Herding in Financial Markets
Background:
Over the last ten years rational herding has become an important tool in analysing how and why economic agents learn through observation in groups, beginning with seminal contributions by Banerjee (1992) and Bikchandani, Hirshleifer and Welch (1992). Economic agents constantly learn from others, through chat, newspapers, and typically in financial markets, also through observing price movements, or buy and sell decisions by others. Financial decisions then seem an obvious candidate for observational learning.
However, a central lesson of herding theory is that what we can observe is not necessarily a direct indication of the information possessed by others. For example, if I observe a sequence of customers entering a restaurant, while I can be reasonably sure that the first one or two have information favouring the restaurant, the others might be herding after the first two, and so provide little new information. So a large herd on one action may be based on the actions of only a few early decision-makers.
Although it seems reasonable to suppose this problem could be endemic to finance, early negativity in the literature (Avery and Zemsky, 1998) stalled the development of financial herding for some years. However, recent work (Chari and Kehoe, 2004; Park and Sabourian, 2004) suggests that herding might be a prime explanation for persistent price spikes and crashes, and often cited "crazy" behaviour on the markets. If we could describe such movements using the tools of herding theory, not only would we better understand financial markets, but we could also develop policy suggestions to try to avoid painful surges and crashes.
We propose to develop a theoretical model which can explain financial crises in terms of rational herding behaviour, as well as potentially incorporating a series of realistic elements such as the scope for delay in decision-making, multiple types of trader, reputations, strategic behaviour and contagion across different financial systems.
Journals, seminar papers and other publications
Park A & Sabourian H, “Herding in Models of Sequential Trade with Monotonic Signals”
Researchers:
Dr Hamid Sabourian
Dr D Sgroi
University of Cambridge
Professor D Gale
New York University
Professor A Park
Toronto University
Contact:
Dr Hamid Sabourian
Faculty of Economics
University of Cambridge
Austin Robinson Building
Sidgwick Avenue
Cambridge
CB3 9DD
Tel: 01223 335200
email: hamid.sabourian@econ.cam.ac.uk
Duration of Research:
January 2005 - December 2006