Here’s the current reality of supervisory stress tests: the supervisor conducts them over a three-year horizon … and everything thereafter is out of scope.
Your investment decision-making could be set to cause a health pandemic in five years, a war perhaps in six, a nuclear disaster in seven, and a climate catastrophe in eight – but the stress-test won’t pick this up.
How one can interpret a financial stability mandate in this way is a mystery. But this reality of short-term supervision is now being challenged: by the supervisors and central banks themselves.
The central banks and supervisors in the Network for Greening the Financial System (NGFS) are starting to send a powerful signal that long-term risks – specifically those related to climate change – matter. And that they should be supervised. The report released by the NGFS yesterday sets out in six recommendations what a bold supervisory agenda can look like in a post-crisis world.
For the recommendations to work however, they have to address the broader question of how to supervise the long-term. This, we just don’t know. Google “Long-term financial supervision” and there are literally three hits and one of them is one of our reports (forgive the shameless plug). It’s not a concept. The broader malaise in financial supervision is that long-term risks are not supervised at all.
The trouble then is we don’t just have to ‘green’ supervision. We have to fundamentally reset it. For the NGFS recommendations to work, we need to redesign financial supervision fit for the 21st century. Four specific issues comes to mind:
First, we need to mobilize ‘suptech’ – supervisory technology — to protect the common good. The traditional mechanisms of data requests and surveys are not high-powered enough in the context of a fundamental technology revolution. Annual disclosures are quaint remnants of a world before total transparency.
Financial supervisors like EIOPA (the European Insurance and Occupational Pensions Authority) are now using line-by-line asset-level data to stress-test insurance companies’ exposure to climate risks. Others are exploring live-monitoring of housing prices, and physical risk analysis of mortgages in bundled asset-backed securities. Big data and real time monitoring of everything from algorithms driving financial transactions to weather events can ensure a swift response when long-term risks come knocking earlier than we expected.
Second, we need to (quantitatively) measure long-term risk management. A lot of the attention (including from organizations like ours) is on measuring the materiality of long-term risks.
We should spend equal amounts of time measuring the extent to which these risks are managed. Are financial institutions actually dedicating research budgets to long-term risk analysis? (Spoiler: Not so much).Natural language processing can help give insight into filings, communications, and other disclosures. Analysis we did in 2017 showed that only 4% of financial institutions in the MSCI World index mentioned fintech in their risk disclosures.
Supervisors can use these techniques to understand whether long-term investors are actually managing and acting on long-term risks in their investment decisions.
Third, we need to supervise long-term risks systematically. As the world around us becomes radically more uncertain and volatile, with (super) artificial intelligence, climate change, and pandemics on the horizon, treating sustainability risks in a silo risks missing the bigger picture.
Our research shows that non-traditional breakthrough technologies could by themselves cut oil demand to 2°C scenario levels in the next 22 years, without a single extra electric vehicle on the road. The relationships and interfaces between various long-term risks, and the broader threat they represent, requires a regulatory response beyond sustainability.
Finally, we need to build the ‘long-term supervision’ adapter: i.e. the plug that connects long-term risks to decision-relevant indicators and supervisory practices today.
This requires new tools such as ‘delayed stress-tests’ looking to capture the materiality of a climate Minsky moment. It also involves developing early-warning signal detection, measuring risk accumulation for example when markets deviate from 1.5°-2°C transition pathways. Other aspects include quantifying the time horizon of the value of financial contracts and understanding non-traditional correlations and network effects.
As we move from reports to action, embedding these issues in a vision for what long-term supervision could and should look like is critical. It ensures both the sustainability of the NGFS recommendations themselves, as well as the plug that drives their agenda across all supervisory practice.
Of course, the task is daunting. And some might say the radical uncertainty of this era makes long-term supervision nothing more than an exercise in make-believe.
The bad news for those voices however is that closing our eyes and yelling ‘you can’t see me!’ – as we did as kids – doesn’t make the problem go away. The good news is that supervisors are waking up. The NGFS message is clear: Watch out world, the long-term supervisors are here.
Jakob Thomä is Managing Director at the 2° Investing Initiative.