Dependence in London Stock Exchange Returns Due to British General Election Results Announcement
Wilson, Alexander (2006) Dependence in London Stock Exchange Returns Due to British General Election Results Announcement. [Dissertation (University of Nottingham only)] (Unpublished)
The research study attempted to examine dependence in the London Stock Exchange due to the "shock" of a British General Election result announcement. This was done in two parts. The first tested for possible Long Memory in the returns by using an ARFIMA (p,d,q) modle to examine for fractional integration and then the ARFIMA (p,d,q)-GARCH(p,q) model to include conditional volatility. The second part of the study examined possible short term dependence in the returns, specifically looking at the market movements on the day of announcement. This was done using the Riley and Luksetich (1980) regression method.The finding for the first section suggest that fractional integration is evident in the election years 1987 and 1997 using the ARFIMA ( p,d,q ) model. However, the following ARFIMA( p,d,q )-GARCH( p,q ) test suggested that this was due to conditional volatility in the returns. The findings of the second test suggest short term dependence in the returns due to the General Election result announcement and show the market to significantly react differently to a Conservative party victory as it does to a Labour Party victory, preferring the former. The results also showed a significant difference between the returns when an incumbent part wins or loses.
Actions (Archive Staff Only)