Use this URL to cite or link to this record in EThOS:
Title: Simulation analysis of learning and expectations in the stock exchange : a case study with the Warsaw Stock Exchange (WSE)
Author: Yau, Louis
Awarding Body: University of Leicester
Current Institution: University of Leicester
Date of Award: 1996
Availability of Full Text:
Access from EThOS:
Access from Institution:
Increasingly, it has become difficult to explain economic phenomena within the neo-classical framework in a period of changes when learning about the changes precedes any costly adaptation. The process of learning has been argued to be a missing element. It is defined as the continuous inference from observable data the unobservable state and structure of the market that are typically unknown. Learning is more detectable during rapid economic changes and when the gain- loss differential is critically enormous, like in the financial markets. Hence, an interactive learning model is formulated to study how learning and the interaction between market and traders can affect price. In particular, the noise trader approach which accounts for the excess volatility and the mean reversion phenomena in share prices, is used as a theoretical framework that allows learning to happen. A case study is done to four chosen shares in the Warsaw Stock Exchange (WSE), a newly-emerged market in a transitory economy. Analysis is done by means of simulation and detailed comparison between empirical and simulated data. The objectives are: (i) to understand the effect of learning and the interaction between market and agents and (ii) to search for the underlying conditions in the chosen markets so as to have a better understanding of them. The results suggest that the four chosen markets in WSE are not efficient. Self-fulfilling inefficient market beliefs of agents who are learning the state of the market with dynamically misspecified models may be a cause. Learning also leads to excess volatility and mean reversion in share prices. Moreover, free participation in the market can produce seemingly deceptive regression results against the objective process if agents are capable of influencing market realisation. This is best reflected in one of our controlled extreme cases with no learning in the market.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available