Why we need a transition integrated credit rating — and how to establish it
Current credit ratings are statistically sophisticated and thoroughly institutionalised. However, they do not capture climate transition risks. That is a major problem that sustainability ratings will not fix. This article discusses why this is a problem and how it can be solved.
Problems: mispricing, preventable defaults and poor allocation of capital
Standard ratings make good use of historical financial metrics, but hardly consider the underlying drivers of those financials, let alone how transition and (un)sustainability issues affect those underlying drivers.
The failure to capture transition risk in credit ratings and credit pricing models is problematic for several reasons. First, relying on these models means that credit risk is mispriced: for newly emerging technologies and business models, credit risk is overestimated and underpriced; and the reverse is true for technologies and business models that need to be phased out, such as coal fired power plants. Second, significant risks are overlooked and financial institutions will incur preventable surprises: defaults that could have been predicted, but are missed by conventional credit models. Third, at the macro level it results in a misallocation of capital, with still too much funding going into maintaining the status quo, a higher amount of stranded assets and hence higher losses.
Why adding a sustainability rating doesn’t work
The seemingly obvious and easy solution to fixing the above problems, is to buy an external sustainability rating and have it somehow affect the credit process, for example by demanding a minimum rating or by adjusting the pricing somewhat. That is not a solution, however. Sustainability ratings have many flaws, most and foremost their tendency to measure the formalisation of sustainability policies, rather than transition risk. How well or ill prepared is a company for transitions in say energy or mobility systems to net zero emissions? A sustainability rating won’t answer that question.
Moreover, such incremental use of sustainability information cannot properly be regarded as integration, though it is often labelled as such. Data is typically used as a traffic light, but does not affect the credit risk in the model. And even if it does, it does so in an artificial way, without clearly linking it to the underlying driver of PDs (probability of default) and LGDs (loss given default). Hence, it does not allow for better assessment of risk, nor for better valuation and better pricing.
The sustainable solution: a transition integrated credit rating
The sustainable solution lies in a transition integrated credit rating, which is a credit rating that truly captures transition risks. It does so by using raw data to estimate the probability that those risks will materialise throughout value chains in weaker financial results and correspondingly weaker PDs and LGDs. The value chain perspective is crucial since even the business model of a low carbon company can become obsolete due to transitions — conversely, a high carbon company might have a high degree of adaptability and be able to introduce new net zero products.
This is a much more forward looking approach than that of current credit ratings and ESG ratings. It better takes context into account, makes better use of available data sources, and allows for adaptation of the core of the credit model: credit ratings, PDs, LGDs and residual values. In addition, it can allow for the processing of valuable yet poorly quantified data — including analyst inputs on how companies differ from their industry peers. The drawback is of course, that this is more difficult trajectory than just buying some sustainability rating. It requires more determination, time, and investment in knowledge building.
What are the challenges?
Establishing a transition integrated credit rating is not trivial. The data challenge is large since the nature of the data is such that assumptions (and hence some degree of subjectivity) are required to link the risk factors to PDs, LGDs, and residual values. But the even bigger difficulty is the behavioural challenge: people are used to the current ratings and reluctant to even consider changing them (or the systems and processes associated with generating them). Current credit ratings are near sacrosanct, and heavily embedded in regulatory frameworks. It is not realistic to simply replace those all at once.
How can a transition integrated credit rating be realized?
Given the above challenges, establishing a transition integrated credit rating must be a journey in several steps. This journey starts with building the knowledge and awareness that such a rating is necessary in the first place. Unfortunately, the awareness of that need is still quite low.
Then, intermediate steps can be taken to create the rating itself. After careful selection of raw data, heatmaps can be constructed, which indicate where transition risks are clustered. These allow a drill-down into high-risk industries and individual loans to better understand and manage those risks.
The next steps include analyst questionnaires and scenario analysis. Analyst questionnaires can systematically gather information on individual companies’ exposures and adaptive capabilities. And scenario analysis can deepen the understanding and quantification of transition risks by calculating PDs and LGDs in specific climate change scenarios.
A combination of top-down (sector data) and bottom-up (per company) can make even the simpler starting modules usable from the start as the data gaps in the individual data can be filled with sector averages. This will often be necessary since individual companies often don’t have data on important metrics like scope 3 GHG emissions.
Jointly, such intermediate steps should provide the building blocks of a shadow credit rating, which can be run in parallel to the traditional rating system. When running this shadow credit rating system, banks learn to assess transition risks and start pricing them. In due course, and after probably several iterations, it can be shown to the regulator that the transition integrated credit system is superior to the traditional one, and it can replace the latter.
Conclusion
Transition risks can bring mispricing, preventable defaults, and capital allocation problems. Therefore, transition integrated credit ratings are badly needed, but not easy to establish. The various transition risks pose multiple data problems, but they can be overcome. The same applies to mindsets. Establishing transition integrated credit ratings requires a journey in several steps, using trial and error in an environment where that is possible, i.e. in an experimental environment where learning is possible. It is up to both bank management teams and regulators to set the bar high and create an environment where this is possible.
