The Principles for Responsible Investment Secretariat recently published guidance for investors on managing risks arising from interactions between climate change and local water supplies and the resulting impacts on the agricultural supply chain. This sophisticated approach to managing risks is illustrative of how academics, civil society groups, beneficiaries and policy makers are increasingly are challenging asset owners and their managers to incorporate complex and interdependent issues in their investment analysis.
This is tall order for an industry that only a decade ago began to recognize that environmental, social and governance (ESG) factors could have an impact on the financial performance of their portfolios. But the emergence of sophisticated intermediaries, sustainability reporting standards, collaborative platforms and technology firms dedicated to structuring big data to provide real-time ESG analytics all provide reason to be optimistic about the capacity of investors to their investment portfolios.
Still, others are calling for a deeper commitment from investors, drawing attention to systemic-level risks and opportunities that are intimately linked to performance at the portfolio-level. In the past, the portfolio-level was the only basis on which investors could reasonably be expected to manage risks. But as the significance of the financial industry relative to the economy has grown, alongside the diminished influence of regulators on global financial markets, investors now require an integrated understanding of the relationship between portfolio-level risks and wider environmental, societal and financial systems-level risks (Lydenberg 2015).
What is surprising about such an ambitious agenda is its relatively low ambition for changing the infrastructure of financial decision-making. When listing the changes that would be required to shift to a focus on systemic frameworks the author suggests that,”asset owners will manage their investments along two parallel, but separate tracks – with differing metrics for measurement.” On the one hand, systems-level measurement would require “directional and qualitative assessments of impacts on the sustainability.” On the other hand, “the measurement of portfolio returns will remain much as it is today – that is, in relation to market prices (Lydenberg 2015, p. 25).”
But this conventional approach to managing investments is what has contributed to making systemic risks so dangerous in the first place. Convention requires ignoring externalities that cannot be understood in terms of risk and uncertainty that can be reduced to risk. It is not difficult to imagine cases where portfolio-level decisions and systems-level decisions would be in direct competition with each other. And we know that when faced with two tasks, agents will dedicate effort to the task that is easier to measure and to reward (i.e., the portfolio-level).
Integrated investing requires integrative thinking. Integrative thinking in management studies is often used to explain a quality of successful leaders that are able to hold “conflicting ideas in constructive tension, and use the tension to think their way toward new and superior ideas.” Among their most defining qualities, integrative thinkers have a higher tolerance for ambiguity.
While integrative thinking is often treated as a skill for leaders to cultivate, this ignores the institutional environment in which financial decision-makers operate. For example, accounting students exhibit a lower tolerance of ambiguity than students in other disciplines. In one study, researchers find that “low ambiguity-tolerant individuals … are more likely to ignore or discount ambiguous information when ambiguity relates to positive information.” Integrated thinking requires institutional and organizational level changes to introduce new decision – making frameworks and incentive systems that accommodate decision problems for which there is no right answer.
But this forces another question: why do our institutions and decision-making frameworks favour decision problems that have a right answer? What would lead us to assign more weight to the portfolio-level risks than to systems level risks in instances where there is a conflict between them?
In a thought experiment conducted in the 1960’s, Ellsberg provides a behavioural answer to this question. Agents are given the choice between betting on drawing a coloured ball from two urns; one urn where probabilities are known and the other where probabilities are not known. The result is that decision-makers have a strong preference for betting on the urn for which probabilities are known over the urn where probabilities are not known. The violated Leonard Savage’s Sure Thing Principle, a pillar of the subjective decision-making theory that allowed decision-makers to extend the ‘machinery of probability’ to guide decisions under conditions of uncertainty (see Gilboa 2009). In other words, decision-makers should be indifferent between betting on the two urns.
A behaviouralist explanation for this result is that individuals are only averse to ambiguity when making decisions that they know will be compared to a more knowledgeable decision-maker. So moving from the realm of coloured balls to politics or sports, when faced with a choice to bet on an event, our aversion to ambiguity depends on who else is betting alongside us. If they know more about sports than we do, we will choose not to bet against them. But when making decisions about events on which no one has sufficient information, ambiguity aversion disappears.
For many observers, behavioural biases justify the continued use of decision-making frameworks that assume all problems can be treated as risk or uncertainty that can be reduced to risk with more and better information and with subjective probabilities. Once we give way to the possibility that uncertainty as permanent, we lose the war against uncertainty.
In his conclusion to “Against the Gods”, Peter Bernstein warns that ” we must avoid rejecting numbers when they show more promise of accuracy than intuition and hunch, where, as Kahneman and Tverksy have demonstrated, inconsistency and myopia so often prevail.”
But as Ellsberg and those following in the behaivoural tradition have shown, our need for certainty (and consistency) is itself a bias, just like myopia, confirmation bias and other pathologies that affect our decision-making abilities. Certainty is an increasingly dangerous bias in a complex environment. It leads us to continue to see problems in terms of rigid trade-offs, and to assume that there is always a right decision and that the remaining decisions are wrong. This thinking is the reason that systemic-level risks arise in the first place. There will always be some externalities that are too messy to incorporate into financial decision-making, no matter how sophisticated our models and how ‘big’ our data.
Until the frameworks that govern investment decision-making change, sustainable investors that are willing to take up the ambitious goals of integrated investing will need to learn to think and to act in shades of grey, but continue to justify their actions in black and white.
If this conclusion is unsatisfying, then perhaps we can find inspiration in other disciplines that have built in a greater tolerance for ambiguity. For example, law has jurisprudence; a conceptual space where legal scholars and student of law can contemplate and debate legal systems and internal contradictions inherent in the theory of law. Finance needs the same: a space where students of accounting and finance can contemplate meaning of financial systems and the problems internal to finance theory and financial systems.*
* This idea is presented by Guy Fraser Sampson (2014) in ‘The Pillars of Finance’, Palgrave MacMillan.
Lydenberg, S. (2015). Portfolios and Systemic Framework Integration: Towards a Theory and Practice. Exposure Draft, November 16.