Académique Documents
Professionnel Documents
Culture Documents
The authors argue that the activities of brokerage house analysts increase
stock liquidity and consequently reduce required stock returns. They study
the return premiums generated by illiquid stocks and find statistically
significant return premiums for both the fixed and the variable cost
components of transaction costs. The authors believe that this finding shows
the social value of security analysis.
When uninformed investors trade in stocks with investors who have private
information, the uninformed traders incur adverse selection costs. These
costs are higher for trading in illiquid than liquid stocks. To compensate
uninformed traders, the risk-adjusted return on illiquid stocks should,
therefore, be higher than that on liquid stocks. The authors measure this
illiquidity premium.
Previous studies measuring the illiquidity premium have used the bid–ask
spread as a proxy for illiquidity. The authors consider the bid–ask spread to
be a noisy measure because many large trades occur outside the spread
and many small trades occur inside the spread. Previous theoretical and
empirical research also suggests that adverse selection costs are associated
with the price impact of trade and, therefore, with the variable component of
trading cost. Consequently, the authors use intraday transaction data to
estimate the fixed and variable costs of transacting, which serve as proxies
for illiquidity.
Previous studies have used the capital asset pricing model to account for
risk. Brennan and Subrahmanyam extend the analysis by using the three-
factor model developed by Fama and French. According to this model, the
factors determining the expected return on a stock are the excess return on
the market, the size of the firm, and the ratio of its book value of equity to
market value.
The authors use transactions data collected by the Institute for the Study of
Securities Markets. The research covers the period from 1984 through 1991.
The sample of stocks consists of 1,629 NYSE-listed firms for 1984 through
1987 and 1,784 firms for 1988 through 1991.
The authors find that both fixed and variable costs have a significant positive
effect on equilibrium rates. The pattern of the return premium (concave for
variable costs, convex for fixed costs) is consistent with a trade-size–
clientele effect (investors trading in small quantities invest in illiquid stocks)
but not consistent with a horizon-clientele effect (investors with long holding
periods invest in illiquid stocks). Previous research, however, has found a
horizon-clientele effect. The authors ascribe their failure to detect this effect
to either an incomplete risk adjustment or to problems arising from not
accounting for price discreteness in their model.
The authors find a negative relationship between the fixed and variable costs
of trading. They cannot adequately explain this finding and believe that it
could be a fruitful area for further research.
Author Information
The abstract was prepared by Johann de Villiers, CFA, the University of the
Witwatersrand, Johannesburg, South Africa.