Emotions often drive financial decisions more than we think. Professor David Tuckett's research reveals how conviction narratives shape risk perception and decision-making. His findings challenge traditional views of rationality, suggesting emotions play
display-middle-ads-here
Introduction:
The Power of Conviction in Uncertainty
What drives
a financial professional to make a risky trade or hold back in times of risk?
We often
assume that risk management decisions are based purely on logic and data, yet
research suggests that emotion plays a much larger role than we might expect.
Back in the
October edition of Connect, we spoke to Brandon Davis, from the University of
Buckingham, about a groundbreaking study to uncover how emotions influence
financial choices—and how understanding this dynamic could reshape risk
management.
In this
edition of the magazine another member of that academic team, Professor David
Tuckett, Director of the UCL Centre on Decision-Making Uncertainty, explains
the extent to which emotion informs the way humans interpret external data in
order to reach good decisions around risk appetite.
The study,
being run in partnership with CeFPro, also includes experts from Princeton
University.
Tuckett’s
Conviction Narrative Theory suggests that financial professionals rely on
conviction narratives – stories they tell themselves to make sense of complex,
uncertain information – when making high-stakes decisions.
Decoding
Conviction Narrative Theory
Conviction
Narrative Theory (CNT) challenges the traditional notion that rationality alone
drives financial decision-making. Instead, Tuckett explains that “under radical
uncertainty – that is, contexts in which at the time you make the decision you
can't know the outcome – people draw on conviction narratives to support their
actions.”
These
narratives act as mental shortcuts, helping us to navigate an overwhelming
amount of data, filtering information, highlighting perceived risks and
opportunities, and creating an emotional response that guides action.
As Tuckett
puts it, “the brain is a constant interaction between thinking and feeling.”
Put another
way, CRO reading a data report isn’t just processing statistics; they’re also
experiencing gut reactions that influence the way they interpret that data.
To test this
theory, Tuckett and his team are conducting an online experiment – open to
Connect readers – in which participants read different financial texts and then
indicate whether they would buy or sell a particular asset, or accept or
decline a particular risk.
By analyzing
how specific words and emotional cues influence decisions, the study aims to
determine whether financial professionals are subconsciously guided by
emotion-laden narratives.
How Words
Shape Financial Risk Perception
The research
builds on previous studies showing that language can signal shifts in market
sentiment. By applying CNT to large text databases – such as Reuters News or
the Bank of England – the team has already demonstrated that emotional language
in financial reports can anticipate market trends.
One striking
example, Tuckett says, comes from their analysis of news coverage in the years
leading up to the 2008 financial crisis.
“In one
result we found was that if you plot a graph showing the relative amount of
excitement minus doubt or anxiety in articles about Fanny Mae, you could see an
upward trend right up until the crash happened,” he says.
“But very
interestingly, if you plot that graph against that the Case-Shiller housing
price index, you see that in 2006, that began to turn down. Yet it had no
impact on sentiment or the share price.”
This
suggests that financial markets may be slow to react to fundamental changes
because prevailing some and exciting “phantastic object” narratives override
emerging warning signs.
If traders
and analysts remain emotionally invested in such exciting stories, they may
ignore factual data that contradicts it.
Testing
the Theory: An AI-Driven Future
Tuckett’s
project is not just about understanding how narratives influence decisions – it’s
also about applying this knowledge to improve financial forecasting.
By
identifying which words and phrases trigger approach (confidence) or avoidance
(fear), the team hopes to build AI models capable of detecting hidden market
risks.
“Supposing
you ran these AI models over analyst reports of different investment
opportunities,” Tuckett explains, “they would probably pick up things that
could not be detected just by reading and making sense of the words.”
This could
be an invaluable tool for financial institutions, allowing those organizations
of scale to analyze internal reports for unseen risks or even predict shifts in
market sentiment before they happen.
The
experiment will culminate in AI models designed to analyze financial texts more
accurately, providing traders and analysts with an early warning system for
emerging risks.
“The survey results
should be completed by March, and the AI models we’re building based on them
should be available in May,” says Tuckett.