European Central Bank’s “climate risk” stress tests need to be better
I’m a big fan of climate stress tests and scenario analysis for banks, as they’re an important means of assessing firms’ vulnerabilities to physical and transition risks.
However, when the results are confusing, misleading, and incomplete, they help no-one, least of all those who want to promote the cause of climate risk management.
That’s why I’m disappointed with the published results of the European Central Bank’s first supervisory climate stress test.
The results include just one aggregate euro loss figure covering 41 of the participants, and then for only a portion of their portfolios. Worse, this figure — €70 billion — is an aggregate of projected losses under a short-term, three-year disorderly transition risk scenario AND two physical risk scenarios.
In my view, aggregating losses from these three scenarios makes no sense, as this mix of severe transition risks and physical risks occurring at the same time stretches credulity. The calibration of the physical shocks is also dubious. For example, the ECB uses estimates from the Network for Greening the Financial System (NGFS) on labour productivity shocks due to heat stress until 2050, and appears to downscale the shock to a one-year horizon.
Naysayers like ex-HSBC executive Stuart Kirk are wrong about the threat of climate risks to the financial system, but supervisors don’t help themselves when they run climate stress tests and then give only a partial — and confused — view of the results to the public.
The ECB must do better next time around.