While waves of takeover activity have been documented in the economics literature for many years, the determinants of aggregate takeover activity are not well understood, and have been subject to only a modest amount of research. It is the aim of this study to investigate the population of takeover bids for listed companies in Australia for the period January 1972 to June 1990, and relate aggregate takeover activity to several macroeconomic and capital market variables hypothesized to influence a firm's decision to seek to expand by external investment.
The economic condition variables investigated are the Melbourne Institute Index of Industrial Production (excluding power), private fixed capital expenditure, and new capital raisings by listed companies. The capital market variables are the Statex Actuaries Accumulation Index and ten year government bond yeilds. Quarterly data for all variables were related to takeover activity, providing 74 quarterly observations from 1972 to June 1990. Contemporaneous, leading and lag relationships were examined for all variables.
Because of the high degree of autocorrelation and nonstationarity in the variances of each of the time series variables, an ARIMA model was used to transform the percentage change for each series into the component predictable from its own past, and the uncorrelated residual component. The residuals for each series were then correlated against the residuals for each other series, including leading and lag relationships of up to four quarters.
From the univarate cross correlation analysis, stock prices and industrial production were the most highly correlated variables with takeover activity, and were included in multivariate models of aggregate takeover activity. From a number of possible regression models, the only variable which was consistently significant was the stock price index. The Granger Causality Test indicated a contemporaneous relationship between stock price returns and changes in successful takeover activity from which no directional causality could be deduced.
Aggregate takeover activity in Australia appears to be largely a stock market conditions phenomenon. However, in interpreting these results, it must be kept in mind that the empirical analysis suffers from a small sample problem in that the number of takeover waves since 1972 is small.