Do Bank Financial Models Disincentivize Investment in Low-Carbon Projects?

If You Read Nothing Else in this Post: 

The financial sector is key in the global push to decarbonize various industry sectors in line with science-based carbon emission reduction targets. Decarbonization could occur when banks transition investments from high-carbon industries (like oil and gas production) to low-carbon investments (like renewable energy). A new study on lending practices for 59 European Union banks found that this transition to low-carbon assets could cause pain to banks’ bottom line, with researchers computing an average 15% 5-year profit loss across the banks they analyzed. The profit loss results from risk models used by banks that rely on historical data and classical financial metrics that do not account for potential future climate policy shifts (e.g., a price on carbon emissions). So unless fundamental changes in lending risk models incorporate these potential future changes, banks and bankers will be left to decide: invest in high-carbon assets and reap higher profits, or transition investments into low-carbon assets and take a cut on my bottom line. Which would you choose?

What’s the Role of Financial Institutions on Greenhouse Gas Emissions?

We’ve discussed sectoral contributions to global greenhouse gas (GHG) emissions. Here’s a reminder:

This post focuses on the key role that the financial sector plays in the movement to a lower-carbon economy in line with science-based targets established to avoid the worst effects of climate change. But you might be wondering—the financial sector isn’t even displayed in the sector-based GHG emissions shown in the figure above, so why is this sector worth exploring?

It turns out that the emissions funded by the finance sector are far greater than the actual, direct GHG emissions caused by all the world’s financial institutions. One recent estimate by the nonprofit Carbon Disclosure Project (CDP), which receives annual self-reported GHG emissions data from more than 20,000 companies and institutions, says the emissions created from projects and sectors financed by banks and other institutions are more than 700 times larger than their direct emissions. Further, the report suggested that banks and financial institutions are not adequately considering potential climate risks in their lending and investment activities. Finally, the report indicated that three-quarters of financial institutions responding to a survey saw major financial opportunities in making low-carbon investments. This story sounds simple - financial institutions’ investments are fueling many of the world’s GHG emissions. Still, these institutions see significant financial upside in making “greener” low-carbon investments, so all institutions should probably transition to low-carbon investments, right?

The story is a bit more complicated. We will highlight the results of a new study in Nature Climate Change (one of the most highly rated journals in the climate and environmental science world) that digs deeper into some practical realities of how banks specifically incorporate potential lending risk into their financial models when making decisions. Their findings provide an eye-opening perspective on how current bank lending constructs may create perverse incentives to continue investing in high-carbon investments.

A Focus on Incorporating Risks in the Investment Decision-Making Process

The following phrases should be familiar to anyone who has ever invested:

• All investments carry risk

• Past performance is no indicator of future results

Financial institutions use various quantitative and qualitative techniques to gauge the attractiveness and risk of a given investment. Likewise, they can use various techniques to de-risk a given investment, thereby increasing the odds that the investment will not only be paid back in full but also achieve the returns it is aiming for.

As an example, when I co-led circular economy investment funds at Closed Loop Partners, part of my initial scouting work involved evaluating many potential investments against a series of metrics we developed to indicate whether the shape of the investment aligned with our investment goals and strategies, which in part included assessing the risk of a given investment. For example, we’d determine a project’s execution risk by drawing on our expertise in the waste and recycling industry. A project at the safe end of the spectrum might be “lending $2 million to a municipality to fully fund some new recycling collection vehicles that will collect recyclables from newly-populated areas of the municipality”. A project at the other, far riskier end of the spectrum might be “lending $2 million toward an overall $50 million project for a new advanced recycling technology that’s been proven at a pilot scale but never demonstrated at a larger, commercial scale”. You might “price in” a risker project into your loan by giving yourself more favorable terms like a higher interest rate, a better position in the loan repayment stack, etc.

New Study on EU Banks and Lending for Low-Carbon and High-Carbon Investments

With some basics of lending covered, let’s dig more into the new study by Matteo Gasparini and the co-authors we referenced above. The authors explored model-based lending risk regulations that exist in the European Union. These regulations go something like this.

1. Many banks failing would be bad for the economy.

2. Financial regulations can foster stability in the banking system by requiring banks to implement certain requirements when they underwrite a loan.

3. One of these requirements is for banks to estimate potential losses from loans they make and then put aside additional capital based on the risk tied to those expected losses. This is a concept called a Loan Loss Reserve (LLR). In accounting, LLRs are considered liabilities.

In this study, the authors explored the question, “Do financial regulations requiring things like LLR systematically result in investments in high-carbon sectors being considered lower or higher risk than low-carbon sectors?” The authors used a novel approach to answer this question by first classifying a significant number of investments made by 59 European Union banks as “high-carbon” or “low-carbon,” then estimating how banks priced in risk using regulatory-defined risk models, then running sensitivity analyses to check that their answers were robust by considering other influential factors. Let’s take a look at their crucial takeaway through the below figure, which looks at a metric called the provision coverage ratio (PCR) for the 59 banks as a function of the total loan size for each bank, represented by four quartiles (the smallest loan size bucket is 0-25 (or first quartile) and the largest is 75-100 (fourth quartile)):

What is the PCR? Mathematically, it’s the ratio of the LLR for each loan divided by the total amount of the loan outstanding. A low PCR means the priced-in risk is lower compared to a higher PCR. Put another way, a lower PCR means the bank thinks it’s less likely a loan will default (i.e., not be repaid) compared to a higher PCR. We can see that regardless of the total amount of loans held by the banks analyzed, the PCR is greater for the low-carbon investments compared to the high-carbon investments. In other words, low-carbon investments are considered riskier. Notably, the authors found this result consistent regardless of the loan size (remember, the loan size is the denominator in the PCR calculation).

Why is this result significant? A given bank loan with a high LLR will be less profitable than one with a low LLR. Here is a simple illustration. Let’s say that a bank has a total revenue of $1 billion, and its total expenses (ignoring any loan loss provisions) are $700 million. Their net income would then be $300 million. However, if the bank increased its loan loss reserves - which, as this study points out, is currently the case for low-carbon investments based on baked-in accounting rules and risk models - by $50 million, that extra loan loss provision would increase its expenses by that amount, resulting in a net income of $250 million.

There’s no explicit regulatory definition of a “high carbon” or “low carbon” investment, so you might think that the authors’ results might reflect how they chose to define high- and low-carbon investments. It turns out that isn’t the case - they modified their initial classifications and found that the effect on the magnitude and direction of the results was minimal.

So then the question is: if current regulatory models used to estimate risk suggest low-carbon investments may result in lower profits than high-carbon investments, how much might profitability be negatively affected if banks fully transitioned to low-carbon investments? The authors answered this question by analyzing five years of historical profit data for a subset of EU banks where such data (along with LLR information) were available, and the results are shown here:

We can see that most individual banks analyzed would have had a decreased (and, in several cases, substantially decreased) profit had they been invested only in low-carbon assets, with the average cumulative profit loss of 15% across the whole data set.

Why Do Financial Regulatory Risk Models Seem to Consider Low-Carbon Investments as Riskier Than High-Carbon Investments?

Now, we need to dig into why low-carbon investments are coming out as far riskier than high-carbon ones. Put simply, banks' accounting rules and risk models rely on risk estimates based on historical data of classical financial metrics and often do not explicitly account for potential future trends. Current model-based regulatory risk computations for a given investment are based on a blend of a given entity’s profitability (e.g., what are their earnings before taxes divided by their total revenue), solvency (e.g., total debt divided by total assets), and liquidity (e.g., the amount of short-term debt carried divided by their working capital). Investigators computed risk ratios for 228 oil and gas firms and compared calculated risk ratios for 235 renewable energy firms based on 11 years of financial data, finding that several traditional financial risk measures were lower for the high-carbon oil and gas sector compared to the low-carbon renewable energy sector (e.g., a solvency measure related to interest expense showed 16% for oil and gas versus 32% for renewable energy, with the lower value equating to lower risk). The authors also explored more than a decade of data for the dozens of EU banks to compute risk another way by looking at the probability of loan default, finding the average probability of loan default for renewables was more than double that of oil and gas investments.

These results confirm that lending risk models that primarily account for historical data result in banks systematically classifying low-carbon investments as riskier than high-carbon investments, illustrating a tension between intentions to decarbonize and the pragmatic goal of achieving profitability (or more profitability). This begs the question: is there a fix for this situation?

It turns out, despite their long-standing use in the financial world of some of the classical financial ratios described earlier to express risk, these traditional measures are likely too simplified, and risk models may be better served by accounting for other factors that enable a more balanced approach to assessing an investment’s risk. For example, classical financial ratios do not contemplate how events in the future (e.g., a policy change related to GHG emissions) could affect the probability of potential losses for a loan. Keep in mind, this factor is crucial because a policy change (for example) requiring lower carbon emissions would be expected to make the high-carbon investment riskier and the low-carbon investment less risky. The authors illustrated this point by examining what happens to the risk profile of the EU banks analyzed if a carbon tax on GHG emissions was $100 per ton. The results were dramatic, with the solvency measure for high-carbon investments jumping from 16% to 46%, and the solvency measure for low-carbon investments remaining static at 32% (the renewables solvency is unchanged because there would be no direct penalty on carbon emissions because direct emissions are zero).

Insights and Implications

So, what can we learn from this study?

1. Emissions from activities financed by banks and other financial institutions are large, so decarbonizing investments from the financial sector is critical to meet global GHG emission reduction goals that will help us avoid the worst effects of climate change.

2. Traditional financial risk models used in EU banking tend to view low-carbon investments as riskier. Although we can’t generalize this study's results to the whole global financial sector, the results are robust given the large number of banks analyzed (59) and the large geographic coverage (14 of the largest countries in the EU).

3. Banks and bankers, therefore, likely have a perverse incentive (i.e., make greater profits) to invest in higher-carbon projects because of the net effect on profits when traditional financial models are used. This may be at odds with banks' intentions or public commitments to decarbonize.

4. If financial risk models contemplate potential or likely future events (e.g., a carbon tax), the risk equation can be flipped dramatically, with low-carbon investments are seen as less risky than high-carbon investments.

Do you know anyone who may enjoy this post? If so, use the widget below to share:

Thank you so much for reading Sustainability at the Frontier. We’ll see you next time.

If you view this in your email app, simply hit reply to provide us with your questions, comments, or feedback - we’d love to hear from you.