Despite teething problems, experimental climate stress tests are forcing a rethink of what constitutes risk in the banking sector, writes Alex Katsomitros

When Stuart Kirk, head of Responsible Investments at HSBC Global Asset Management, stepped onto the podium to deliver a speech at a Financial Times conference last May, few expected what would follow. Not one to mince his words, Kirk launched a full-frontal assault on conventional wisdom about green finance, accusing central banks of exaggerating potential credit losses from climate change in an attempt to “out-hyperbole the next guy”. Figures announced by the Bank of England in particular were misleading, Kirk argued, as the bank’s analysts used in the models “gigantic interest rates” that would lead to heavy losses in debt portfolios.

The speech sent shivers through an industry that is not just increasingly wary of the impacts of climate change, but also goes to great lengths to declare its commitment to green causes. Kirk was suspended by HSBC a few weeks after his speech, and eventually resigned to take a position at the Financial Times. In his recent columns, he fumes against the insidious capture of the finance industry by “woke” culture. “Stuart Kirk was deliberately provocative, flying a kite that was very quickly shot down,” says Michael Wilkins, Executive Director at Imperial College’s Centre for Climate Finance and Investment. However, many believe that these views are not fringe in the banking sector, even if few dare to share them in public. “Kirk’s views might be more common than we think among bankers – or at least the view that central banks’ concerns on climate risks are too strong,” says Francesc Rodriguez Tous, a banking expert teaching at Bayes Business School who has previously worked for three European central banks.

The first cut is the deepest

One reason why Kirk’s speech caused a kerfuffle is that the mood across the banking sector is shifting in the opposite direction. The world’s largest central banks have launched climate stress tests to assess how banks are preparing for climate change. A 2020 survey by the Bank for International Settlements, an organisation monitoring central bank policies, found that 24 out of surveyed 27 central banks and regulators had conducted research into climate-related risks or were planning to do so.

Some of these exercises have already borne fruit, revealing the huge challenges facing banks. Last spring, the Bank of England announced the results of its first climate stress tests, warning that UK banks and insurers could suffer credit losses of up to £225bn by 2050 if they failed to manage climate risks. The bank’s analysis tested participants’ balance sheets against three scenarios, from a moderate increase of 1.8°C compared to pre-industrial levels, to more extreme increases of up to 3.3°C. The results of a climate risk stress test conducted by the European Central Bank (ECB), announced last summer painted a similar picture, as most eurozonebased banks did not meet EU climate disclosure and management criteria, while just one out of five included climate risk in their criteria to grant loans. An earlier ECB “thematic review on climaterelated and environmental risks” had revealed that less than 10% of institutions were using “sufficiently forward-looking and granular climate and environmental risk information in their governance and risk management practices.” On the other side of the Atlantic, the Fed has announced that the country’s six largest banks will participate in a “pilot climate scenario analysis exercise” in early 2023.

Although still in an experimental stage, initial climate tests have received criticism from a wide range of sources. In his speech, Kirk accused central banks of designing these tests to produce alarming scenarios, likening the climate crisis to the “millennium bug”, an IT glitch at the turn of the century that caused far less damage than originally feared. “Even with a carbon tax, they couldn’t make climate risk move the needle,” he said. The timing of the tests has also been questioned, as banks are still recovering from the pandemic and are tasked with the implementation of a new batch of international banking rules, known as “Basel IV”, perceived as onerous in parts of the sector. Other critics have highlighted the use of static portfolios that don’t take into account future changes in balance sheets and other events that are nearly impossible to predict. “Climate stress tests are very stylised exercises that, while somehow informative, might not provide an accurate enough view of these risks to warrant the costs of running them – especially in the current environment,” says Tous of Bayes Business School.

Proponents of a tougher approach have argued that climate stress tests should have an international scope, given that natural disasters may spark migration and conflicts with global spillovers, causing abrupt changes in the value of bank assets. A recent report by the G20’s Financial Stability Board and the Network for Greening the Financial System (NGFS), a central bank group that helped shape initial climate tests, highlighted this concern: “Metrics are not capturing second-round effects, potential climate non-linearities, and the costs and potential further externalities from risk management measures taken by financial and non-financial firms”, the report said. For banks, one of the greatest challenges is to “expand modelling capabilities to take a holistic approach that factors in indirect impacts caused by climate-induced disruption to whole ecosystems,” says a UK banking source. Another worry is that different standards may create scope for regulatory arbitrage, with some banks gaining a competitive advantage by facing laxer regulation. “If we see a set of standards being imposed on banks in the OECD [countries], but not in developing countries, that is going to create an uneven playing field,” says Wilkins from Imperial College.

Time for change

Like many other businesses dealing with the impacts of climate change, banks face a conundrum: how to define something so broad in scope and collect the relevant data. The ECB reported that ten of the banks that took part in its test were constrained by a lack of “sufficiently granular or reliable” data for their risk models; some institutions had few or no data at all. “The main problem that banks are facing is computing the extent of ‘Scope 3’ emissions [emissions firms are not directly responsible for], broadly the emissions from their lending and securities portfolio,” Tous says, adding: “Different banks are approaching this problem differently, but at the end of the day regulators will have to decide whether the approach to estimating these emissions is valid or not.” Part of the problem is that banks rarely collect non-financial data from clients, which makes it difficult to assess their own climate risks. “One of the major challenges is the availability and granularity of climate and emissions data,” says Joris Hoff, a spokesperson for Rabobank, a Dutch bank. “We aim to make the most out of the available data at any point by combining externally available scenario data, like NGFS data, with internally available emission and exposure data on our clients.”

Another challenge is the integration of climate risk analysis into internal decision-making processes, along with training staff on the importance of climate considerations. Some banks have already started hiring analysts with a climate science and sustainability background. Robobank has set up special “Paris Alignment” and climate risk teams that include climate analysts and scientists, feeding data and other inputs into the bank’s credit risk calculations, physical risk scenarios, target-setting and financed emissions forecasting. According to Hoff, the information is ultimately used by the managing board for budget decision-making purposes.

Climate change is already here

Most banking experts recognise the limitations of initial tests, given the volatility of government policies and the climate itself. However, many expect these exercises to help calibrate regulation and promote best practices, possibly by increasing capital requirements and diverting investment from polluting to green industries. Future versions are also expected to be more targeted, focusing on specific banking activities like trading and lending. Perhaps their greatest contribution would be the establishment of widely accepted standards of what constitutes climate risk and how to manage it. “There still needs to be convergence and consensus amongst the industry about how to do it on a unified basis,” Wilkins says. “Currently, banks are doing it alone to some degree, without working together. But that is changing.”

What most experts acknowledge is that this is a problem banks face now, rather than in the distant future. “Banks, insurers and other corporations are facing billions of losses due to climate change risk. So we are not even talking about 20-30 years, we are talking about today,” says Wilkins, adding: “This is not like the millennium bug [in 2000] when nobody really knew what was going to happen. People can see the impact of climate change.”