Climate risk-weighting: the devil and the deep blue sea

Should capital charges be calibrated to climate risk? European banks test the waters

Risk 1219 RWA Stephen Lee NB illustration
Stephen Lee,

  • European policy-makers are warming to the idea of adjusting asset risk weights to reflect their impact on the climate.
  • The idea is stoking controversy among commercial and central banks, which are divided on the wisdom of using capital requirements to drive sustainable investment.
  • Critics say mixing sustainability and credit ratings could incentivise banks to increase the riskiness of their loan portfolios in pursuit of ‘green’ laurels.
  • Questions surface as to how green assets eligible for capital discounts can be appropriately and uniformly identified. 
  • Some banks prefer carbon pricing as a mechanism to green loan portfolios.
  • US banks remain largely adrift from European discussions.

Climate change is real. So is climate-related financial risk. Yet prevailing prudential standards fail to capture this, meaning lenders needn’t capitalise against potential losses either from physical risks – such as extreme weather events – or transition risks, which erode the value of carbon-hungry assets as green technologies supplant them.

“We know there are risks linked to climate change. We know those risks haven’t been properly assessed in the past. We know they’re still not properly assessed,” says Karen Degouve, head of sustainable business development at Natixis in Paris. “If they were, it would necessarily result in a change in the allocation of capital.”

The French lender is among banks and regulators that are making the case for a green supporting factor (GSF) for allocating capital based on climate impact. The idea is gaining traction in Europe. In September, Natixis launched its own green weighting factor, offering the industry a test case on how a GSF could work in practice.

Put simply, a GSF increases the credit risk-weighted assets (RWAs) for carbon-intensive ‘brown’ loans – those vulnerable to physical and transition risks – while lowering those for sustainable ‘green’ projects, to incentivise financing of climate-friendly businesses (see box: Model in motion).

The GSF model has the support of the European Commission (EC) and also has commercial bank cheerleaders – Spain’s Santander among them. “Policy initiatives for the banking sector, such as a GSF, should be explored, since they could contribute to making financial resources available to put us on a path towards a more sustainable future,” says Santander head of sustainability, Federico Gomez. “It is important that banks can support the transition of other sectors, and are not penalised for doing so.”

But should credit risk-weightings be used to promote such policy goals? Not everyone thinks so.

“The danger with a GSF is you then frontload the balance sheet with deals on the basis of capital requirements that are supposed to reflect credit attributes – but actually don’t,” says Hervé Duteil, chief sustainability officer for the Americas at BNP Paribas in New York. “Why should a green project get X% less RWA against it? To be honest, it’s arbitrary. In particular, there is no evidence that its credit will be more resilient, especially as it could be displaced by a new technology.”

Consensus is still a long way off – and may possibly never come.

Measurement minefield

Many of those who question the GSF model’s merits recognise climate change as a risk factor. The difficulty lies in measuring and plugging it into existing risk management tools.

“From a pure credit risk perspective, it is easy to [intuit] that climate change has an impact. But an internal ratings-based [IRB] model is not – and cannot be – based on conviction. It has to be based on facts, on demonstrable numbers sustained by backtesting,” says Laurent Chédin, head of credit valuation adjustment and scarce resources at Crédit Agricole Corporate and Investment Bank in Paris.

The challenge is linking a borrower’s climate-related externalities to its creditworthiness. Climate risk may be real, but measuring it is hard – not least because data is patchy. “There is a growing consensus that there is a link between the genuine credit riskiness of a counterparty and its green footprint – the difficulty is translating this into something that can be quantified, that is robust enough to sustain regulatory review and validation,” says Chédin.

Credit risk models compute default probabilities based on decades’ worth of financial statistics. The same wealth of information does not exist for climate-related risks.

“While in principle it makes sense, provided that an institution can measure that risk, the reality is that it may be challenging to do that because relevant historical data is limited,” says Miroslav Petkov, a climate risk expert at Parker Fitzgerald’s global financial services practice in London.

Using projections to estimate climate risk is also loaded with obstacles. The 2015 Paris Agreement objective, to limit rising global temperatures to less than 2°C above pre-industrial levels, provides a marker for banks’ simulations. But the range of potential policies and innovations for reaching this goal complicate any asset-level assessment of climate risk exposure.

“You have to recognise that even green assets may face transition risks in the future. Think of wind energy. There may be more efficient renewable energy technologies in future that make some wind energy economically unviable. It could also be the case that subsidies are taken away from renewable energy sectors, leading to an increase in their credit risk,” says Petkov.

Alexandre Petrov, Nordea

However, these obstacles need not frustrate all efforts to develop a GSF. Some banks already have working models in place. “Currently, they are run in parallel to classic IRB credit risk models, used to determine capital requirements,” explains Alexandre Petrov, head of group credit risk corporate and IRB models at Nordea in Stockholm. “From a modelling perspective, they are quite similar to IRB models, having a number of risk factors and outcomes,” he adds, noting that he has seen such climate risk models, and has worked to reduce their dependence on expert judgement and place them on firm statistical foundations.

Meanwhile, the French prudential regulator, the ACPR, is known to be working on draft models to test bank portfolios’ susceptibility to climate transition risks, due for December delivery.

Assuming the climate risk of an asset can be estimated fairly, it’s still unclear to what extent this should alter its credit rating – and its capital charge.

Physical and transition risks will unfold decades in the future, but the weighted average duration of bank loan portfolios is much shorter. Current risk assessments use these shorter time horizons to determine creditworthiness.

“Typically, we are looking at a time horizon of three to five years. There might be sustainability considerations which are much longer-term but may not be material from a credit perspective within that time horizon,” says Carlos Wong-Fupuy, a senior director at credit rating agency AM Best in London.

Model in motion

Natixis’s mechanism adjusts the credit risk-weighted asset (RWA) value of each loan the bank makes in line with its climate impact, using a seven-level “colour rating”. The adjusted RWA is used to set the bank’s internal economic capital requirement for each asset. The colour scale runs from dark brown, for carbon-intensive assets, to dark green, for those with impeccable sustainability credentials. The adjustment factor ranges from +24% for the former to –50% for the latter.

By modifying RWAs this way, Natixis is explicitly tying climate and credit risks together. All else being equal, loans with like credit ratings, but different colour ratings, generate different economic capital requirements.

Natixis’s head of sustainable business development, Karen Degouve, believes this approach respects the data: “The very reason for implementing an adjustment factor is to reflect an accurate level of risk. Climate transition risks are here, and have not been sufficiently addressed by banks in the past – everyone recognises that today. You really should have less capital allocated to a wind farm than you should a gas power plant, [entirely] because of its climate transition profile.”

For climate risks to be properly incorporated in credit assessments, these time horizons need to shift. For its green weighting factor, Natixis analyses climate risk over the entire lifecycle of the asset, rather than just the term of the proposed financing. So, if a bank is considering a five-year loan to a power plant, it won’t be assessing the environmental impact of the loan’s maturity, but the 40- or 50-year life of the power plant itself, explains Degouve.

She believes such an approach accurately captures how client relationships unfold over time. A short-term financing with a valued customer is likely to be renewed by a lender on expiry, and corporate bankers typically ballpark their profit-and-loss estimates by referencing a core group of borrowers.

“Even if the maturity of a loan is one to three years it is likely to be renewed after it has expired. If we take a picture of our corporate loan book today and take it again in five years, my guess would be that 80% would be with the same group of clients for roughly the same amounts,” she adds.

Every asset subject to Natixis’s green weighting factor has to slot in somewhere on its colour scale. The differences in adjustment factor between categories can be large. But are such jumps justified?

Chris Cormack, co-founder of Quant Foundry, a consultancy developing climate change models for the financial sector, doesn’t think so.

“The academic literature shows that it’s statistically very marginal – the difference between brown and green – as it stands. When I reviewed what was out there, it was garbage,” he says. “It’s based on herd mentality on current price movements. What are the fundamentals? What the industry needs is a genuine forward-looking way of categorising a company’s credit (or investment risk) that can provide a reliable, transparent view of the likely capital structure for the company, depending on its investment choices.”

The academic literature shows that it’s statistically very marginal – the difference between brown and green – as it stands. When I reviewed what was out there, it was garbage

Chris Cormack, Quant Foundry

Once the research, data and models can be calibrated, it’s possible the differences in credit risk between brown and green assets alone will be too small to meaningfully reorientate capital flows towards sustainable projects.

“If we manage to integrate into our IRB models the proper level of credit riskiness from the climate transition, is that sufficient to save the planet? The answer is no,” says Crédit Agricole Corporate and Investment Bank’s Chédin. “Therefore, we would still need to act more on the commercial policy level at each bank, by making additional sustainability commitments and integrating those into our policies.”

Carbon pricing

While proponents of the GSF model include the European Union’s high-level expert group on sustainable finance, the EC and the European Banking Federation trade body, other mechanisms are jostling for recognition.

BNP Paribas and ING both favour carbon emission metrics to promote sustainable lending. According to Duteil, BNPP committed two years ago to introduce shadow carbon pricing.

“We are still currently in a pilot phase, where we recompute Ebitda [earnings before interest, tax, depreciation and amortisation] of companies in eight sensitive sectors – like oil and gas or transportation – to include the effects of a hypothetical tax of $25 to $40 on each metric tonne of carbon emissions,” he says. “This way, we can test their resilience to a direct or indirect carbon tax in the future and analyse whether they have built a resilient business model. This is gradually being incorporated as part of our credit process.”


ING is reshaping its global loan portfolio in line with the goals of the Paris Agreement. The Dutch lender has developed a way – dubbed the Terra approach – of measuring the greenhouse gas output of the clients to which it is exposed. Using numerous methodologies to assess technology usage, it populates client-specific ‘climate alignment dashboards’, which allow for comparisons of borrowers’ climate-related performance across peer groups. It uses these to steer clients towards climate-friendly activities.

According to its first Terra approach update, ING’s commercial real-estate and power-generation portfolios are already on track for alignment with the less-than-2°C goal.

Leonie Schreve, global head of sustainable finance at ING in Amsterdam, says the bank also supports a carbon tax to “drive the market”, claiming it would “have more impact” than a GSF: “If you have the carbon impact and the carbon pricing, you will have a measure of the actual impact of a client. That’s a balanced approach, instead of a general discount for everything that is green and add-on for everything that is brown.”

Brown sticks and green carrots

Assigning capital charges to specific exposures is a tried and tested means of swaying bank decision-making. But as a blunt instrument, it has been more effective at deterring banks from allocating capital to politically undesirable assets, rather than stimulating investment in desirable ones. Post-crisis rules have set punitive capital requirements for opaque securitisations and correlation trading portfolios, which have caused markets to dry up for these trades.

Would a GSF that lowers RWAs – and therefore capital requirements – calculated for climate-friendly assets be any more successful?

The EU hopes so. In a 2018 report, an EC-sponsored expert group said that a capital discount for sustainable investments “could give a strong policy signal to re-engage the banking sector” after “years of tightening capital regulation”.

Too-low risk weights increase the probability of financial crises, which will push work against climate change to the future

Otso Manninen, Bank of Finland

Yet Natixis is an outlier in making the case that green projects are a safer bet than brown ones. Other lenders are sceptical – including local competitor BNPP.

Duteil represents a school of thought that a GSF is a policy tool, rather than a risk management mechanism. He is not alone.

Professor Kern Alexander, of the University of Zurich, and Paul Fisher, formerly of the Bank of England, published a paper in January on banking regulation and sustainability. They voiced concerns that if the prudential regulatory regime became “a more general mechanism for politicians seeking to deliver wider social objectives, however desirable, it could potentially exacerbate sustainability risks”.

Why? If GSFs created a bubble in green loans that later go bad and bring on a banking crisis, political focus would shift to short-term risk mitigation, rather than the long-term objectives of limiting climate change.

“Risk weights should only reflect financial risks, not our endeavour to promote all that is good,” says Otso Manninen, a senior economist at the Bank of Finland in Helsinki. “Too-low risk weights increase the probability of financial crises, which will push work against climate change to the future. Therefore, trying to incentivise good causes with lower risk weights could turn out to be counterproductive.”

Green finance

The divide is between: those who believe capital requirements should be brought into play to promote green financing; and purists who argue they should be exclusively about financial stability – to ensure banks have sufficient buffers to absorb losses.

European policy-makers are split between opposing camps.

On November 11, the EU’s chief overseer of financial stability, Valdis Dombrovskis, said “prudential rules could favour ‘green’ investments and loans”, citing efforts by the European Banking Authority (EBA) to assess the possibility of a GSF.

A clutch of EU central banks has voiced support – but not everyone is singing from the same hymn sheet.

Jens Weidmann, president of German central bank Deutsche Bundesbank, has said it would be “wrong to want to use banking regulations to set climate policy incentives, for example, by granting risk-weighting ‘discounts’ for ‘green’ assets”, insisting it is the job of politicians to create such incentives.

The European Central Bank is also leery of the idea. In April, Sabine Lautenschläger, then member of the governing council, said central banks “should not be obliged ... to promote green finance by granting banks preferential capital if this is not justified by the specific risks linked to green finance”.

But while there is disagreement in Europe, there is dialogue. In the US, the topic is not even on the table (see box: US and them?).

The colour of risk

EU policy-makers’ efforts, meanwhile, are also focused on developing a taxonomy that will give financial institutions a common language with which to define and describe climate-friendly investments. This could serve as a first step towards sorting which assets could be eligible for a GSF in future.

The taxonomy will be absolutely key, and we will be referring to it

Piers Haben, EBA

“The taxonomy will be absolutely key, and we will be referring to it,” says Piers Haben, director of banking markets, innovation and consumers at the EBA. He is leading its efforts to integrate the EU’s sustainability agenda.

“We think appropriate classification and disclosure will bring market discipline to bear on encouraging banks to make more green investments. This is the key first step. And then we’re not going to guess and say we like green assets so they should be of a lower prudential risk – we will need evidence,” he adds.

The hope is the taxonomy will end disharmony between existing market-based classification systems and reduce the risk of “greenwashing” – making misleading claims for a product’s environmental benefits.

Once a standard is adopted, it could be used to calibrate risk weights in a uniform, transparent, and evidence-based manner. A technical expert group published a report on an EU taxonomy in June, and on December 5, the European Parliament and EU officials reached an agreement on a package that would bring the taxonomy into force. It now awaits approval by the EC.

Leonie Schreve, ING
Leonie Schreve, ING

“The main challenge will be in how wide or small the definitions are set,” says ING’s Schreve. “If it’s too small, [investment] will be focused and directed only at a niche market. If it’s too broad, then we will not move the needle on funding sustainable business. Then we have other challenges, in terms of implementation. One is data availability. At the moment, there is not enough data out there to determine if and what portion of our client activities are meeting the taxonomy definition.”

This is crucial, as banks need to identify brown assets and clients, but also green assets originated by brown clients, and vice versa.

Quant Foundry’s Cormack explains: “There are firms that may be brown now, but have a huge amount to invest in green. It doesn’t make sense to penalise companies that have the intention to invest, as long as they stick to it.

“We did a study of European utilities as to whether they could actually produce enough renewable energy in the next 30 years. There is just not enough new capital to achieve this, unless companies that are sitting on brown cash switch their investments.”

Existing renewable energy companies are too small to absorb a sudden influx of financing.

“Imagine that amount of investment borne by those companies currently operating in the green, renewables sector. For these companies, funding the transition purely through debt, without also receiving significant injections of equity, just leverages up their balance sheets and makes them very risky,” says James Belmont, a director at risk consultancy Baringa Partners in London.

Banks face pressure to divest from carbon-spouting industries, and some have either stopped or are drawing down their support to the dirtiest producers, such as thermal coal mines. But stopping funding for all non-renewables-only energy companies could actually inhibit the transition to a zero-carbon economy.

Natixis understands this. Its GWF won’t be used to identify ‘brown’ lenders with the objective of cutting them off. “We want to give our clients – even the big emitters – some time to progress, to go along the route of transition,” says Degouve.

US and them?

The green supporting factor debate may be heating up in Europe, but in the US, it’s virtually non-existent, despite polarised views on climate change. Of the eight large US lenders asked to comment for this article, just one offered an interviewee – but who declined to speak on the record.

The political backdrop may explain US financial institutions’ absence from climate-related initiatives. President Trump has repeatedly called climate change a “hoax”, and withdrew the US from the Paris Agreement, citing its objectives as a conflict with his ‘America First’ policy. Individual states have since signed the agreement.

In November, US Federal Reserve chairman Jerome Powell told lawmakers: “We are not going to be the ones to decide society’s response to [climate change]. That’s going to be elected legislators, not us.”

The Fed is also conspicuously absent from the Network for Greening the Financial System, a group of 51 central banks and supervisors committed to financial sector development of climate risk management.

Among global systemically important banks, 10 European and just one US firm – Citi – signed up to the United Nations’ Principles for Responsible Banking this September.

“[The US banks and regulators] are likely to be followers. If there are ideas out there in Europe on the regulatory side and on the toolbox side that can be wholesale adopted, the US may do so at some point. I don’t see the innovation in regulation coming from the US regulators,” says Stacy Swann, chief executive officer of Climate Finance Advisors, a consultancy in Washington, DC.

Gregg Gelzinis
Gregg Gelzinis, Center for American Progress

A few US lawmakers are trying to prod the Fed towards a more proactive stance. Senator Brian Schatz (Democrat, Hawaii) is working on a bill to compel large financial institutions to undergo special stress tests to test their resilience to climate change. Democrat-aligned groups are also clamouring for action.

“Regulators should consider climate stress-testing, explicit risk management requirements for climate risk, and incorporating climate risk into risk-weighted capital requirements. This isn’t about regulators stepping outside of their mandates and dictating the allocation of capital towards social ends. Bolstering the resiliency of the financial system in the face of future climate shocks falls squarely within their jurisdiction,” says Gregg Gelzinis, at the Center for American Progress, a liberal think-tank.

We’re going to need a bigger boat.


Editing by Louise Marshall 

Only users who have a paid subscription or are part of a corporate subscription are able to print or copy content.

To access these options, along with all other subscription benefits, please contact or view our subscription options here:

You are currently unable to copy this content. Please contact to find out more.

You need to sign in to use this feature. If you don’t have a account, please register for a trial.

Sign in
You are currently on corporate access.

To use this feature you will need an individual account. If you have one already please sign in.

Sign in.

Alternatively you can request an individual account here