Relationship of striatal and anterior insula activity with real-life investment success

896 

Abstract Submission 

Stephan Nebe^1,2, Nick Sidorenko^1,2, Alexandra Bagaïni^1,3, Steve Heinke³, Silvia Maier², Alexandre Ziegler⁴, Isabella Kooij², Kevin Trutmann³, Thorsten Hens⁴, Jörg Rieskamp^3,5, Rui Mata^3,6,5, Philippe Tobler^2,5

¹Shared First-Authorship, *, ²Department of Economics, University of Zurich, Zurich, Switzerland, ³Department of Psychology, University of Basel, Basel, Switzerland, ⁴Department of Banking and Finance, University of Zurich, Zurich, Switzerland, ⁵Shared Senior-Authorship, *, ⁶Max Planck Institute for Human Development, Berlin, Germany

Stephan Nebe  
Shared First-Authorship|Department of Economics, University of Zurich
*|Zurich, Switzerland

Nick Sidorenko  
Shared First-Authorship|Department of Economics, University of Zurich
*|Zurich, Switzerland

Alexandra Bagaïni  
Shared First-Authorship|Department of Psychology, University of Basel
*|Basel, Switzerland

Steve Heinke  
Department of Psychology, University of Basel
Basel, Switzerland

Silvia Maier  
Department of Economics, University of Zurich
Zurich, Switzerland

Alexandre Ziegler  
Department of Banking and Finance, University of Zurich
Zurich, Switzerland

Isabella Kooij  
Department of Economics, University of Zurich
Zurich, Switzerland

Kevin Trutmann  
Department of Psychology, University of Basel
Basel, Switzerland

Thorsten Hens  
Department of Banking and Finance, University of Zurich
Zurich, Switzerland

Jörg Rieskamp  
Department of Psychology, University of Basel|Shared Senior-Authorship
Basel, Switzerland|*

Rui Mata  
Department of Psychology, University of Basel|Max Planck Institute for Human Development|Shared Senior-Authorship
Basel, Switzerland|Berlin, Germany|*

Philippe Tobler  
Department of Economics, University of Zurich|Shared Senior-Authorship
Zurich, Switzerland|*

As the proportion of households with stock ownership has increased over the past decades, more and more individual investors make their own investment decisions. Poor financial decisions can lead to financial insecurity, stress, and declining mental health, raising the question how more successful investors differ from less successful ones. As part of a pre-registered study on the foundations of successful financial decision making (Heinke et al., 2021), we used a newly developed investment task (Figure 1) to examine individual differences in neural markers of value and risk sensitivity and their correlation with investors' trading behavior and performance. Previous research has implicated the ventral striatum and the anterior insula in the representation of outcomes (rewards and punishments) and risk, respectively, in a neural network associated with value-based decision making. Therefore, we focused on the contribution of these two regions of interest (ROI) in our fMRI analyses.

154 participants with a mean age of 39 years (SD=12; 9% female) participated in the MRI arm of the study. In the task, they had to repeatedly make decisions between a risky option, which was presented to them by showing the average expected outcome magnitude and volatility, and a safe option that included no risk but also had a lower outcome magnitude. Computational models of expected value, expected utility, prospect theory, and mean-variance-skewness were fitted to the behavioral data to further explore the behavioral task data. Meta-analytically generated masks of the ventral striatum and anterior insula were used as ROI for fMRI analyses of outcome value and risk, respectively.

A simple expected value model with no distortions of outcome value or risk showed the best fit to the behavioral data at the group level. The other three models performed comparably but required more free parameters. Note that our convenience sample consisted predominantly of active retail investors, which could explain why a simple expected value model best fitted the behavioral task data. ROI analyses testing associations between outcome magnitude and BOLD signals in the ventral striatum and between risk and BOLD signals in the anterior insula revealed weak associations. Specification curve analyses (SCA) predicting real-life trading behaviors and outcomes did not show substantial associations with individually extracted, aggregated BOLD signals from either ROI. To further investigate the relationship between activity in the predefined ROIs and key real-life trading measures, we conducted exploratory analyses. By directly contrasting BOLD responses to outcome magnitude and risk, we aimed to identify neural activity that was more specifically associated with reward over risk and vice versa. Activations obtained with these contrasts were then extracted from ventral striatum and anterior insula for each contrast. We performed the same SCA to test the association of these exploratory neural indicators with portfolio risk, return, and performance. Activity in this striatum ROI not only showed a preferential relationship to reward over risk, but activity in voxels identified with this contrast was also significantly associated with portfolio return and variance (Figure 2).

Our pre-registered analyses provided little explanatory power for real-life trading behaviors and outcomes. However, additional exploratory analyses within the same ROIs showed that ventral striatal BOLD signals with preferential relation to outcome magnitude over risk are positively associated with trading outcomes in real-life financial decisions. Thus, neural outcome sensitivity in a core region of the brain valuation system explains investment success in part. This study is to our knowledge the first demonstration that BOLD responses to outcome magnitude have predictive power for real-life trading behavior.

Reward and Punishment ²

Decision Making ¹

Cognition

FUNCTIONAL MRI

Other - risk, value-based decision making