The causal relationships in trading volume among stock options

Abstract

This paper examines the trading volume of call and put options listed on the CBOC for IBM. The objective is to study the flow of information among the various options by adopting a Granger causality model to investigate the lead-lag relationships in trading volume. Each class of options is first Sorted into thirty-nine groups that differ from one another in terms of either moneyness ( 13 categories) and/or time-to-maturity (3 categories). Cluster analysis is then used to identify the presence of any homogenous groups and reduce the number of moneyness groups in each of the three maturity categories from 13 to 5. The Granger causality results reveal in general the existence of causality with feedback, and this occurs more often for those groups of options with close level of moneyness in each maturity category and for those groups of options with further apart expiration elate in each moneyness category. The strong presence of two-way causal directions suggests that information asymmetry does not exist among the various options. Besides, the results indicate that the informed traders arc more likely lo use a spread strategy that either combines options with different expiration dates or options with different strike prices rather than the use of a single option to exploit their information.