net Zero Banking Alliance show no meaningful changes in lending patterns or borrower engagement (Sastry et al. 2024), while tightened Equator Principles have not shifted project finance away from brown projects (Berg et al. 2025). An exception comes from Green and Valleé (2024), who demonstrate that voluntary coal exit policies can reduce carbon emissions.

Against this somewhat disheartening backdrop, in a recent paper (Bernad et al. 2025) we provide new evidence from emerging markets. More specifically, we examine how banks’ climate commitments relate to their internal practices across 33 low- and middle-income countries in Emerging Europe, Central Asia, and North Africa. Our unique survey data allow us to directly observe internal green lending practices. Our approach addresses some key gaps in existing research, which focuses almost exclusively on high-income countries with strong environmental regulation. In our emerging market context, different institutional constraints, financing patterns, and climate vulnerabilities may yield different outcomes (De Haas 2025).

We construct our dataset by combining two main sources. First, we track banks’ participation in three climate initiatives: the Principles for Responsible Banking/Investment (PRB/PRI), the Science Based Targets initiative (SBTi), and the Task Force on Climate-related Financial Disclosures (TCFD). In our sample, 29% of banks had committed to at least one initiative – with 10% joining one, 13% two, and 6% all three. Second, the Third Banking Environment and Performance Survey (BEPS III), conducted across 33 economies during 2020-2021, provides unique data from structured interviews with 644 CEOs and credit heads at 335 banks about their environmental practices – information unobservable in studies relying on public disclosures. We use these responses to construct a Green Management Index (GMI) and Green Lending Index (GLI), analogous to indices developed by De Haas et al. (2025) for firms, while also incorporating geospatial data on bank branch locations.

Green banks in poor countries: Stylised facts

Figure 1 reveals that climate-committed banks score substantially higher on both indices than non-signatories. Foreign banks also outperform domestic ones on green management, as do large banks compared to smaller ones. Banks emphasising development and innovation show greener lending than those prioritising efficiency and profitability, while relationship-oriented banks outperform transaction-focused institutions – suggesting that long-term relationships facilitate environmental screening. These basic patterns already indicate that public climate pledges align with some genuine differences in banks’ internal managerial practices and may be more than just greenwashing.

Figure 1 Distribution of Green Management Index and Green Lending Index, by bank type

Notes: Figure 1 presents kernel density distributions to examine how banks’ GMI and GLI vary across five organisational dimensions: climate signatory status, ownership, size, corporate culture, and lending technology. Both indices are normalised to zero, with GMI ranging from -1 to 3 and GLI from -1.7 to 3.3.

Main findings

Our further analysis reveals three key findings. First, banks signing international climate initiatives score 0.83 standard deviations higher on green management and 0.47 standard deviations higher on green lending practices than non-signatories. These differences persist after controlling for ownership, size, culture, and lending technology. Climate-committed banks are more likely to employ dedicated environmental managers, maintain climate risk frameworks, and screen loans for environmental impact, suggesting genuine organisational differences rather than pure greenwashing.

Second, firms borrowing from climate-committed banks are 5.4% more likely to make green investments (over a three-year horizon). While we do not establish causality, this pattern – whether reflecting selection, influence, or both – demonstrates that banks’ environmental commitments tend to reflect some meaningful differences in client composition and behaviour.

Third, using geocoded locations of 46,843 potential firm-bank pairs within a 5km radius, we document local assortative matching: green firms preferentially borrow from nearby climate-committed banks, particularly when firms make actual green investments (for example, investments in renewable energy generation on site, waste management, air pollution control, among others) rather than just maintaining green management structures. These geographic patterns reinforce evidence against greenwashing: purely cosmetic commitments would not produce systematic spatial differences in lending relationships. Our local-level findings extend Degryse et al.’s (2023) documentation of green-meets-green matching in global syndicated loans, to SMEs in emerging markets.

Conclusions

Combining unique survey data on banks’ internal practices with firm-level information and geocoded locations, we document three findings suggesting banks’ voluntary climate commitments may represent more than just greenwashing. First, climate-committed banks score 0.83 and 0.47 standard deviations higher on green management and lending practices respectively, employing environmental managers, maintaining climate risk frameworks, and screening loans environmentally. Second, their borrowers are 5.4% more likely to make green investments. Third, geographic analysis reveals environmentally oriented firms preferentially match with nearby climate-committed banks. These results contrast with the predominantly null findings in prior work. We attribute this difference to two factors. First, emerging markets provide greater scope for differentiation: baseline environmental standards are lower. Second, our unique survey data capture actual managerial practices – environmental managers, risk frameworks, loan screening procedures – that remain unobserved in studies relying solely on public disclosures or lending outcomes.

Source: cepr.org