Guest post: Climate risk prudential regulation: "a-one, a-two, a-you know what to do"

Jérôme Courcier of the Scientific and Expertise Council of the French Sustainable Finance Observatory looks at whether the three-pillar Basel framework is up to the task of tackling climate risk

***Guest post*** I am grateful to Jérôme Courcier for sharing the below article for publication. Feedback? Get in touch with Jérôme at

As an Earth Day special, this article is free for all subscribers. Thursday comment articles are typically for paying subscribers only. If you’re currently only receiving Friday’s weekly round-ups, do consider becoming a paid subscriber for the full Climate Risk Review experience:

Banks run on risks. They invite them in, warehouse some, repackage others, and sell still more to an array of public and private customers. These risks come in all shapes and sizes: counterparty, credit, interest rate, liquidity, geopolitical — all are welcome at the modern lender.

However, too much risk can be a dangerous thing. To prevent banks from overindulging, the Basel Committee on Bank Supervision (BCBS), a global standard-setter, established a regulatory framework to which all major economies subscribe. It’s made up of three components, or ‘Pillars’: minimum capital adequacy requirements (Pillar 1), supervisory review (Pillar 2) and market discipline (Pillar 3).

Banks’ requirements under Pillar 1 are simple enough. All must maintain a minimum ratio of capital to risk-weighted assets (RWAs) of 8%. Why? Because it’s believed this ensures banks have enough loss-absorbing resources to survive if their risks turn sour. The riskier an asset, the higher its risk-weight under the Pillar 1 regime. For instance, a fully-collateralised home loan would attract a lower risk-weight (and therefore capital charge), than a complex securitisation of unsecured corporate debt.

Pillar 2 empowers bank regulators to scrutinise their supervisees’ internal risk-weight calculations, and pass judgement on the suitability of the resulting RWA amounts. In the European Union and UK, if regulators believe a bank’s own RWA assessments are deficient, they can impose capital add-ons at their discretion. 

Finally, Pillar 3 puts power in the hands of market participants, by forcing banks to publish periodic risk disclosures. These are supposed to give investors the relevant information they need to make informed trading decisions.

It’s an open question as to whether this three-pillar approach is up to the challenge of climate change. Last week, the BCBS proposed a “comprehensive analysis” of how climate-related financial risks could be incorporated in the present framework, and pledged to identify those gaps where they “may not be sufficiently addressed”. 

Right now, few financial watchdogs are willing to openly invoke the Basel system in their supervision of climate risks. A BCBS survey, also published last week, found that respondents — all financial supervisors — engaged with banks on climate in several ways outside the Basel framework, mentioning surveys, conferences, workshops and supervisory engagement.

One reason for this may be a lack of confidence in the framework itself. In the same survey, BCBS said a number of respondents saw a need to re-evaluate Pillars 1, 2 and 3 in order to fully take into account climate-related financial risks. 

Another could be a feeling that climate risks themselves are incapable of being captured, monitored and reported on like traditional risk categories. Put simply, that they resemble a whole new breed of risks that require a unique approach.

Are such concerns well-founded? Let’s start with Pillar 3. Climate risk reporting has garnered plenty of support among financial watchdogs. Indeed, several jurisdictions have already mandated, or are in the process of mandating, climate disclosure for financial institutions under their sway. Many are also encouraging banks to examine and disclose their climate-related exposures using scenario analysis, stress testing, and forward-looking metrics.

However, in a recent consultation run by the Task Force on Climate-related Financial Disclosures (TCFD) covering over 200 companies, three in four financial institution respondents said they were uneasy with the methods used to create forward-looking portfolio metrics, and 53% that they distrusted the reliability of their outputs. 

Source: TCFD

Most seem to share the Bank for International Settlements’ view that traditional approaches to risk management and measurement — based on extrapolating historical data and assuming normal return distributions — are not suitable for assessing future climate-related risks. After all, the downside effects of climate change, both transition and physical, may become non-linear when certain thresholds are crossed, and all but impossible to calculate. 

In other words, Pillar 3 may make investors aware of the existence of climate risks, but alone may not be robust enough to influence investors’ decision-making, or shift financial flows towards sustainable activities. This theory will be put to the test soon enough — the European Banking Authority (EBA) issued a proposal last month to integrate ESG factors into existing Pillar 3 rules across the EU. 

What about Pillar 2? This component seems to be the favourite of supervisors when it comes to climate risk management. In the abovementioned BCBS survey, those respondents who did think the three-pillar approach was capable of tackling climate risks said Pillar 2 was the best tool for the job.

Certainly the ECB is in favour. It’s come out in support of incorporating climate-related and environmental risks into its Internal Capital Adequacy Assessment Process (ICAAP) for banks, and for having climate risks capitalised using Pillar 2 charges through the Supervisory Review and Evaluation Process (SREP).

What’s unpalatable about this approach, though, is that it takes the form of a private negotiation between regulator and institution, based on the outputs of a bank’s own internal models — the inner workings of which are not disclosed to the wider world. Furthermore, when it comes to the size of Pillar 2 add-ons much depends on supervisors’ own judgements. Who’s to say that a regulator’s climate risk analyses should be any more (or less) sophisticated than those of banks themselves?

A more transparent approach to Pillar 2 add-ons could be built on top of the EBA’s effort to get banks to publish a ‘Green Asset Ratio’ (GAR). Under this proposal, banks would have to publicly report the activities they fund that are compatible with the EU’s sustainable taxonomy as a percentage of their total loans. Potentially this new KPI could be used by regulators to set a ‘Green Asset Floor’, which would require banks to hold a fixed percentage of their assets in ‘green’ loans. Banks that fall below this floor could have their Pillar 2 add-ons increased mechanically, in line with the perceived risk of holding too many ‘non-green’ assets.

And so to Pillar 1. Ultimately, incorporating climate risks into minimum capital requirements would be the most effective way to both penalise, or even prohibit, ‘brown’ financing while supporting ‘green’ activities. A panel of 50 sustainable finance experts said as much to the Climate Safe Lending Network earlier this month.

Some regulators think so, too. Anna Sweeney, the executive director of the insurance supervision division at the Prudential Regulation Authority (PRA), said in a speech at the Moody’s Insurance Summit Webinar in September 2020: “Risk mitigation in a prudential regulation authority is largely conducted through the lens of firms’ capital adequacy […] It is therefore possible that the incentives to address climate change risk for both firms and supervisors could be enhanced if it were incorporated explicitly into firms’ capital requirements”. 

Conveniently, this interpretation of the purpose of the prudential regulator’s job aligns neatly with European regulation as already written. Article 128 of the EU’s Capital Requirements Regulation states that institutions should apply a 150% risk weight to exposures associated with risks that are particularly high or difficult to assess — which could and should include climate risks. Existing rules also allow for risk-weights to climb to 1,250% for certain assets, meaning they have to be entirely equity-financed. As new fossil fuel exposures accelerate climate change, such assets could be weighted at this limit, as the non-profit Finance Watch recommends, to deter financing of coal, oil, and other climate-harming energy activities.

Sounds good, right? Still, there’s a potential dark side. As the Institute for Climate Economics argues, the sudden imposition of new Pillar 1 charges has the potential to trigger a rapid readjustment of financial and economic actors’ expectations, and a stampede out of carbon-intensive assets. This in turn could set off a chain reaction of loan defaults and bankruptcies, which would deplete banks’ loss-absorbing capital buffers and unleash market risks throughout the financial system. 

Such an outcome would be devastating not only for the banking system, but the cause of the low-carbon transition itself. The banking sector has to be able to finance real economy participants throughout this transformation after all. That means supporting the companies committed to changing their ways, even if today they’re among the most carbon-intensive in operation. This is why the EU taxonomy is not limited to activities that are already ‘green’, but also covers activities considered ‘in transition’ and those helping other sectors of the economy to go ‘green’. 

What we need, then, is a Pillar 1 approach that isn’t just tough on climate risks, but is smart on them too. Regulators should be able to distinguish between those assets that elevate systemic risk, and those that reduce it — even if both are carbon-intensive today. 

Similarly, they should not be afraid to lower capital requirements for explicitly ‘green’ assets that minimise or reduce a bank’s climate risks. Chenet et al argue that capital discounts for ‘green’ assets are a bad idea because the banking system isn’t holding enough capital to begin with. However, in the same way that one can reduce global road speed while maintaining a difference between city centers, country roads and highways, one can increase banks’ protection buffers while making a distinction between climate-friendly and climate-harming activities. 

Think about it: if a home is well-insulated and has a low-energy heating system, this should translate to lower running costs. If it’s designed to be resilient to natural catastrophes, its value will hold even as climate-related physical risks escalate. From a bank’s perspective, this makes it a ‘safer’ asset than one without these features. Applying lower capital requirements to such an asset is not letting a bank dispense with its usual prudence in risk management — it’s simply accurately reflecting the home’s higher creditworthiness.

This is not merely theoretical. In a recent working paper, economists at the Bank of England found that mortgages against energy-efficient homes are less frequently in payment arrears than their non-energy-efficient counterparts. A similar differential has been found in Italy, and credit ‘outperformance’ is observable in the green bond market, too.

Economists, activists, and regulators themselves are buzzing with ideas on how financial regulation could be updated to encompass climate risks. Indeed, the idea of lowering capital requirements for ‘green’ assets was proposed by the French Banking Federation way back in 2016. This means the Basel Committee need not break new ground with its “comprehensive analysis”. It should instead embrace the good ideas already being put into practice, and implement those for which validating evidence already exists — while being mindful of the limitations of the three pillars enumerated above.

Jérôme Courcier is a Member of the Scientific and Expertise Council of the French Sustainable Finance Observatory

Please send questions, feedback and more to

You can catch climate risk management updates daily on LinkedIn

The views and opinions expressed in this article are those of the author alone

Image under free media licence from Canva