'Complex adaptive economic systems in the laboratory' with Cars Hommes
This event is organised by the Institute for New Economic Thinking at the Oxford Martin School
A distinguishing feature between complex natural and complex social systems is the role of adaptive behavior and learning. We survey laboratory experiments with groups of human subjects to test theories of expectations and learning. Subjects must repeatedly forecast a market price, whose realization is an aggregation of individual expectations. Emphasis is given to how individual forecasting rules interact at the micro-level and which structure they co-create at the aggregate macro-level. Does learning converge to equilibrium? It turns out that the type of expectations feedback is crucial for the (in)stability of experimental markets. Negative feedback systems, such as supply driven commodity markets, are rather stable and quickly settle down to the rational equilibrium. Positive feedback systems, such as speculative asset markets, are rather unstable and fluctuate wildly around the fundamental equilibrium with repeated bubbles and crashes. We also discuss a behavioural heuristics switching model, where agents switch between strategies based upon their relative performance, that matches individual as well as aggregate behaviour of the lab experiments quite nicely. Finally, we discuss some recent experiments with larger group sizes.
Speaker: Professor Cars Hommes, University of Amsterdam
Cars Hommes (CH) is one of the pioneers in complexity economics. His work on nonlinear complex economic systems challenges the traditional neoclassical paradigm of the representative rational agent in economics. In a rational world the economy is characterized by an average agent (consumer, producer, investor, etc.), who is a perfect optimizer with rational expectations about the future. CH’s work develops an alternative complexity paradigm based on agent-based behavioral complexity models. His PhD thesis in 1991 is an early pioneering contribution to complexity economics and his recent book in 2013 provides a comprehensive survey of the state of the art of the field. The interactions of boundedly rational micro decision rules create the emergent macro behavior that may cause sudden critical transitions and crises, similar to what has been observed in the financial economic crisis since 2007. His Center for Nonlinear Dynamics in Economics and Finance (CeNDEF) is a multi-disciplinary research center in complexity economics. CeNDEF not only has developed an alternative theory of complex economic systems, but has tested this alternative paradigm successfully in laboratory experiments with human subjects and estimated complexity models using financial and macroeconomic data sets.
In an influential article Brock and Hommes (Econometrica 1997) developed an alternative heterogeneous expectations hypothesis, based on evolutionary selection among different forecasting strategies, as a new paradigm for a bounded rationality research program in economics. The class of heuristics is disciplined by evolutionary selection and reinforcement learning, with agents gradually switching to strategies that performed better in the recent past. In a second widely cited contribution Brock and Hommes (J.Econ.Dyn.&Contr.,1998) applied the heterogeneous expectations hypothesis to a standard financial market model and explained temporary bubbles and sudden market crashes and crises in this new framework. Through a NWO Pionier grant in 1998 CH founded the Center for Nonlinear Dynamics in Economics and Finance (CeNDEF) to further develop this new paradigm and test it empirically in laboratory experiments with human subjects and with economic and financial data. A comprehensive survey of this research program is given in CH’s recent book at Cambridge University Press. CH has published more than 100 articles in leading international journals and book chapters and he is the author and editor of four books. He has given more than 200 lectures all around the world, including many invited key notes, seminars and contributions to workshops, conferences, popular science lectures and policy debates. CH has been Editor of the Journal of Economic Dynamics and Control 2002-2012 and is member of the Editorial Boards of six international journals. He is the president elect of the international Society of Computational Economics. CH has been advisor of more than 20 PhD students and 20 postdocs, most of them now holding international academic and research positions throughout the world.
Many economists nowadays agree about the limitations and shortcomings of the rational paradigm. But what is the alternative? In the view of mainstream economists, abandoning the rationality paradigm leads to a `jungle of irrationality’ and a million degrees of freedom to model bounded rationality. CH and his CeNDEF group have developed a compelling and parsimonious alternative that fits laboratory experiments and empirical financial and economic times series. Economics is close to a tipping point and the transition of a new behavioral complexity paradigm into mainstream economics is near. The financial-economic crisis has stressed the urgent need of a new paradigm for society and for policy makers, as echoed by then European Central Bank (ECB) Governor Jean-Claude Trichet in November 2010: ``As a policy-maker during the crisis, I found the available models of limited help. In fact, I would go further: in the face of the crisis, we felt abandoned by conventional tools’’. It is now time to use the fundamental and empirically relevant insights of the new complexity approach to develop more realistic financial-economic models to assist policy makers in more effective management and prevention of socio-economic crises.