Neuroeconomics: What Brain Scans Can Say About Investing
Table of Contents
Suppose a headline says a brain scan explains why investors panic when prices fall. What did the researchers actually observe: people’s choices, a difference in a scan, or selling during a real downturn? Neuroeconomics combines economic decision tasks with measurements or other evidence about the brain. It can help explain how people evaluate risk. The strength of the explanation depends on what the researchers changed, what they measured, and whom they studied.
What Is Neuroeconomics?
In a typical experiment, researchers give people choices with specified payoffs and probabilities. They might record which option each person picks and use functional magnetic resonance imaging (fMRI) during the task. These studies use a signal that depends on changes in blood oxygenation. It is an indirect measure related to brain activity. The scan does not record thoughts or tell the researchers, by itself, how someone valued an option.[1]
Behavioral finance and prospect theory help describe patterns in observed choices. For example, prospect theory evaluates gains and losses relative to a reference point, such as the status quo. Loss aversion means an equivalent loss can carry more weight than a gain relative to that point. Neuroeconomics can ask which measurements track that pattern and whether other evidence supports a proposed mechanism.[2]
A Framing Experiment: Same Payoffs, Different Choices
In a 2006 study of 20 university students or graduates, De Martino and colleagues gave participants a starting amount and a choice between a sure payoff and a gamble. Consider the study’s £50 example. Keeping £20 and losing £30 of the starting amount both leave £20: £50 minus £30 equals £20. In either frame the gamble was the same: a 40% chance of keeping £50 and a 60% chance of keeping £0. Its expected payoff was 0.40 × £50 + 0.60 × £0 = £20. Only the wording of the sure payoff changed between these frames.[3]
Across the framing trials in that experiment, participants chose the gamble 42.9% of the time with the keep wording and 61.6% with the lose wording, a difference of 18.7 percentage points. The scan analysis found a stronger amygdala signal for choices that followed the prevailing pattern: taking the sure payoff with keep wording or gambling with lose wording. Changing the wording tested its effect on choice. The scan comparison linked a signal to the pattern of choices; it did not establish that the signal caused those choices. The guide to framing in finance shows how this distinction between wording and information matters in portfolio reporting.[3]
Behavioral Findings, Neural Correlates and Causal Evidence
Start by asking what the researchers varied. In a mixed gamble, accepting can mean gaining or losing money relative to the starting position. That lets researchers vary the size of a gain or loss, whereas the framing comparison above varied how the same amount was described. A study can change what participants see while only measuring how their brains respond.
| What researchers studied | What the evidence supports |
|---|---|
| Advice shown or withheld | Engelmann and colleagues (2009) showed an economist’s recommendation on half the scanned trials. The 24 participants chose between a sure payment and a lottery. Choices shifted toward the recommendation. The authors interpreted changes in some associations between estimated option values and scan signals as consistent with shifting part of the valuation task to the adviser.[4] |
| Scan signals during mixed gambles | Tom and colleagues (2007) studied 16 people recruited through UCLA campus ads. In several brain areas the signal rose with larger possible gains and fell with larger possible losses. The difference in sensitivity was linked to estimates of loss aversion within this sample. The study measured an association during a specific task.[5] |
| Choices with existing brain damage | De Martino and colleagues (2010) found lower estimated loss aversion on mixed gambles in two people with existing damage to both amygdalae compared with 12 matched controls. The comparison supports a possible role for the amygdala in the task, but the researchers did not assign the damage and one person’s damage extended beyond the amygdala.[6] |
These designs answer different questions. The advice experiment tests what happens when a recommendation appears. It does not show that participants stopped thinking or that following financial advice improves portfolio returns. The damage study bears on how a brain structure might contribute to a choice, but it cannot tell us whether a particular reader will turn down a gamble.
How to Evaluate a Brain-Based Investing Claim
- Identify the outcome. Did the study record a choice between two known payoffs, observe actual trades, or measure portfolio returns? A result about one does not establish the others.
- Ask what changed. Changing the wording or displaying a recommendation can test the effect of that change. A scan signal that varies with the choice may still be a consequence of the decision process, or it may reflect something else the task evokes.
- Check who and what were tested. Many repeated decisions from the same few people help describe their responses. They do not turn a small group into a representative sample of investors. A claim about a new group or real decisions needs evidence in that setting.
Common Mistakes When Reading Scan Results
Giving a brain region a single job. Saying “the amygdala was active, so participants felt fear” is an example of reverse inference: working backward from a brain signal to a particular mental state. Fear can be associated with an amygdala signal, but so can other states. A persuasive inference needs evidence that the measured pattern distinguishes fear from plausible alternatives in that setting.[7]
Reading a stable trait from one choice. Suppose a person can decline and gain $0, or take a hypothetical 50/50 gamble to gain $30 or lose $20. The expected gain is 0.50 × $30 + 0.50 × (−$20) = $5. Declining could be consistent with loss aversion, but it could also reflect a preference for certainty. One choice does not separate those explanations. The guide to cognitive biases in investing covers other behavioral patterns researchers may want to explain.[2] [6]
Where the Laboratory Evidence Stops
The decision experiments described here used small groups and tasks with stated odds. A portfolio decision may involve unknown odds, a longer horizon, and consequences outside the experiment. Even if a scan and a choice move together reliably in the lab, that does not show the scan adds useful information beyond observed choices for a new person. The studies here do not test a trading rule, estimate an individual investor’s risk tolerance, or establish a clinical diagnosis.
Frequently Asked Questions
Sources
- Logothetis et al. (2001), Neurophysiological investigation of the basis of the fMRI signal, Nature.
- Kahneman and Tversky (1979), Prospect Theory: An Analysis of Decision under Risk, Econometrica.
- De Martino et al. (2006), Frames, Biases, and Rational Decision-Making in the Human Brain, Science.
- Engelmann et al. (2009), Expert Financial Advice Neurobiologically “Offloads” Financial Decision-Making under Risk, PLOS ONE.
- Tom et al. (2007), The Neural Basis of Loss Aversion in Decision-Making Under Risk, Science; study methods and supplement.
- De Martino, Camerer and Adolphs (2010), Amygdala damage eliminates monetary loss aversion, PNAS.
- Yarkoni et al. (2011), Large-scale automated synthesis of human functional neuroimaging data, Nature Methods.
Disclaimer
This article is for education. It does not provide investment advice, medical advice, or a diagnostic assessment.
