A Markov Switching Model for UK Acquisition Levels Sian Owen*

This paper examines the time series properties of UK acquisition numbers in the period 1969 to 2003 using a three-regime Markov Switching Model. The majority of the data is characterised by a relatively stable series, and this regime has by far the longest duration. It is necessary, however, to also include regimes that represent both the beginning and end of the waves to accurately model the data. The expected duration of the regime that marks the end of a wave is longer than that characterising the start, revealing that the start of a merger wave is marked by more extreme changes in behaviour. This somewhat surprising result is then confirmed with further analysis of the characteristics of the regimes marking the beginning and end of the waves