Narrative Economics

Citation

Forbes, W. (2023), "Narrative Economics", Qualitative Research in Financial Markets, Vol. 15 No. 2, pp. 217-223. https://doi.org/10.1108/QRFM-02-2023-234

Expanding upon his 2017 Presidential address to the American Economic Association in Shiller (2017), Nobel Laureate in Economics 2013, in his latest book (Shiller, 2019) tries to build a consistent set of research practices for understanding the evolution of narratives. To understand an economic event, or value a company, we need to have a (valuation) story capable of being convincing and thus “going viral” amongst many investors. Such a story might explain how in the future consumers will demand “mobility services” rather than buying cars themselves, or open a “Libra” crypto-currency accounts.

Such stories might be especially needed when little is known about the prospects of the company being valued, or economic event being evaluated. Barker (1998) explains how fund managers understand this well enough (see Figure 1).

Barker (1998) reports the rankings given to fund analysts to the services analysts provide. He finds that the two most highly rated skills of analysts are the provision of company specific information and ideas that add value to a financial analysis of the corporation and a stock market recommendation to buy/sell/hold the stock. Earnings forecasts, a quantitative indicator of company value, is ranked fourth by fund managers as a service analysts can offer them.

1. The nature and purpose of narratives

So in constructing our measure of asset value stories form part of our calculus. Stories structure how we interpret quantitative measures of value and perhaps precede these quantitative measures in constructing an overall image of company value. In an uncertain world, outside the “small world” of Bayesian risk assessments, narratives allow us to consider possible states of the world, even if attributing frequencies to those outcomes is impossible (Brighton and Gigerenzer, 2012).

This leads to the central argument of Shiller (2019) that:

Economists can best advance their science by developing and incorporating it into the art of narrative economics (Shiller, 2019, p. xv).

In developing the study of narrative economics, Shiller (2019) focuses on two particular themes. These are:

1. word-of-mouth contagion in the form of stories, and

2. the effort people make to generate new contagious stories.

To understand economic developments entirely through the lens of data and valuation models can often seem rather misleading and almost pointless. Shiller (2019, pp. 71–72) stating:

Trying to understand major economic events by only looking at data on changes in economic aggregates; such a gross domestic product, wage rates and tax rates, runs the risk of missing the underlying motivations for change. Doing so is like trying to understand a religious awakening by looking at the cost of printing a religious tract.

Examples of such stories which Shiller considers are those surrounding the transformative power of Bitcoin, or the employment displacement that artificial intelligence (AI) will cause. Bitcoin is just one of a series of examples of “bubble psychology” and the way in which the impact of AI is now discussed is not that different to the way the influence of the steam engine, or computers, were perceived in their time. In 2017 alone over 900 initial coin offerings were made. Almost half failed within a year, demonstrating the power of a good story over sustained economic analysis as a driver of investment.

Much of what we understand about business practice and business success is gathered from legends of business life, Josiah Wedgewood, Howard Hughes, Steve Jobs, etc.

What Mokyr (2017) elsewhere has called the “cultural entrepreneur”, who creates, new markets and awareness of new needs (to socially network, or have access to “mobility solutions”). Mokyr (2017) cites the English Philosopher, David Hume, who stated:

What depends on a few persons is, in great measure, to be ascribed to chance, or secret and unknown causes; what arises from a great number may often be accounted for by determinate and known causes. [Hume, On the Rise of Progress in the Arts and Sciences, 1788, quoted in Shiller (2019, p. 71)].

This is especially true when we can construct a coherent, clearly bounded, reference class, like Unicorn (private companies worth $bn plus) in the social media space, or US defence contractors. We cannot say much about the yearly fortunes of Facebook perhaps. But maybe we can talk sensibly about how social media as a consumer/investment space, will evolve. This requires the development of compelling narratives of future competitive plays and their results.

Other narratives become reduced to powerful epigrams to be customised for personal use, like the “American dream”. Some storytellers outrank others. Shiller (2019, p. 254) notes stories about investment decisions, largely made by men, are more commonly discussed in historical accounts of the 1929 Crash and the depressed 1930s, rather than consumption decisions that were largely made by women. Yet President Roosevelt addressed himself to the woman of the household in seeking support for the “Buy in August” campaign of 1933, which aimed to raise the nation out of its depressive state.

So a powerful, viral, narrative needs to get the chain of transmission just right. The speaker needs to be seen as authentic, convincing, the narrative itself have vividness, moral authority and the hearer needs to be suitably primed to receive it. Configuration of these elements is key for, as Edgar notes in Hamlet, “Ripeness is all” [1].

Judging ripeness is a precarious business. Shiller (2019, p. 267) states the difficulty thus:

The various narratives that share the stage at any point are in a biological analogy, many cellular receptors and signalling molecules. Modern communication means that new and different types of epidemics are possible, and economic forecasting requires close attention to many different narratives (1).

Another strength a focus on narratives might bring is in liberating us from too restrictive a view of causality. So often when undertaking an accounting-based valuation we seek to construct a “theory of price without reference to price” (Penman, 1992). But it is far from clear how you do this in a world where balance sheet assets/liabilities are “marked to market”, putting market value on both the left- and right-hand side of the proposed valuation model. Thus to speak of price variations being caused by accounting variables seems a bit odd really.

Narratives can help here because they recognise the simultaneity that “everything causes everything else”. Events drive narratives. But narratives shape/motivate events. So the closure of the US border to European entrants has depressed European stock markets. But both that closure and its effects are joint products of populist nationalism and the need to “make America/India/Poland/Brazil/(fill in the blank) great again”. We might recognise these divisions in the Brexit debate, or the rise of Donald Trump and his appeal to the “forgotten man”.

2. Fitting facts to stories not stories to facts

Narratives can even be retrofitted to our past to suit our current need to glorifying winners and justify the impoverishment of losers. Shiller (2019, pp. 236–238) points to the emergence of a narrative of those who dodged the bullet of the 1929 Crash (sometimes JD Rockefeller, Bernard Baruch or Joseph Kennedy to one’s taste) bailed out of the market after a shoe shine boy gave them stock tips. Shiller points out this story seems first to have emerged in Bernard Beruch’s 1957 memoirs. This suggests narratives can be both post hoc ergo propter hoc rationalisation, as well as drivers of our current actions. Shiller (2019, p. 238) notes:

[…] certain stories that recur with mutations play a significant role in our lives. Stories and Legends of the past are scripts for the next boom or crash.

Thus, our Queen in reassuring the nation that we will defeat COVID-19 recalled her youthful address to evacuate children in the Second World War. The contexts were different, but the same personal resolve to endure hardship was required of the common (wo)men.

The diffusion of narratives can proceed along the lines of pre-existing social economic divisions. Shiller (2019, pp. 157–161) recalls this was to be the case in the dispute between “gold-bugs” and Silverites, after the imposition of the gold standard by the Coinage Act of 1873.

That Act provided for the conversion of dollars into gold at a rate of one 120.67 of gold for each dollar surrendered. As an amendment to the Coinage Act there was much discussion of simultaneously allowing the conversion of silver into gold at a rate of 16 ounces of silver for one ounce of gold. This was suggested as a reformed bimetallism standard for the imposition of much needed monetary discipline in the USA, which was then experiencing high inflation after the completion of the civil war.

By the late 1890s silver was trading at a price of roughly 30:1 with gold. So a debtor could clear their debts at half the price by converting silver to gold before clearing their debts up. Advocates of the bimetallist amendment were often based on the West coast and formed part of wild western tribes, who had often migrated from the East, now pushing the boundaries of the Union from coast to coast.

Eastern intellectuals loved to denounce silverists as rubes and charlatans. As Shiller (2019, p. 160) puts it:

Supporters of the gold standard tended to appreciate symphony performances, while Silverites liked to watch boxing matches.

This association between being a gold bug and at least tolerating, if not perpetuating, social injustice was embedded in William Jennings Bryan’s plea in 1896 not to “crucify mankind upon a cross of gold”.

3. Crafting the story

Narratives reflect and shape events, forming a cohesive whole outside causal chains. Many economic narratives form part of a self-fulfilling prophecy, when fears of a recession, induced by a global pandemic, induce an actual recession. So Shiller (2019, p. 114) reports, with some glee, the story of a man who goes into a bank and confronts the bank teller saying “If my money is here I don’t want it. But if it is not here I want it”.

To attain such viral traction, influencing economic narratives are constantly being edited, recirculated and withdrawn. In this they might be seen to form part of a broader metanarrative of our times, or chosen tribe.

Shiller points to an iterative process by which narratives mutate and recur. So the same narrative that predicted we would all be replaced by machines was redeployed in the dot.com era of the late 1990s to describe an era of untold riches (Mosco, 2005). Some narratives get transformed and resurface in new form.

The austerity, modesty and personal responsibility for poverty/ill-heath are very much out of fashion in our irreligious age. Perhaps dieting, working out and a return to natural, organic products, serve some of the same function for us. But some relics of the austere, virtuous life do remain. The popularity of bicycles and blue jeans dates from this time, when the educated middle class could identify with some part of the poor’s daily suffering.

One such recurrent narrative is that of the impoverishment of the masses at the feet of a technocratic elite. While Shiller points out this goes back to the 19th century and the “Swing riots” by city mobs displaced, or undercut, in the labour market by incoming machinery. But popular concern about our replacement by a technocratic elite really got going during the descent into the slump of the 1930s, with the emergence of the first large trust companies and their embrace of modern technology like the dial-up telephone.

Initially telephones when lifted simply put you through to the operator of the telephone company, who took the number of the person you wanted to contact and placed the call. Self-dialling phones cut out a swathe of humanity who acted as operators connecting calls. Senator Carter Glass of the US Senate surely spoke for all decent folk (quoted in Shiller, 2019, p. 191) when he said:

I object into being transformed into one of the employees of the telephone company without being compensated.

3.1 Some stories of our time

Currently the rise of AI is being fused with concerns about intensifying economic inequality as the lifestyles of the educated middle class are being threatened by the rise of AI applications. Past technology threatened the jobs of manual workers, in shipyards, coal mines and transport. Today teachers, radiographers and accountants are losing their jobs to have their judgements replaced by machines. Here, technological displacement forms part of a bigger discussion of maintaining social cohesion in a period of technological progress.

The arguments used to justify the establishment of the Federal Reserve were originally understood to apply to industrial titans, like JP Morgan, who corralled other industrial “robber barons” into buying up US stock during the 1907 stock market panic.

Here, the moral rectitude of the most wealthy, willing to reverse an emerging national panic, was transformed in a justification of (Keynesian) active macroeconomic policy management (Shiller, 2019, p. 119). When a story alerts us to perceived injustice in can be far more powerful in initiating action. Shiller (2019, p. 256) points to the OPEC Oil embargo, in response to US support for Israel during the 1973 Yom Kippur war, as an example of a powerful, anger-driven narrative.

Elsewhere, Shiller (2019, pp. 258, 260) points to the way in which the short-sighted greed of unions, sometimes led by criminal figures, like Jimmy Hoffa, discredited industrial action as a way to extract rents for “insiders”, as opposed to unorganised “outsiders”, excluded from competing for union member’s jobs (Lindbeck and Snower, 1989). Fans of British Cinema’s “I’m Alright Jack” and “The Man in the White Suit” may recognise the style of such charlatans.

Shiller (1997) in his study of why people dislike inflation marvels at the fact that in his interviews, with ordinary citizens in three countries, nobody seems to like inflation, despite the fact that many folk holding substantial assets, must have benefited from its presence. All seemed focused on how inflation raised their costs, without ever reflecting on how inflation raised their incomes and the value of their assets relative to the debt they took on to acquire those assets initially. This suggests powerful, viral narratives may not be true ones.

Vosoughi et al. (2018) examine 126,000 news stories spread on Twitter by some 3 million people over 4.5 million times (with a few retweets in there it seems). The authors conclude:

Contrary to conventional wisdom, robots accelerated the spread of true and false news at the same rate, implying that false news spreads more than truth because humans, not robots, are more likely to spread it. (Vosoughi et al., 2018, p. 1146)

Hence, our consideration of the power of narrative in forming asset value must consider both its role in generating noise around true asset value and in indicating true, long-term value itself. But investors can display a robust distain for even obvious warnings of asset mispricing if they jar with the prevailing market trends. Shiller (2019, p. 232) notes how investors were blind to the expansion of implied volatilities, embedded in stock option prices, as well as the surging aggregate P/E ratio in the run up to the 1929 Crash. Perhaps we just forget how intoxicating the joy of the bubble felt after it imploded (Galbraith, 1990).

4. The story unfolds

For those seeking a “how to” guide to using narratives to explain/predict economic/financial events Shiller (2019) will be something of a disappointment. Shiller (2019, p. 271), clearly states:

When it comes to predicting economic events, one becomes painfully aware that there is no exact science to understanding the impact of narratives of the economy. But there can be exact research methods that contribute to such an understanding.

So it seems what lies ahead of us is a struggle of competing visions, enabled by competing methods. As in much of social science, often substance will be bent and deformed to shoe the economic/financial problem into the chosen method’s grasp (Gigerenzer, 1991).

Nevertheless, Shiller (2019, p. 279) remains optimistic about what can be ultimately achieved:

Economists can now collect data on economic narratives themselves, on their essential elements of meaning, without being overly focused just on words, and they can model the transmission of narratives. If we maintain quantitative rigor, we can make narrative epidemics a part of economic science.

In defence of this view we can see some early fruit of the application of textual analysis to the analysis of financial markets (Boudoukh et al., 2019; Bowden et al., 2018; Nekrasov et al., 2020). I hope we can participate in the establishment of some rules of the road for examining textual data to unravel and employ important economic narratives as part of our investment management process. Shiller’s book convinces me it is an opportune time to do so.

Like any new shiny toy, the question arises how does it fit into the many excellent econometric methods we already have and which made the Finance group the most revered group in many Business Schools worldwide. Gabriel (2000) makes the point that the power of narratives emerges from melding key facts together to give meaningful stories about our past and possible, shared futures. The most post-modern of colleagues can sometimes give the impression that “mere facts” are beneath their search for metanarratives. But narratives can never transcend observed data and should not want to. So a narrative economic approach must seek to enhance, refine, our standard financial econometric tools that many of us have spent the past few decades learning and now teaching. So the future is neither for financial ecometricians or textual analysts but an exciting, enabling, mixture of both.

Figures

Figure 1.

Bar plot based on Table 1 page 10, Barker (1998)

Note

1. Hamlet, Scene 5, Scene 2.