Tomorrow – October 1 – sees a number of new regulations take effect under the UK’s Pension Schemes Act of 2021. Among the most noteworthy of these is a requirement for UK pension plans to carry out climate scenario analysis. Although initially mainly applying only to plans with more than £5 billion of assets (roughly the largest 50 or so), this nonetheless represents a massive jump into the unfamiliar: as is the case for most institutional investors worldwide, previous use of scenario analysis (for this or any other purpose) by UK pension plans has been minimal.
In theory: an excellent risk tool. In practice: largely unused.
That scenario analysis has not been much used in the past by institutional investors is perhaps surprising, since it is an approach well-suited to those situations where the full range of possible future outcomes is difficult to capture, when the risk we are dealing with is not easily defined or modeled. These situations are sometimes referred to as “Knightian uncertainty”, sometimes as “radical uncertainty”. Keynes wrote of “uncertain knowledge”, and described this as things for which “there is no scientific basis on which to form any calculable probability whatever. We simply do not know.” That’s clearly the case for many aspects of investment, and certainly so for the future trajectory and impact of climate change.
Scenario analysis has a richer qualitative element alongside the quantitative, a story-like feel that is missing from so much investment work. And human beings love stories: we remember and learn better from stories than we do from numbers. The story-like aspect helps to turn an abstract risk into something more real. And scenario analysis can be an effective way to break out of narrow frames of thinking, making it a valuable risk management technique in a world where the biggest risk so often turns out to be the thing that you didn’t think of.
Despite this, scenario analysis has been, in practice, a largely untouched tool in the institutional risk toolbox. It’s been used only to a limited extent, and notable successes with it do not spring easily to mind.
Scenario analysis is different from what we’re used to
Perhaps this lack of use is due to scenario analysis simply being too different from what trustees are used to seeing. Trustees are accustomed to top-down modeling, modeling that summarizes entire distributions in measures such as expected return and VaR. But scenario analysis goes deep not wide, looking for lessons from particular outcomes rather than trying to capture the whole range of all possible futures.
Hence, scenario analysis is about the particular, not the general. It’s about the qualitative as well as the quantitative. And, in my opinion most importantly, it’s about the journey not the destination; not about the result, but rather about what caused the result.
If a trustee gets lost during a presentation of the detail of an actuarial valuation or a strategic asset allocation review, it may not matter much: the punchline comes at the end. But with scenario analysis, it’s the other way around: by the time you get to the end, the important takeaways have already passed. For scenario analysis to be useful, it must be hands-on. Indeed, the UK Department of Work & Pensions’ guidance for trustees issued in connection with the new regulations even goes as far as suggesting that: “Trustees should also consider whether they could gain more insight from undertaking a simpler form of scenario analysis in-house than out-sourcing it to a third-party.” It’s not about how brilliant the scenario you build is, but about what you learn from it.
How it works
To get a sense of how climate scenario analysis works, imagine two possible futures. In one, the Paris climate objectives are achieved as a result of global policy action, technological innovation and economic realignment. In the other, business as usual leads to a continued increase in global average temperatures and significant societal disruption.
Those two possibilities are not simply different outcomes on a temperature spectrum, but qualitatively and fundamentally different paths. The first implies a massive economic transition on a short timescale, a rapid realignment of how business is done. The second implies massive physical changes to the earth, with far-reaching consequences for every aspect of how we live. Already, we have two different stories, two narratives that each opens up a list of what-ifs with big implications for how we position portfolios.
A scenario analysis exercise puts flesh onto the bones of stories like those. What factors seem to matter most in each version of the future? What sort of policy and technological developments might be involved in the Paris-aligned outcome? What sort of geo-political considerations could come into play under business as usual? How do these play out over different time horizons? How do they affect investment outcomes? What might they mean for our liabilities? For our plan sponsor?
Done well, this will add insight into key risk drivers and potential tipping points, helping to identify factors that need to be monitored. It can enhance trustees’ understanding of what is plausible and what needs to be taken seriously, informing strategy and investment decisions and strengthening the risk management process.
A worthwhile experiment
Since this is a technique that has not been widely used by these plans in the past, the first round of scenario analysis exercises is going to be a learning process. It won’t be done perfectly and may even end up being little more than going through the motions in some cases. But climate risk is too big to ignore. The most widely-used existing risk tools and techniques just aren’t well suited to an area in which past experience tells us little, in which so many factors are at play, each connecting with each other, creating an expanding funnel of uncertainty the further out we look. I’ll be surprised if this process doesn’t help at least some trustee boards to break out of narrow frames of thinking, and produce at least a few examples of valuable new insight.