Although infrequent, extreme events can lead to significant financial losses, as the collapse of financial markets following the pandemic emergence in March 2020 has recently shown.
How do investors react to such dramatic events? To answer this question, we developed an experiment that allowed us to better understand the reaction of investors to extreme negative events.
In particular, we show that the market exacerbates risk-taking in times of extreme losses, which can be explained by the fact that the market amplifies the anger of investors facing large losses.
Taking risks and emotions
In an extreme event experiment aimed at studying individual investment decisions, investors tended to lower their bids after observing extreme events without incurring losses, as this increased their expectation of observing such an event again.
On the other hand, investors who have suffered extreme losses have increased their offers, which is explained by the increase in risk-taking in the area of losses. In fact, according to the prospect theory (prospect theory), investors who suffer extreme losses seek risk because they suddenly enter the realm of losses.
Furthermore, reactions to extreme events are linked to specific negative emotional reactions, such as anger and fear. The role of emotions in the study of extreme events is crucial because, by definition, their occurrence is rare, resulting in surprise and a strong emotional reaction.
Extreme risk is an essential feature of financial markets that has been used, for example, to explain shareholders’ demand for a high return on investment relative to bond yields.
In our recent experimental study, we investigated extreme events in a market where investor decisions are interdependent. We therefore sought to assess whether investor reactions to extreme events and large losses are attenuated or amplified in a market context versus an individual decision-making situation.
213 participants in this experiment were assigned the task of making 300 successive offers to acquire a financial asset that offered a small positive reward (i.e. 10, 20, 30, 40 or 50 cents) in more than 99% of cases and a big relative loss (1000 cents) otherwise. The loss was both highly unlikely (0.67%) and large enough to wipe out the accumulated gains of participants and bankrupt them.
Participants were assigned to three different treatments: (i) pure individual decision (without interaction with other participants), (ii) individual decision with period-by-period information about the other participants’ price choices, and (iii) decision at all. internal market (also including the same information referred to in point (ii)).
With respect to individual decision making, the market is characterized by a social context in which investors observe each other’s auction information and compete with each other. By comparing individual decision (i) and information processing (ii), we can isolate the information dimension. By comparing treatments (ii) and market (iii), we can isolate the competitive dimension of the market.
Because the treatments were designed to be as similar as possible, the study shows that the offerings do not differ significantly between treatments at normal times. On the other hand, in the event that extreme events occur, our results confirm previous studies: Investors who observe but do not experience an extreme event run less risk, while investors who experience an extreme event run more risk.
Emotions play a key role in explaining investor reactions to extreme events. These emotions are influenced by the interdependence of investor decisions, which is inherent in the treatment of the market.
We found that the information dimension of the market did not exacerbate the emotions and therefore the reactions of investors to extreme events. On the other hand, we show that competition exacerbates the anger of investors who suffer extreme losses.
In fact, we found that extreme losses trigger a stronger anger response in the market treatment than the other treatments. As anger increases risk taking, we see a more pronounced increase in offers after extreme losses in the market treatment than the other two treatments.
This result is consistent with the theory of the evaluation of emotions (Theory of the evaluation of emotions), that an emotion tends to anger if another person can be held partly responsible for the negative outcome an individual faces. In a market where, unlike individual investment choices, decisions are interdependent, it is easy to attribute losses to the behavior of other investors.
While markets did not encourage risk-taking in normal times, they exacerbated risk-taking when investors suffered significant losses. Our results are therefore consistent with the idea that markets promote emotional contagion in times of crisis. The emotional contagion is probably even stronger if we consider a richer social context, characterized for example by the use of chat platforms in which participants share opinions and feelings about the market.
The phenomenon of emotional contagion in the markets could therefore justify the use of “automatic switches” that allow investors to vent their anger when market values fall sharply.
Professor, EM Lyon.
08 October 2021, 7:32