The Choice-of-Future Problem in Systemic Investing (Part III): Discussion and Special Topics
This is the third article in a three-part series on the Choice-of-Future Problem in systemic investing. Part I introduces the problem and Part II presents a framework for addressing it. These three parts are meant to be read sequentially, as each builds on the preceding one.
Part II of this article series presents a set of strategies and a decision tree that together form a framework for working through the Choice-of-Future Problem in systemic investing. This framework is a stylized model, making assumptions and raising questions that warrant discussions.
On the Rigor of the Framework
Issues of Measurability
Our framework assumes that the relative probabilities and preferabilities of different plausible futures are measurable. In practice, this might be tricky to do with accuracy.
When it comes to probabilities, the key issue is that predicting the future is a famously uncertain undertaking. But this does not mean that each plausible future should be seen as equally probable.
In fact, we believe it is possible to develop a sense of relative probabilities that is tangible enough to provide useful guidance for decision-making. This should be particularly true when considering competing versions of the future that are “far apart” from each other.
For instance, it is difficult to reliably predict whether all-electric vehicles or hydrogen-based vehicles will be the dominant technology platform of the future for passenger vehicles. But at the moment, EVs are out-selling hydrogen cars by orders of magnitude, and researchers predict that hydrogen cars will be unable to compete with EVs. So even in the absence of quantified probabilities, these two signals suggest that an all-electric future is more likely, or at least that betting on hydrogen is a more risky investment proposition.
Similarly, our model assumes that it is possible to evaluate the relative preferabilities of different futures. In practice, this might also be difficult. One reason is that the tools we have at our disposal to assess the relative benefits of different technologies — such as life-cycle analysis — often evaluate only a single metric (e.g., CO2e emissions) and therefore only allow an incomplete assessment.
We posit that the difficulty of comparing relative preferabilities increases as a function of the complexity of the desired outcomes. It is relatively straightforward to compare the well-to-wheel CO2e emission profiles of all-electric vs. hydrogen-based vehicles. It is much more difficult to assess both technologies on a system’s level while considering a broader set of societal impact factors, such as transportation equity and transportation justice.
And yet, in many cases, it should be possible to approximate the question of relative preferability in a way that leads to actionable insights. And where this is not possible, a pragmatic approach could be to consider each version of the future as (more or less) equally preferable and thus choose a future-neutral or multi-future investment strategy (cf. Part II).
The “It Depends” Problem
Even in cases where a particular version of the future can be deemed preferable on an ex-ante basis, just how preferable it actually is will depend on how, exactly, this particular version is going to eventuate. Electric vehicles with batteries that wreak environmental and social havoc in the countries that mine rare earth materials will solve one problem but create many others in a way that makes one question the net benefit of EVs.
The “it depends” problem does not just exist within a system but also arises as a result of the interdependence of systems. For instance, electric mobility in a country that relies on fossil fuels for its electricity production is not going to have a much better environmental footprint than a mobility system based on cars burning gasoline or diesel.
In any case, this dependency on implementation does not mean that all versions of the future are equally desirable. For example, even within the boundaries of the many evolutionary possibilities of a specific energy system and allowing for unintended consequences, negative second-order effects, and the ambiguity emerging from systemic interdependencies, a system configured around, say, a combination of renewable energy and battery storage is almost certainly going to be better — in all relevant aspects — than one based on fossil fuels with carbon capture and storage.
So yes, in many instances, the answer to the question of which future is better will indeed be “it depends”, but investors can still approximate the question of relative preferability in a way that allows defensible choices.
What the “it depends” problem points to is the importance of continually trying to ensure that the system evolves in a desired direction in a suitable manner. This lies at the heart of the concept of nesting — the continuous, proactive strategic alignment of investments with other actors in society and the engagement of a diverse set of levers of change.
The “Fruit Basket” Problem
A classic fruit basket contains several types of fruit (apples, oranges, bananas), several varieties of each fruit type (for apples: Pink Lady, Gala, Granny Smith), and several units for each sub-type (multiple Gala apples). If you want to determine which particular piece of fruit amongst all the pieces in the fruit basket is the best, you first have to decide the basis on which to make your decision.
Will you compare an individual piece of fruit with all the fruits in a basket (all apples, all oranges, all bananas)? With all the fruits within the same fruit category (all apples)? Or with all the fruits within a specific fruit variety (all Gala apples)? Or will you compare the fruit basket at large with a vegetable basket or a cheese board? Depending on the answer, you might end up comparing an apple with an orange, a Granny Smith with a Pink Lady, different units of Gala apples with each other, or fruit with vegetable or dairy products.
The nature of the “Fruit Basket” problem is one of drawing system boundaries. And is a problem that systemic investors face when trying to evaluate different versions of the future.
Consider the case of net-zero emissions technologies (NETs). Are investors interested in NETs going to compare different approaches to kelp farming (one Gala apple with another Gala apple) with each other? Are they going to evaluate kelp farming with rock weatherization (Gala vs. Pink Lady)? Or are they going to compare the entire group of NETs with climate mitigation strategies that are not focused on the removal of emissions but on their avoidance, such as renewable energy (apples with oranges)?
Answering these questions is not trivial.
The “silver buckshot” metaphor implies that we will need multiple different solutions (types of fruits). So, from that perspective, any set of solutions passing the Improvement Test (cf. Part II) would qualify, including NETs.
However, research shows that the promotion of too many technological solutions creates the risk of delaying the adoption of clean technologies more broadly, mainly because the “silver buckshot” approach makes it harder to unseat the incumbent technology. So from this perspective, focusing on one type of solution (apples) is preferable.
There is also the issue that a particular solution might inadvertently prolong the viability of an incumbent system that needs to be replaced. For instance, the ability to artificially remove CO2 from the atmosphere might delay the phase-out of fossil fuels, both directly (because it mitigates the harm fossil fuels do) and indirectly (because it gives fossil fuel executives a narrative to fend off stricter regulation). So considering second-order effects becomes key (like whether not eating the apple will spoil the bananas).
What all of this boils down to is that there is no straightforward resolution to the “Fruit Basket” problem — or at least not one we can present in the context of this article. What is clear is that this particular debate is a great example of why climate change is called a wicked problem.
On the Practical Implications of the Framework
Generalizability
The question of which future to support is relevant for any impact investor; in fact, it is relevant for all investors. This is because each individual investment embodies a vision of the future and is impactful in its own right, whether those impacts are intended or not.
In other words, it is not possible for investors not to have impact, whether they invest systemically or according to traditional paradigms. Consequently, the model presented in this article series should be relevant for any (impact) investor.
That being said, there is an implicit question here about what imperative should dominate an impact investor’s choices. Is it a moral imperative to promote the most sustainable future possible? Is there an additionality imperative dictating that impact investors should go where nobody else is going? Or is there a liberal imperative dictating that impact investors are free to choose which future to support?
This is a fundamental question — with important moral and ethical dimensions — whose discussion is beyond the scope of this essay.
Values and Legitimacy in Determining Preferability
The concept of preferability is also fraught with another challenge: Who determines what is preferable and what is not?
In general, preference is a subjective concept. What we prefer depends in large part on what we value. In theory, this leads to an infinite number of preferences across the societal spectrum.
We posit that, in order to manage this complexity, purpose-driven investors should do two things. First, they should subscribe to a high-level philosophy that supplies a set of first principles for evaluating trade-offs and informing decisions, such as humanism.
Second, they must look for commonalities across this large set of preferences to reduce complexity to an actionable level. For instance, most people value clean air, toxin-free water, and nutritious food. These shared values form the bedrock on which a more nuanced definition of preferability can be conceived.
That said, the real issue is not so much the lack of commonality in describing a future worth building but the difference in opinion of how to get there. Here, it is important that investors do not impose their own worldview onto the rest of society. As wielders of financial capital, they possess outsize power. If they get to set the agenda, they will create a plutocracy.
So how can purpose-driven investors ensure the version of the future they are championing is legitimized?
In our systemic investing prototype around net-zero mobility in Switzerland, we have solved this issue by adopting the long-term policy plan of the Swiss federal government. Switzerland is a well-functioning democracy (ranking #12 on the Economist’s Democracy Index), so the government’s long-term policy plan is a democratically legitimate source of preferability. In contrast, in Vietnam — which ranks #137 on that same index and where we have also done some work — sticking to a government policy plan would be more problematic.
Irrespective of how the problem of legitimization is resolved — whether by adopting an official policy or through some other mechanism — we believe the position of power from which investors deploy capital creates an imperative to ensure their preferred version of the future is democratically legitimized.
The Lure of Utopia
Some might argue that purpose-driven investors should, by default, push for the most radically sustainable future possible. For instance, in mobility, they should not just aim to accelerate the substitution of one drivetrain with another but also tackle structural issues of urban planning, aspects of inequities in access to mobility, and the many environmental and social problems in the battery supply chain. In other words, the argument is that impact investors should focus on Horizon 3 futures without compromise.
Those visionaries are needed. And sometimes, the most radically desirable future is indeed within reach and thus investable. However, we posit that more often than not those Horizon 3 futures are too speculative for investment capital to engage with. This is because part of the DNA of (impact) investing is the idea that capital preservation is the minimum objective and risk-adjusted capital multiplication is the ideal scenario. This is what differentiates investing from philanthropy.
So there is something to be said about financial pragmatism in investing, and it would be unreasonable to expect investors to invest in futures that are too speculative. In practice, this might inevitably lead to a bias for Horizon 2 solutions, creating the need to find creative ways of stretching into Horizon 2+/3 possibilities.
This is where the concept of funding architecture comes into play. By blending market-rate investment capital with concessionary capital, philanthropic funding, and public finance, it becomes possible to finance propositions across the entire “futures horizon”.
On the Imperative to Generate Additionality
One interesting problem in systemic investing is the question of whether investors deploy capital only into bottlenecks (or leverage points) within a system and whether pure scale-up plays are off-limits. This, essentially, is a question about the imperative of additionality.
In sustainability, additionality “refers to the extent to which an action or project contributes to sustainability goals beyond what would have happened anyway in the absence of that action or project” (Source: ChatGPT).
Consider the case of the Swiss mobility system. Our system analysis has shown that there is a lack of technicians capable of installing charging stations (a bottleneck) and a lack of pace in building charging infrastructure (a scale-up problem). Should systemic investors deploy capital only in the hard-to-solve problem of training technicians or also in the straightforward infrastructure project finance space?
We posit that whether there is an additionality imperative in impact investing is largely a moral question. There is nothing in the theory or rationale of systemic investing that suggests only additional actions should be funded.
In fact, we believe there is a compelling case to be made to use investments with favorable risk-return characteristics (e.g., a project finance transaction) to minimize the overall risk in a portfolio. Indeed, the existence and benefits of such cross-subsidization effects have been demonstrated in the context of urban climate projects by a study published by the European Investment Bank: from a whole-of-portfolio perspective, adding projects with favorable risk/return characteristics enables investments in more audacious (i.e., risky) propositions.
Special Topics
Dealing with Uncertainty: Hedging and TEP-Hopping
Our model raises important questions about how to mitigate the risk emerging from the fundamental uncertainty inherent in trying to steward the evolution of systems. How can investors reduce those risks? What are hedging mechanisms?
Organic hedging might happen when a portfolio contains assets that are either future agnostic (e.g., a start-up developing a fleet management app that can be used both for electric vehicles and hydrogen-based vehicles) or adaptive (e.g., a software product monitoring battery performance that can be tweaked to monitor hydrogen fuel cells).
Another potentially promising approach to hedging adverse evolutionary risk is adding a public equity component to the portfolio. For instance, a systemic investment portfolio focused on electric mobility could mix in a basket of public stocks relevant to the hydrogen economy.
It might also be worth considering “hopping” onto another TEP midway through a systemic investment program’s term should it start to become increasingly clear that a system to which investors have committed initially is going to be outcompeted by another. Whilst this will likely mean that the initial investments will not perform well, such change-of-mind is in essence no different than a venture fund writing off the majority of its investments or public investors changing their asset allocations to reduce exposure to under-performing sectors.
Providing a detailed framework for risk hedging in systemic investing is beyond the scope of this article series, though something we are already working on.
Questions of Diversification
In traditional investing, Modern Portfolio Theory suggests that the risk-minimizing capital allocation strategy is that which achieves diversification, i.e. spreads investment capital across a wide range of economic sectors, technology concepts, companies, projects, and asset classes.
In contrast, systemic investing is concentrated and high-correlation investing. It is concentrated because investors choose to focus on specific systems and, within those, on one or more specific TEPs. It is high-correlation investing because investors seek to bring investments in strategic alignment with each other in search of combinatorial effects.
What does this mean for how investors need to think about diversification?
What should matter to investors is that their assets are diversified at the level of their entire investment portfolio. This means that as long as they do not put all their capital into a single systems transformation challenge, they can still achieve adequate diversification by (i) focusing on multiple TEPs within a system, (ii) focusing on multiple systems, and (iii) investing capital in more traditional ways (e.g., publicly-traded stocks)
Conclusions
What all of these musings across the three parts of this series suggest is that the Choice-of-Future problem is much more complex than it appears. So, in our view, it is surprising — and perhaps a bit concerning — that there is not a greater normative debate about how purpose-driven investors should go about choosing which future to promote.
Some approaches, such as Effective Altruism, are providing guidance on the basis of high-level principles. However, the devil is not so much in the principle (“Focus on issues that are important”, an EA tenet) but in the implementation of the principle. This is because it is in the specific context of a system in which issues like measurability and legitimacy become meaningful.
All of this points to a structural problem — that we are not reflecting enough about where to direct impact capital. Perhaps that’s a reason why there’s such a large gap between where money needs to flow and where it is actually going.
