Market-based metrics for systemic risk in climate change

The transition to a low-carbon economy has potential implications for financial stability. Unexpected changes in climate policy and regulation, technology, and consumer and investor preferences can cause sudden readjustments in market prices. Javier Ojea Ferreiro, Juan C Reboredo, and Andrea Ugolini have developed an empirical framework to assess different sets of scenarios. That’s what they write An early warning indicator is essential to understand the extent to which financial institutions are exposed to different transition paths.


The transition to a low-carbon economy brings challenges and poses risks to the value of financial assets, with potential implications for financial stability. Changes in climate policy and regulation, technology, and consumer and investor preferences, if unforeseen, can lead to sudden readjustments in market prices, which could have major implications for financial stability. Different transition scenarios could trigger different types of revaluation effects, so identifying and quantifying the impact of these impacts on financial stability is of great interest to regulators, investors and policymakers alike. The development of an early warning indicator is an essential tool that could provide useful information to identify the risk of financial institutions in realizing different transition paths that would not be apparent outside the framework provided by a scenario analysis.

We develop an empirical setup to quantify the impact of climate change risk on financial firms. In particular, we model the distribution of financial firm returns under the condition that different transition scenarios occur: disorderly transition, orderly transition, and greenhouse world.

Borrowing from the Network for Greening the Financial System (2020) narrative, a disorderly transition scenario is characterized by abrupt policy curbs on the use of high-carbon energy, which can lead to operational difficulties for companies exposed to higher risk, ultimately affecting affects the value of their assets (e.g. assets may be lost). In contrast, companies with a lower transition risk face a privileged market position (unless highly exposed companies adapt their production processes to the new circumstances in the meantime). As a result, market expectations for the asset prices of low-commitment companies (green companies) rise, while the opposite happens for the value of high-commitment companies (brown companies). This impact can be described using the quantiles for fixed returns, assuming that the green portfolio returns are above their highest value and the brown portfolio is below their lowest value. In a greenhouse scenario, policies to support the transition will be implemented slowly and with a lag, and investors will adjust their expectations accordingly. As brown companies have more time to offload stranded assets without suffering a large price impact, brown asset prices rise while green asset prices fall as green companies lose the opportunity to boost their businesses. Thus, the relative price impact of a greenhouse scenario can be described as brown portfolio returns above the peak and green portfolio returns below the trough. Finally, in the orderly transition scenario, the policy constraints to achieve the climate targets are implemented smoothly, allowing the companies to gradually adapt to the new business environment. Investors would therefore expect asset returns to hover around their median levels (ie without abrupt price changes).

We identify three market portfolios: green, brown and neutral, which are labeled for their exposure to carbon transition risks based on their Sustainalytics-calculated Carbon Risk Score (CRS) rating information. We refer to brown/neutral/green portfolios to indicate high/medium/low climate change risk portfolios to avoid cumbersome notation. We construct the brown (green) portfolio annually by using the highest (lowest) quintile from the cross-sectional distribution of CRS from a dataset of more than 900 European listed companies. The neutral portfolio is built up from the remaining assets. Based on the financial firm’s return dependency on these portfolio returns, we can quantify the impact of each transition scenario on the financial firm’s value and uncover vulnerabilities from climate-related risk events.

Following the systemic risk literature (Acharya et al., 2017; Browless and Engle, 2017; Adrian and Brunnermeier, 2016), we measure systemic impacts using the Climate Transition Expected Return (CTER), the Climate Transition Value-at- Risk (CTVaR) and the metrics of the Climate Transition Expected Shortfall (CTES). These metrics are calculated from the conditional distribution of financial company returns. Chart 1 summarizes the key metrics constructed according to our approach, which are technically similar to conditional risk metrics (Adrian and Brunnermeier, 2016; Girardi and Ergun, 2013), although in our setup the trigger is provided by the combined action of green and neutral , and brown portfolio returns. (For more details on dependency structures, data, formulas, and mathematical proofs, see Ojea-Ferreiro, Reboredo, and Ugolini, 2022.)

Figure 1. Structure of risk measures for climate change

Notes: This figure summarizes the goal of our framework, by setting stock market scenarios coherently with the NFGS narrative, we obtain the conditional stock distribution of financial firms (orange bars) which would differ from the unconditional distribution (blue bars) if financial firms and Stock market are not independent from which the climate change risk measures are calculated, ie the mean (CTER), a quantile of the distribution (CTVaR) or the mean below a certain quantile (CTES).

We present some results on sector-level CTvaR and country-level CTER for European financial institutions. Chart 2 shows the median CTVaR by subsector in the solid line, while the area indicates the cross-sectional interquartile range. (We have presented the results this way because VaR is a non-additive measure, so the charts are more informative when the interquartile range is taken into account.) The behavior of the European banking sector differs from that of the non-banking sectors, with the highest losses in the greenhouse -world scenario, while the remaining sectors, in contrast, suffer extreme losses in a greenhouse world scenario. Also, the cross-sectional behavior of financial firms appears to be more heterogeneous when in a greenhouse world scenario, as shown by the broader bands.

Figure 2. CTvaR at sector level

Banks:

Insurance companies:

Financial Services:

Property:

Sources: Ojea-Ferreiro, Reboredo and Ugolini (2022). Notes: These charts show the median CTVaR within each subsector (solid line) along with the interquartile range of the cross-sectional distribution (area). The red color refers to the greenhouse world scenario, green to the disordered transition and blue to the orderly transition. Each scenario has a distress of 0.2, e.g. For example, in the disorderly transition scenario, the green portfolio’s returns are above the 80th percentile, while the brown portfolio’s returns are below the 20th percentile. The CTVaR is calculated by looking at the 10th percentile of the conditional distribution.

Chart 3 shows the mean of the CTER over the sample period at the country level, with values ​​aggregated by relative market capitalization. We can observe that Southern Europe, Ireland and Poland are the countries most affected by the disorderly transition, while France, the UK and the Scandinavian countries are the most affected by a greenhouse world scenario. The losses (gains) in the greenhouse scenario are 100 (500) basis points higher than the losses in the disorderly transition scenario, as shown by the figures’ heat bar. This is the result of dependency asymmetries in relationships with market assets.

Figure 3. CTER at country level

Disorderly Transition:

greenhouse world:

Sources: Ojea-Ferreiro, Reboredo and Ugolini (2022). Notes: These charts show the weighted average CTER using relative market capitalization as the weighting factor versus the sample. Warmer color means higher losses. Each scenario has a distress of 0.2, e.g. For example, in the disorderly transition (greenhouse word) scenario, green portfolio returns are above the 80th (below 20th) percentile, while brown portfolio returns are below the 20th (above 80th) percentile.

By redefining the SRISK metric from Browless and Engle (2017) to our climate change metrics, we can assess the impact of transition risk on capital. As shown in Ojea-Ferreiro, Reboredo and Ugolini (2022), the largest capital losses in the banking sector occur during the disorderly transition, where capital requirements could reach 140 billion euros, which seems manageable by the European banking system.

Overall, our new empirical setup is easily reproducible and flexible to assess different scenarios for the transition to a low-carbon economy.

Authors’ disclaimer: The views expressed here are our own and do not necessarily reflect those of the European Commission.

♣♣♣

Remarks:

Leave a Comment