Telling better stories about financial markets08.04.2017
In my recent blog posts, I referred to ‘narratives’ frequently, and it happens that the new AER issue of April 2017 contains Robert Shiller’s Presidential Address that is entitled ‘Narrative Economics’! And indeed, his examples mostly refer to historical events that involve a lot of coordinated behaviour in large groups, namely depressions and crises. His point is that narratives shape expectations and define the situations in which economic actors find themselves, and thus also influence their actions.
As I argued previously, in a mechanistic approach to financial markets, narratives might play a crucial role. Following Shiller’s account, this mainly relates to two aspects.
Firstly, narratives are carriers of contagion, they literally ‘go viral’. This implies, that we can use narratives to delineate the boundaries of a mechanism. A ‘herd’ must be defined according to certain shared characteristics. Thus, simply, the herd may be conceived as the group that is ‘infected’ by a peculiar narrative now shared in the group. That means, a mechanistic approach explicitly integrates the approach of epidemiology. Narratives are like ‘memes’ whose diffusion can be tracked via the analysis of media, what Shiller also shows. But this can be only the starting point. Serious epidemiology includes the analysis of network structures of contagion, for example, the role of ‘hubs’ in spreading a virus, which could be opinion leaders in spreading a narrative. I think that standard models of herding which refer to the incentives of financial actors in organizations can be included here.
Secondly, a narrative has a meaning, which helps actors to interpret a complex economic environment and may also trigger certain actions, operating like a ‘script’. That means, the narrative may be a reason to produce a certain action that actors present to themselves (in a sort of ‘inner speech’) and to others. But even more so, the narrative may contain certain images of actions which may become models for real action taken. In this sense, a narrative may also be a template for a mechanism that really works.
The second aspect is much more difficult than the first. I think that Shiller commits a serious mistake in taking only times of crisis as examples of narratives. This reflects the general tendency of behavioural economists to treat all the phenomena that are central to their concerns as dysfunctional, because they still regard them as being deviations from the standard of rationality. Narratives are ‘stories’, and as such do not reflect serious information. I think that narratives also undergird normal economic behaviour, and may also be the ground on which the healthy economy flourishes. The real economy is much too complex even for modern economics to fully understand its workings. Hence, narratives are the essential medium to ‘make sense’ of it and enable meaningful action: Sense and meaning are essential conditions of enabling rational decisions. Shiller himself analyses the diffusion of economic models as ‘narratives’, after all! So, I suggest that in a mechanistic model of financial markets narratives should become an indispensible building block also for understanding their regular operations.
This means that a special kind of collective behaviour, ultimately related to language, would obtain fundamental theoretical status. This is what the French ‘economics of conventions’ postulates (André Orleans also refers to financial markets). For the topic of herding, that raises intriguing questions about the role of narratives in establishing the collective that is demarcated as a ‘herd’. I think that shared narratives support the transition from what philosopher Tuomela calls the ‘I mode’ to the ‘We mode’ of behaviour, a kind of phase transition. The standard economic models of herding assume individual rationality and individual optimization. But if we conceive of herds as genuine collectives, what are the implications for the decision-making process?
One interesting possibility is that collectives may start to drive individual behaviour towards a more altruistic stance. Certainly, this is not supported by the professional culture in financial markets. But a ‘we group’ may increase levels of altruism, and we know from models of herding that this might overcome dysfunctions of information externalities: An altruistic trader might start to use her private information because she expects to improve decisions of other traders, and because their utility is integrated into her own utility function.
Perhaps we need to invent new narratives of financial markets and their stories of success? The market hero would not be someone who generates fortunes for herself, but who, while doing this, also takes responsible action towards other actors. Perhaps a better narrative of finance would look more like narratives about judges and doctors. Given the damage that havoc on financial markets can cause, that would not be a bad story, don’t you agree?
Shiller’s paper is at: https://www.aeaweb.org/articles?id=10.1257/aer.107.4.967