Regularities in Australian Share Returns

Abstract

The inverse relation between total equity value and return is confirmed for Australian ordinary shares over the 1974–1984 period. We find this so-called size effect to be robust with respect to several technical methodological adjustments. In particular, the pattern persists when we account for: fluctuations in the returns of nominally riskless assets; varying levels of systematic risk of size ranked portfolios; the index selected to represent the common factor in the return generating mechanism; the form of returns, i.e., simple versus continuously compounded; possible initial mispricing of new issues; possible seasoning of new issues; shares that leave the data base for bankruptcy, winding up, or other reasons; the third moment of the return distribution; and the levels of portfolios' unsystematic risks. We do find several economic variables that seem in part to account for the magnitude of the size effect, viz., accounting for: yearly rather than monthly portfolio revision; nonsynchronous trading; and varying transactions charges, volume of information available, and liquidity between large and small firms.