Evaluating market supervision through an overview of trading halts in the portuguese stock market

• Autor(a) principal: Duque, João
• Data de Publicação: 2003
• Outros Autores: Fazenda, Ana Rita
• Tipo de documento: Artigo
• Idioma: eng
• Título da fonte: Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos)
• Texto Completo: http://hdl.handle.net/10400.5/23636
• topic: Market Supervision; Capital Markets; Trading Halts; Stock Price Returns; Volatility; GARCH Models

In this study we analyse security-trading halts in the Portuguese capital market to measure the goodness of trading halts imposed by market authorities as well as its timing ability on both interrupting and restarting trading. A sample of 54 equities whose trading has been halted from some hours to some days between 1992 and 1999 was collected. We used stock price returns, abnormal returns and volatility to compare and test the significance of differences for pre and post-halt periods. Firstly we consider the global sample to analyse stock abnormal returns and afterwards we split it into good and bad news halts. Concerning volatility, we applied t-test as a GARCH (1,1) model to complete volatility analysis. The GARCH(1,1) model has two additive dummy variables on the independent term of the mean and variance equations. We found that justification for trading halts tend to increase as event window size increases, suggesting that supervisory authorities tend to spot better the dominant changes (mainly for the latest years considered in this research). In fact, when very short time sampling periods are used we found weaker justifications for stock halting. We also found a GARCH (1,1) model useful and a more sensitive instrument on justifying trading halts. The opportunity of market authorities to interrupt trading seems to be increasing, but in terms of timing, we found that they seem to be most of the times retarded when imposing trading halts or anticipated when authorizing restart trading. Nevertheless when considering "good" news, although the halt tends to be late, the restart seems to be on time. Additionally we found that the methods under scope in this study are not redundant and should be jointly used by stock watch departments of supervision authorities for detecting trading under asymmetric information.