Corporate Finance

Big Data Will Yield Finance’s Big Stories, Nobelist Predicts

"Narrative economics" can reveal the truth behind the numbers, says Robert Shiller.
David KatzJune 13, 2017

As a rule, CFOs are rational individuals. Adept at thinking in terms of logic and numbers, they tend to look for quantitative explanations for the ups and downs in the valuation of their companies and of markets as a whole. They’re likely to assume that reality will conform to those models.

In that tendency, they aren’t alone. Since the 1970s, investors have increasingly subscribed to the notion that markets themselves are the bedrock of rational wisdom — that active investors can’t beat them without taking on a whole lot of risk

Drive Business Strategy and Growth

Drive Business Strategy and Growth

Learn how NetSuite Financial Management allows you to quickly and easily model what-if scenarios and generate reports.

As a response to this way of looking at things, known in economic terms as the efficient market hypothesis, mutual fund titans like Vanguard and Fidelity have assembled huge portfolios of index funds based on the idea that algorithms that basically track the capital markets can never be beat by the stock pickers.


Robert J. Shiller

Robert J. Shiller

Although this has become the dominant way of thinking about the movement of asset prices, CFOs may be overdoing it if they don’t account for the illogical, the irrational and, yes, the creative element in the human mind, says Robert J. Shiller, a co-winner of the 2013 Nobel Prize in economics and Sterling Professor of Economics at Yale University.

In fact, Shiller thinks that market volatility may well be driven by not only numerical factors, but by the stories people tell each other about financial conditions — no matter if they’re true or not.

Long a student of the irrational element in economics — some have suggested that Shiller should share credit with former Fed Chairman Alan Greenspan as the source of the concept of “irrational exuberance” — Shiller has turned his attention of late to the related concept of what he calls “narrative economics.”  

In a January discussion paper, he defines the term as “the study of the spread and dynamics of popular narratives, the stories, particularly those of human interest and emotion, and how these change through time, to understand economic fluctuations.”

Regarding recessions, he said, “[w]e have to consider the possibility that sometimes the dominant reason why a recession is severe is related to the prevalence and vividness of certain stories, not the purely economic feedback or multipliers that economists love to model. The field of economics should be expanded to include serious quantitative study of changing popular narratives.”

Ironically, Shiller notes, narrative economics runs counter to the ideas of Eugene Fama, the discoverer of the efficient markets hypothesis and one of two other economists with whom he shared the Nobel Prize. “I think one has to learn not to take glib explanations that make no reference to narrative seriously — like the efficient market hypothesis, for example,” Shiller said in an interview with CFO last week.

“CFOs probably take that too seriously as a general rule,” he added. “One has to respect one’s intuitive and human judgment.”

To be sure, finance and business people have always gone with their guts to some extent, Shiller acknowledges. What he wants now is to help prevent them from going to “the other extreme” of relying too much on automated, strictly market-based  solutions.

Referring to hard-line, efficient-market-type analysts, he asks, “in evaluating a company, you have to start thinking about why [an asset] is priced as it is. Do these people know what they’re doing?”

A Sudden Switch

To illustrate the concept of narrative economics, Shiller likes to point to the 1920-1921 Depression, an event not nearly as well known as the Great Depression that began in 1929, but one which nonetheless laid the groundwork for the narrative of the latter. The earlier downturn, in fact, “was the sharpest U.S. recession since modern statistics [were] available,” he noted in his discussion paper.

“The U.S. Consumer Price Index switched suddenly from inflation to deflation: between June of 1920 and June of 1921 … it fell 16%, the sharpest one-year deflation ever experienced in the United States,” he added.

The most influential explanation of that depression conforms to the rational model of economics. In their 1963 Monetary History of the United States, Milton Friedman and Anna J. Schwartz wrote that the 1920-21 contraction stemmed from the Federal Reserve’s error in raising the discount rate to trim inflation in 1919. The Fed, which had only been created in 1913, had to make that move to stem the high prices triggered by aggressive monetary policy.

Shiller disputes what he thinks of as Friedman’s purely quantitative explanation. More likely, “a multiplicity of factors, whose confluence produced such severity, caused people to cut back substantially on their expenditures, and these factors in turn often have the form of epidemics of narratives,” he writes in the discussion paper.

The availability of big data techniques will provide a quantitative basis for narrative economics, Schiller thinks. In trying to find out what stories were particularly contagious around the 1920-21 recession, for example, Shiller did word searches on  Proquest, which offers a database of newspaper articles going back as far as 1740.

Particularly illuminating was a search on the word “profiteer,” a word which began to see a revival during World War I. “I discovered by doing a word count that this word … had exploded in 1920,” the economist explains. “There were complaints about businesses making millions of dollars while our boys were fighting and dying.”

At the time, businesses weren’t really doing anything other than their normal practice of attempting to maximize profits, Shiller explains, acknowledging that the Fed’s monetary blunders had boosted inflation after the war.

On their own, however, those economic factors may have not been enough to trigger the crisis that ensued. Fueled by the narrative of excessive profit-taking, citizens suspected that “some kind of evil was taking over the the business world,” in Shiller’s account. Citizens boycotted businesses and refused to buy goods, exacerbating the situation and perhaps leading to a depression.

“What happened in 1920 was the great narrative basis for the Great Depression of 1929 because [people] really remembered 1920,” Shiller adds.

“This sounds like [a] crazy story, but I think it’s true,” he observes. “Economists don’t like it because it’s too undignified a story.”

One thorny issue Shiller struggles with in terms of narrative economics involves choosing which narratives will have a sustained influence on markets and which will turn out to be ephemeral. “How do you choose which narratives are significant? You can ask which narratives are on the rise and which ones are big,” he says.

Here again, Shiller points to the use of analysis of textual word counts as a tool to provide a quantitative basis for making such calls. For example, in using a database that counts mentions in the media, he found that the word “the” appears about 7% of the time in a typical article of 1,000 words. In December 2016, he noted, the name Trump rose to appearing about 0.10% of the time in a typical article.

“That’s amazing,” he says, suggesting that the story of Trump was a particularly contagious and largely self-created narrative. “Why is he tweeting? He probably shouldn’t do that as president, but he knows that keeping the gossip going is [essential to] his success story.”

Still, the question remains how CFOs and other business leaders can make practical use of narrative economics in performing such essential tasks as budgeting, capital allocation, and strategic planning. “You have to be a compiler of many narratives” and use “the combined effect of multiple narratives” to help with such decisions, Shiller says.

But a certain humility is a requirement, too. “You have to have some kind of respect for the magnitude of the problem  in understanding narratives, and to understand that there are many of them,” he adds.