‘Study macroeconomics with realistic behavioural models'

11 March 2014

Nowadays for many of us it’s much easier to imagine the link between ‘chaos and economics’ than in 1999 when this was the title of Cars Hommes’ inaugural address. He referred to scientific chaos, but what is that precisely?

The most striking example of the scientific chaos theory happens in the atmosphere, the best-known chaotic system. It’s the 'butterfly effect'; the fluttering of a butterfly’s wings can cause a small disturbance in the atmosphere that can lead to completely different weather, maybe even a hurricane. "Chaotic, complex systems have nonlinear processes, which simply means that a small cause can lead to big consequences," says Hommes, professor of economic dynamics at UvA's Amsterdam School of Economics. 

To Hommes the economy is also a complex system that can be studied with research methods from the sciences. In January, the mathematical economist won the Distinguished Lorentz Price & Fellowship of the Royal Netherlands Academy of Arts and Sciences for innovative research that builds bridges between scientific disciplines. Hommes is also founder and director of UvA research institute CeNDEF (Center for nonlinear Dynamics of Economics and Finance). Hommes: "The financial crisis was a blessing for us from a scientific point of view. Everyone now has to admit that the standard economic models didn’t see the crisis coming. Surely all economists already knew that the rational model isn’t completely accurate. It’s just so difficult to find an acceptable alternative. It's what we call the ‘wilderness of bounded rationality’, meaning that the model needs to deal with the fact that people make mistakes. How can you possibly put that into a model?” 

Hommes advocates the use of realistic behavioural models in macroeconomics and financial markets, which are considered nonlinear systems. Hommes: "Instead of assuming rational behaviour on the part of an average agent such as a consumer, producer or trader, we take into account the group behaviour of many individuals. These individuals are by definition limited in their rationality since they can never all have all the relevant information. In the laboratory of CREED (University of Amsterdam's Center for Research in Experimental Economics and political Decision making) we did experiments designed to test and model the decision making process of groups of individuals." 

During these experiments the participants had to predict prices. Hommes says the experiments showed clearly that financial markets never reach a price equilibrium as a result of their positive feedback system. Positive feedback means positive expectations build up the realised market price and with negative expectations the price drops. In the experiment with a negative feedback system (as seen in the agricultural market where high expectations lead to increased production and lower prices) a realistic price was reached in three steps by matching supply and demand. In the case of positive feedback, the fluctuations kept increasing while the average price remained well above the  rational equilibrium price. Hommes: "It shows that financial markets are doomed to form bubbles and crashes. And the financial crisis made quite clear that these fluctuations are so intense that some form of regulation is required. But what should be done? Regulations can do more harm than good so first there should be thorough research based on this new economic approach.” 

One of the supporters of the new way of economic thinking is George Soros. This American super-investor of Hungarian descent wrote his first book in 1987, about reflexivity: how participants in the economy influence what happens in the future with their expectations. It is exactly what the positive feedback system experiment shows. If people expect the prices to rise, this will actually happen and vice versa. Soros is one of the sponsors of INET, the Institute of New Economic Thinking that Hommes is actively involved in, and from which he receives several research grants. 

In an article in the special January edition of the  Journal of Economic Methodology dedicated to reflexivity, Hommes explains the laboratory experiments as well as the theory that is based on the experiments. In April Hommes will give a speech at the INET conference in Toronto. "Soros could only describe his ideas based on real life experience. With our behavioural model we can now explain how his intuitive ideas about reflexivity actually work. People have a number of decision strategies. They quickly switch to strategies that work and repeat rules that proved to be successful. It’s an evolutionary process. In a positive feedback system trend-following behaviour proves to be successful. That’s why expectations and the subsequent herd behaviour have such a large impact. If you deny that fact in your policy models, you will draw all the wrong conclusions," says Hommes.

Amsterdam School of Economics