One of my reflections following the 2025 Global Tipping Points Conference was how little the term ‘systemic risk’ was used whereas, unsurprisingly, ‘tipping points’ was probably the number one phrase. Is the answer in the conference title (probably), or do Earth-system scientists use ‘systemic risk’ in other contexts (don’t know), or are they essentially the same thing and therefore interchangeable (worth exploring)?
The first step in any exploration is to define our terms. Here I hit my first problem. My first browser offered me definitions for systemic risk which all involved the financial system. I had to scroll down to find a bit of economics creeping in. I tried a second browser, and economics joined finance a bit higher up the list, but no other system featured. So, am I trapped in a particular filter bubble? Or are we collectively trapped in a distortion of reality? Surely a change in the behaviour of an Earth system (e.g., climate) would qualify as systemic risk.
To cut a long story short[1], I will define systemic risk as the chance a system behaves in a way that we don’t want it to. Note that this is different from the system delivering us an outcome that we don’t like – we can label those as tail events from a normally-behaving system. Systemic risk describes an unexpected change in system behaviour.
The typical definition of a tipping point is that it is a threshold, or transition zone, beyond which is a new ‘equilibrium’ state, and the crossing of which is effectively irreversible. This definition introduces some terms that we should also clarify. I placed equilibrium in quotes because (1) it is not strictly necessary (the irreversibility implies the new state will persist for a while), (2) it may not be a pleasant state, and (3) it may not be a permanent state. The latter point introduces the idea of timeframe. Systemic risk implies that the threat is within a timeframe that matters to us – largely because I have chosen to define it around “us” and what we want. Tipping points, on the other hand, could be referring to geological timescales. And this, in turn, relates to the irreversibility part. In theory, any tipping point can be reversed, if you have enough time and enough force at your disposal to push in the other direction. Hence the use of the word ‘effectively’ – we generally don’t have the necessary forces, and the time scales are generally beyond human lifetimes.
We can now use these definitions in our exploration. First, we can note a significant similarity – a change in system behaviour. For systemic risk it is in the definition; for tipping points it is implied. For a system to cross a tipping point, the reinforcing feedback loops must overpower the balancing loops, and this is nothing other than a change in system behaviour.
Second, we can note a meaningful difference. Systemic risk emphasises the consequence (an outcome we do not want), and not the trigger. Conversely, tipping points are all about the trigger (do we tip, or not), and not really about the consequence. This last statement requires clarification. A large part of the conference focused on ‘positive tipping points’. In other words, on system changes that are desirable, and that might keep us away from ‘negative (undesirable) tipping points’. We can therefore think of tipping points as ‘sensitive intervention points’ within a system and its dynamics. At a sensitive intervention point a little extra effort produces a big payoff in terms of systems change. For positive tipping points, the subsequent irreversibility is our friend. For negative tipping points, it is our enemy.
Having identified both similarity and difference, can we conclude anything about how tipping points and systemic risk are related? I propose that it is a relationship of the parts (tipping points) to the whole (systemic risk). From our earlier work on systemic risk (see the footnote) the hardest concept to explain was that systemic risk was a property of the system, and therefore inseparable from it. This is alien to how most of us have been trained, and to how we practice investment. We explain what we can and use a residual to carry what we can’t explain. But that is not how systemic risk works. Either the system works as we want it to (systemic risk is just a threat), or the system doesn’t work (the risk has manifest)[2]. Knowing which tipping point we crossed to stop the system working is perhaps less valuable at that stage.
Systemic risk emphasises the consequence… Conversely, tipping points are all about the trigger.
That said, if my interest is in what might cause systemic risk to manifest, rather than in a broader understanding, then tipping points become an important analytical tool. Assuming we agree that ‘systemic risk manifesting’ means ‘systems collapse’, then whatever the route – running out of resources, collapsing under its own weight of complexity, running out of insurance etc – we will have crossed some tipping point or another.
A final question we should probably ask and answer is ‘does any of this matter?’. One of our team mantras is ‘we write to think’, and so I am grateful for this opportunity to improve my thinking. Our previous work deepened our understanding of systemic risk, while this piece has sharpened my attention on the possible triggers. However, perhaps more important was my initial web search, which suggests that systemic risk is a term used by the insiders of the financial system. Rather than worry about definitions, we should perhaps look outside more, where scientists are telling us that Earth-system tipping points are now very close. Crossing them will change how we eat, how we trade, and how the financial system behaves … forever.
Tim Hodgson is co-founder and head of research of the Thinking Ahead Institute at WTW, an innovation network of asset owners and asset managers committed to better mapping the future.
References
- We have written extensively on systemic risk. See Systemic risk | deepening our understanding and Systemic risk | adapting our practices, both Thinking Ahead Institute
- The separation route can be traced back to Knight’s distinction between risk (the knowable) and uncertainty (the unknowable). A better analogy for systemic risk, perhaps, is Schrödinger’s cat – in the sense that we can consider the system to be poised between the two states of working and not working.