The stock market is an emotional affair. A simple look at some footage from the earlier days of the stock market - when trading still took place in person and not mostly online - seems to prove that point: traders’ faces display the whole spectrum of human emotion, ranging from fear and anxiety when the market starts a downward spiral, to hope and joy after a long rally. It is easy to imagine that the degree to which market developments trigger an emotional reaction largely depends on how invested an individual is in the stock market. In turn, how much investors invest likely depends on their previous emotional experience with the market – if they have experienced mostly frustrating and disappointing emotions, they are likely to withdraw from the market, while they are likely to invest heavier after experiencing positive emotions like hope and joy.

What seems obvious from an intuitive perspective has been largely ignored for a long time by finance research. Only recently have researchers turned their eye on investigating the role of emotions in financial decision making. While a lot of the literature in the field of behavioral finance deals predominantly with behavioral biases and heuristics in decision making, emotions and their consequences for behavior have been included mostly implicitly (Taffler 2017). This is in part due to the fact that emotions are difficult to observe or manipulate in empirical data. Recently, experimental research has emerged that tackles the role of emotions more explicitly.

The vast majority of experiments investigating the role of emotions in financial decisions focuses on the effect that emotional states have on decisions. To study this effect, emotional states are induced by external stimuli like electro shocks, horror movies, or shocking images (Cohn et al. 2015, Guiso et al. 2018). Afterwards, participants make financial decisions (e.g., investment decisions or return forecasts) while still in the emotional state triggered by the cue. Those experiments consistently show that investors invest much less after experiencing a negative emotion like fear, suggesting that emotions can be one ingredient explaining the phenomenon of countercyclical risk aversion, i.e., higher risk aversion in times of financial crises. However, by externally triggering emotions and only investigating the effect it has on investments, this type of experiment can only provide one aspect of the dynamic interplay between returns, investments, and emotions.

A key reason why few experiments have tried to approach this question more holistically is that an emotional state is fairly easy to induce, but difficult to measure. However, technological advances have made measuring techniques that are already established in other fields dealing with human emotion, like medicine or psychology, available for economic research as well. For example, by scanning brain activity using fMRI while participants make investment decisions, Frydman and Camerer (2016) show that investing (and also not investing) can cause the emotion of regret. Breaban and Noussair (2018) utilize masks that measure and record facial activity to show that trading in a market during a price bubble causes fear, and that bubbles are amplified if enough market participants experience fear.

In a recent paper (Cordes et al. 2020), we have approached the question how returns, investments and emotions interact by using the fact that negative emotions in particular trigger a strong response in the autonomous nervous system. We measure skin conductance (the sweat rate at the fingertips), heart rate variation, and pupil dilation as proxies for activity in the autonomous nervous system while participants decide how much to invest. We track our measure of emotional arousal and the investment behavior over several periods to be able to analyze their dynamic interplay. Our key findings can be summarized as follows: strong market developments, in particular market busts, cause a strong emotional reaction. Positive short-term returns reduce emotional arousal (the participants relax), while negative returns increase arousal (participants become more anxious).  High emotional arousal triggers lower investments, so when participants are afraid, they invest less. In turn, higher investments are followed by states of increased emotional arousal, because higher risk exposure increases excitement and anxiety.

What do these results mean? First, financial models trying to describe (and potentially predict) behavior should incorporate this dynamic back-and-forth between emotions, investment, and returns in order to paint a complete picture. An investment decision is not a static decision taken by non-emotional robots. Second, people planning to invest should be aware that emotions affect their behavior, and that the very process of investing may put them in a different emotional state, with potential consequences for their investment decision. And finally, the results may help us understand how and why the stock market is appealing to some and appalling to others – because it is, after all, a highly emotional affair.