A couple of people have asked me to opine on the scenario exercise published by the ECB last month. The exercise was trying to address a key shortcoming of earlier stress test attempts by focusing on the effects of climate transition in the short to medium term. It makes little sense to imagine bank portfolios remaining static over a thirty-plus year horizon. The ECB researchers considered scenarios primarily concerned with the timing of transition efforts. They ranged from an accelerated transition to a late push and a delayed transition, which would see the region fall short of the Paris targets by 2030. The paper then went on to assess the energy intensity of various sectors and it attempted to translate this information into the impact on PD for corporate and household debt. The ECB is clearly of the opinion that an accelerated transition would be beneficial, and this is precisely what the researchers found. The paper projected some bold increases in credit risk. Under the late push scenario, corporate risk was found to increase by 80% by 2030 relative to today’s levels. The scenarios involving more measured transition efforts, in contrast, recorded increases of around 40%. Even these 40% increases are quite stunning - the authors are saying that some form of transition effort is inevitable and, even under the best case scenario, banks stand to suffer a substantial hit to the bottom line. Given that this is projected over the next handful of years, you would expect results to force banks to sit up and take notice. This is, after all, within the extended budget planning horizon for many institutions. And yet the ECB’s scenario-based study has had very little impact. It’s been several weeks since publication and few people seem to have been discussing the results. It’s very difficult to determine the veracity of the ECB’s numbers. We’ve never seen a systematic level shift in credit performance, though we have observed a long sequence of cyclical fluctuations. One wonders whether investors would tolerate trend increases in losses or whether they would demand managers to rein in their lending practices. The results read more like the ECB’s opinion than anything backed by hard evidence. Had results like this been generated using a data-driven method - perhaps a case study showing that Martian banks suffered significant losses during a similar transition phase - the study would be taken far more seriously, Put a coherent confidence interval around the 80% prediction and, as long as it's not too wide, bankers will know what they stand to lose. When faced with hard evidence, they would need to be able to explain to investors why they were not taking transition risk more seriously. As it is, the projections can be dismissed quite easily. Is it just me or does the shelf-life of scenario-based studies keep getting shorter and shorter? Picture: Headquarters of the Bank of Mars, generated by AI.