For best experience please turn on javascript and use a modern browser!
You are using a browser that is no longer supported by Microsoft. Please upgrade your browser. The site may not present itself correctly if you continue browsing.
uva.nl
Bubbles in Amsterdam

Imagine, on a beautiful morning many people expecting a sunny day, so they take their sunglasses as they go outside. This will not prevent a thunderstorm from occurring; the expectations do not affect the weather. However, if many people expect prices on a financial market to rise, demand – and thus the share price – will rise. Here, the connection between expectations and share price movement is evident. Myrna Hennequin’s layman’s talk, given during her PhD ceremony at the University of Amsterdam late May and illustrated with smileys with sunglasses, price charts and a few tables, is perfectly clear. Her thesis, Expectations and bubbles in asset market experiments, may be full of formulas and charts, but the findings likewise speak for themselves.

Individual expectations and group behaviour

How are expectations formed? And when do these expectations cause stability or instability in financial markets? In her doctoral research, Hennequin conducted three experimental studies. One of the studies investigated bubble formation in experiments with larger groups. An earlier study involved six participants. Hennequin raised this number to between 21 and 32. Participants were asked to predict the price of a share. There were no major differences in how expectations were formed compared to smaller groups and the results were similar for both group sizes, she found. In larger groups – as in small groups - individual forecast errors were not cancelled out: participants collectively aimed too high or too low. 

'If you don’t immediately know what would be best in the controlled environment of the lab, why would you be able to make the best decisions outside the lab?’

Experience with bubbles

In another study, Hennequin found that if all the participants in a complex market experienced stability before, the market becomes stable sooner. If all subjects experienced a bubbly market before, they expect more bubbles, which can lead to the formation of new bubbles. The past experience thus affects expectations, a conclusion that had already been drawn from earlier empirical research. However, the outcome that bubbles do not disappear with experience is surprising; this clashes with results of earlier experiments in less complex markets.

Fascinated by behaviour economics

During her Master’s, Hennequin wrote an essay on bubbles in experiments with financial markets. An intriguing subject, she found. When her future promotor Cars Hommes suggested the theme for her doctoral thesis, it did not take her long to agree. Hennequin opted for econometrics because she liked economics and maths in high school. At the UvA she soon became hooked on the experimental research of the UvA’s Center for Research in Experimental Economics and political Decision-making. ‘Those experiments were really about behaviour: what do people do in economic situations? The situations in these experiments seem relatively simple, but they turn out to be quite complicated when you enter them and need to make decisions. And if you don’t immediately know what would be best in the controlled environment of the lab, why would you be able to make the best decisions outside the lab?’ From then on she was fascinated by behaviour and experimental economics.

Myrna Hennequin
Myrna Hennequin

Monetary policy and the market 

Hennequin’s study on the effect of a ‘leaning against the wind’ monetary policy on asset price bubbles provides food for thought beyond scientific circles. ‘We wanted to shed a light on the possible impact of monetary policy on stock markets.’ Hennequin is unfazed by the fact that central banks are not eager to use monetary policy to target anything but inflation. ‘It is important to reflect upon policies that might limit bubbles, and also to investigate alternatives’, she says. ‘We wanted to see if the monetary policy might work, and that seems to be the case.’

Bubbles caused by trend-following expectations might be deflated by a strong interest rate hike, she found. Clear communication about the interest rate changes and the goal of the policy is crucial. ‘The results seem to indicate that expectations can be managed to some extent, but a lot of additional research is needed.’

Research results are of social significance

Although she does not expect central banks to overhaul their policy because of her findings, Hennequin feels they are of social significance. A lot remains unclear about bubbles in financial markets, about both their formation and ways to reduce them. ‘And no one wants a return of the housing market bubble and the subsequent crash, she stresses.’ With some adaptation, her findings can be extrapolated to other markets - ‘to all markets that have a positive connection between expectations and the price development, in fact.’ Besides, future research on this topic can be conducted at a lower cost now that it is clear that experiments with a limited group size are representative.

Hennequin will not pursue this herself, but she is confident there will be further research, possibly by her promotor or other colleagues. Last fall, Hommes published a paper in the ECB Working Paper Series on behavioural & experimental macroeconomics and policy analysis. He told her that people within the ECB are fascinated by the research findings.

'No one wants a return of the housing market bubble and the subsequent crash.'

How do people react in the real world

Hennequin is starting new projects. Since May, she has been doing research within UvA into the dynamics and effects of free trade and import duties in commodity markets.
But she has not let go of the theme of her thesis. Recently, an article about her first study was accepted for publication in the Journal of Economic Dynamics and Control and Hennequin continues to work on papers about the other two. She also continues to investigate the impact of group size: she is collaborating on an online experiment with markets of 500 people. 
This summer, she wants to take steps that might build a bridge between theory and the real world: she intends to further investigate the link between expectations and the actions that people actually take. ‘Currently, what’s missing in my research is the possibility that people will enter or exit the market when there is a bubble. In the real world, markets are much more complex than in the lab. It would be interesting to include what people do if they actually act based on their expectations.’