Leader: Thorsten Hens
The researchers in the Zürich node of the network (Marc Chesney,Thorsten Hens, Felix Kübler,Jean-Charles Rochet, Karl Schmedders) are from three different areas: General equilibrium (Hens, Kübler and Schmedders), Mathematical Finance (Chesney) and Contract Theory (Rochet).
Past research of this group has covered many topics from different perspectives. Hens has gained some experience on expectational coordination while he studied the relation between sunpot equilibria and the multiplicity of equilibria. When he started this research it was well known that a coordination of expectations is needed when the underlying economy has multiple equilibria. However, Hens (2000), Hens and Pilgrim (2004) and Hens, Mayer and Pilgrim (2004) showed that even when the initial underlying economy has a unique equilibrium then trade in incomplete asset markets might lead to an ex post wealth distribution that allows for multiple equilibria so that the coordination problem among multiple equilibria was generated endogenously by asset trade. In his papers he also shows that the conditions on the underlying economy that allow for this surprising phenomenon are closely related to those needed for the transfer paradox known from international trade literature.
Kubler and Schmedders are the leading experts in computational economics and finance. Concerning the problem of expectational coordination, as early as 1999, Schmedders had already studied conditions for multiple equilibria with Hens and Voss in Hens, Schmedders and Voss (1999). That paper shows that assuming Cobb-Douglas utilities is not sufficient to rule out multiple equilibria when markets are incomplete. More recently Kubler and Schmedders (2011a) and Kubler and Schmedders (2011b) have developed new methods to track multiple equilibria, also for steady state equilibria in dynamic models.
Chesney is a leading expert in mathematical finance who recently studied the impact of exogenous events like terrorism on financial markets (Chesney, Reshetar and Karaman (2011)). Also, he is the leading expert in environmental finance. Rochet is one of the world best researchers in contract theory as applied to banking. His studies of the incentives in the financial crisis will help us to find a micro-foundation of the general equilibrium view.
In this research project our group will join forces to focus on the following research question:
Soros` Principle of Reflexivity and Speculative Bubbles in Production Economies
With hindsight, the history of speculative bubbles in financial markets seems to be good evidence that from time to time financial markets lead to a huge misallocation of resources. One is wondering for example how people in the Netherlands in 1636 were ready to pay more than the value of a house for a single rare tulip onion or why, more recently, Spain has doubled its hotel capacity from 2000 to 2008. On the other hand some of those speculative bubbles achieved the break through of important innovations like the rail road, the mass production, globalization, and information technologies. See Schleifer (2000, Chapter 6) for an overview of speculative bubbles from 1630 to 1999.
The aim of this subproject is to develop models that help sorting out the pros and cons of speculative bubbles so that policy makers get some help in acting on them. Unfortunately, speculative bubbles are an under-researched area of financial economics: Tirole`s (1985) analysis of rational bubbles in discrete time over lapping generations models is a useful benchmark case for us which however does not simultaneously explain the build up and the crash of a bubble. More recently, Abreu and Brunnermeier (2003) show that with continuous time exiting a bubble imposes too high synchronisation risk on the side of rational investors so that bubbles can last longer than one would think from a purely rational perspective. This is an important aspect of expectational coordination but it seems to hinge on the continuous time framework. On the other hand, speculative bubbles are known to arise in models with quite some ad-hoc assumptions, e.g. in the model of Brock and Hommes (1998) or Lux (1995). We think that these are seminal contributions but there is still a huge gap to be filled between these extremal points.
Despite the work of Brock and Hommes (1998), Lux (1995), Shleifer (2000) and Shiller (2000), the leading paradigm of financial economics has still worked with the rational expectations hypothesis, according to which economic fluctuations are not created from within the economic system but appear as exogenous surprises – just like the “black swans” described in the celebrated book by Taleb (2007). This deficiency of the leading paradigm of economics is often used by policy makers for excusing themselves not having acted on a speculative bubble. E.g. Alan Greenspan was always very sceptical to act against asset market bubbles. He said “How do we know when irrational exuberance has unduly escalated asset values?”. We hope that our project creates such knowledge.
The purpose of this subproject is to develop models of speculative bubbles balancing the pros and cons of speculation. At the core of the problem we see the interaction between speculation, investment and innovation. During a speculative bubble some firms get cheap credit which they can invest in new technologies creating new innovations. These innovations have positive externalities for other firms. Thus the question we need to analyze is how to trade off the inefficiencies due to cheap credit and the positive externalities from innovations.
Since we ultimately intend to do a normative analysis giving advice to policy makers it is mandatory to use a general equilibrium perspective. Many aspects of speculation already arise in exchange economies, but as we see it, the core of the problem can only be analysed in production economies. Because in those economies Soro`s principle of reflexivity is most severe:
“Financial markets attempt to predict a future that is contingent on the decisions people make in the present. Instead of just passively reflecting reality, financial markets are actively creating the reality that they, in turn, reflect. There is a two way connection between present decisions and the future events, which I call reflexivity.” (Soros (1998), page xxiii)
To be successful with our research project we divide it into the following milestones
We need to develop a general equilibrium, model with production and
- credit constraints.
- credit constraints and speculation.
- credit constraints, speculation and externalities.
In a sense we intend to build a general equilibrium version of the famous Kiyotaki and Moore (1997) model of credit cycles. For the first milestone we can build on the general equilibrium literature modelling collateral, e.g. Dubey, Geanakoplos, and Shubik (2005) and Geanakoplos, Zame (2010). These models need however be extended to production economies. We are aware of the problems incorporating production in a general equilibrium model when markets are incomplete, as pointed out by Dreze (1974) and many others thereafter (see Magill and Quinzii (1996), Chapter 6 for a systematic account). In order not to overload, in a first attempt the model we are willing to assume that markets are complete. For the second milestone we need to blend in second order beliefs that give rise to speculative bubbles, as suggested by Keynes (1936) a long time ago and used recently in e.g. Bacchetta and Van Wincoop (2004) who build on the framework of Morris and Shin (2002). Finally, we want to extend the model towards production externalities, e.g. in the sense of Allen (1982).