Abstract
This dissertation presents an examination of the trading behaviour of active Australian fund
managers. The thesis begins with an analysis of how fund manager trades relate to stock
returns in the past, the present, and the future. The dissertation next proceeds to investigating
how fund size affects fund performance, trading and portfolio construction. Finally, using
earnings announcements as the locus for trading sequences, we analyse the nature of the
information used by fund managers to predict stock returns.
This research is presented in the form of three essays. The first essay investigates how active
fund manager trades relate to stock returns. Using a unique database of daily transactions
from Australian equity managers, we document that our sample of institutional investors
exhibit statistically and economically significant predictive power in forecasting future stock
returns over the ten days following their trades. Furthermore, detailed analysis indicates that
manager style is important in understanding the link between institutional trading and stock
returns. The essay finds growth-oriented managers are momentum traders, while style-neutral
and value managers are contrarian. Further, the contemporaneous relation between
institutional trading and returns depends on trade size, broker use, and investment style.
Finally, the study documents that trades and returns are inversely related for value/contrarian
managers and directly related for style-neutral and growth managers.
The second essay presents an analysis of how fund size affects investment performance.
Recent studies find evidence that small funds outperform large funds. This fund size effect is
commonly hypothesized to be caused by transaction costs. Due to the lack of transactions data,
prior studies have investigated the transaction costs theory only indirectly. This study
however, analyses the daily transactions of active Australian equity managers and finds
aggregate market impact costs incurred by large managers are significantly greater than that
Finally, the third essay examines the nature of price-sensitive earnings information used by
package formation and portfolio characteristics consistent with transaction cost intimidation.
An analysis of the interaction between transaction cost intimidation and the fund size effect
documents that large managers pursue a highly active trading strategy, and accordingly suffer
more from the fund size effect than is the case for large funds following a less active trading
strategy. This suggests the fund size effect is related to transaction costs as trading activity is
a good proxy for expected market impact.
Finally, the third essay examines the nature of price-sensitive earnings information used by
fund managers to trade. While a number of recent mutual fund performance studies find
data, prior studies have investigated the transaction costs theory only indirectly. This study
however, analyses the daily transactions of active Australian equity managers and finds
aggregate market impact costs incurred by large managers are significantly greater than that
Finally, the third essay examines the nature of price-sensitive earnings information used by
package formation and portfolio characteristics consistent with transaction cost intimidation.
An analysis of the interaction between transaction cost intimidation and the fund size effect
documents that large managers pursue a highly active trading strategy, and accordingly suffer
more from the fund size effect than is the case for large funds following a less active trading
strategy. This suggests the fund size effect is related to transaction costs as trading activity is
incurred by small managers. Furthermore, large managers exhibit preferences for trade
evidence of outperformance relative to suitably constructed benchmarks, limited research
exists as to whether such outperformance is due to privately collected information, or merely
expedient interpretation of publicly released information. In this essay an examination of the
trade sequences of fund managers around earnings announcements is performed, and evidence
is presented revealing an increased Occurrence of buy-sell trade sequences around good
announcements and vice versa for bad announcements. The results also show an increase in
the frequency of fund managers not trading before announcements, only to subsequently
purchase during good announcements. Taken together, this evidence suggests managers are
reliant on private information before earnings announcements, as well as them engaging in
'interpretation' of earnings announcements when they do not receive a private signal.