Stock Market Participation: When No Priming Works Best

Abstract

Typically, individuals tend to under-invest in financial markets. Meanwhile, financial market participation always takes place with some form of nudging, which can particularly be introduced as priming of potential clients. Through a survey of 2107 individuals from 5 European countries, we show that, in line with the somatic marker hypothesis, priming significantly reduces participation in the stock market for individuals who fail to be emotionally moved while significantly increasing stock market participation for those who feel strong emotions due to this priming. Moreover, emotions and financial literacy interact so that less financially literate investors typically participate more in financial markets when they are impacted by emotions compared to highly financially literate investors. In terms of managerial implication, the results testify to the need for targeted priming of individuals.

Keywords:

• Behavioral Finance
• Emotions
• Nudge
• Priming
• Stock-Market Participation

JEL CLASSIFICATION:

• G41
• D14

Acknowledgements

The authors thank the Savoie Mont Blanc University, the Odonatech company and Linksium, the Grenoble Alpes SATT, for funding support.

Ethical approval

All experiments conducted in this research were in accordance with the ethical standards of the 1964 Declaration of Helsinki.

Informed consent

All participants provided informed consent prior to completing the experiment.

Authors’ contributions

All authors equally contributed to the article: conceptualization, conception and design of the experiment, data collection and software analysis, writing – review & editing.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Figure 7. Attention Check Used in the Study.

Figure 8. Measure of Feelings.

Figure 9. Priming - Short Term Results (ST).

Figure 10. Priming - Long Term Results (LT).

Figure 11. Priming - Material Expectations (Mat).

Figure 12. Priming - Negative Expectations (Neg).

Figure 13. Priming - Regret (Reg).

Figure 14. Priming - Lifetime Goals (Goal).

Figure 15. Measurement of the Dependant Variable - Common to Control Groups and All Treatment Groups.

Notes

1 https://www.pwc.com/us/en/industries/financial-services/library/retirement-in-america.html.

2 Note that this initial experiment of Bechara et al. (1997) did not use any priming. The authors initially used the Iowa Gambling Task to prove their point as it relates to the somatic marker hypothesis. The Iowa Gambling Task was thus just a means to an end, and the somatic marker hypothesis has implications well beyond the Iowa Gambling Task, including for priming.

3 The somatic marker hypothesis does not posit that emotions are always beneficial to decision making, as clearly explained later on in Bechara and Damasio (2005). In particular, if the somatic state is unrelated to the task at hand, it can lead to disadvantageous decision-making.

4 Note that long-term regret (LT Reg) is simply the merger of the treatment long-term results (LT) and of the treatment regret (Reg).

5 The scale of answers for the literacy test comprised 7 points in total with 0 points for the incorrect answer and 1 point for each correct answer in 6 questions. The share of participants who did not provide any correct responses accounted for 1% of the sample.

6 To retain only attentive participants, we used an attention question in the middle of the survey and screened out participants who failed to answer properly. About 36% of initial participants were either screened out or left the survey without answering fully and are thus not included in the analysis.

7 The other half was faced with a different task. We split the sample to cover all the various research questions the survey was aiming at while keeping total survey time reasonable.

8 Note that this payment is below what the respondents would have made had they simply worked in their respective countries. We recognize that some could consider it questionable from the perspective of research ethics. However, research has shown that money is often not the main driver of the decision to participate in a research survey. As recently shown Desmoulins-Lebeault et al. (2022) or nicely summarized in Kaufmann et al. (2011), other factors, including intrinsic motivation or interest in the study, play a large role in the decision to spend time answering a study.

9 Please see descriptive statistics on all variables in “Descriptive statistics”and regression analysis in “Impact of priming treatments”.

10 Please, see “Robustness analysis for the impact of emotions of different categories”.

11 According to the survey of the European Central Bank (“Household sector report 2022 Q4”, 2023), the mean level of household financial investment in the euro area in 2022 reaches 2.5% in transactions as a percentage of outstanding financial assets. We recognize that in our survey the market participation share reported by participants is considerably higher than the official statistics figures. This difference appears to be one of the limitations of the current work and hypothetical experimental surveys in general.

12 Calculated effect size is in our case equal to d = 0.079.

13 Please note that participants from the baseline treatment were not questioned about their emotional state. In line with the dual-process theories that mark out the key role of emotions underlying the priming effect in the decision-making process, we aimed to analyze the impact of emotions induced by various priming treatments. Consequently, our further analysis highlights the effect of emotions only in the context of priming.

14 The division in the three categories was performed to ensure a balanced distribution of responses across these groups.

15 This finding is in line with several works stating lower risk-aversion in men compared to women, and proving higher motivation for profit-making and financial investments in men population (e.g., Barber and Odean 2001; Bernasek and Shwiff 2001; Felton et al. 2003; Martenson 2008).

16 We did not observe any quadratic effect of age. The lower market participation of older participants compared to the young population is coherent with the findings of Korniotis and Kumar (2011), marking out the adverse effects of age on investment and deteriorating investment skills despite the increased expertise.

17 This result is in contrast with several studies. Thus, the high impact of financial literacy on stock market participation is highlighted by Van Rooij et al. (2011) among others, showing that low-literate people invest less in risky financial products like stocks compared to sophisticated participants. As well, Lusardi and Mitchell (2011b) documents a strong correlation between the level of financial literacy and retirement planning. We are going to see later on that financial literacy actually interacts with emotions, which might explain this finding.

18 This is coherent with the idea of stocks as lotteries of Barberis and Huang (2008), who underline that investors who overweight small probabilities are expected to take undiversified positions in positively skewed assets.

19 Country-level differences might be explained by various factors, such as cultural norms or the attractiveness of particular stock markets (Thomas and Spataro 2018). As well, participants might have different levels of risk tolerance and risk perception in financial decisions.

Additional information

Funding

This work was supported by the Savoie Mont Blanc University, the Odonatech company and Linksium, the Grenoble Alpes SATT and Université Savoie Mont Blanc.