Summary: In 1992, turnover on the New York Stock Exchange was 48 percent. While there is no convincing theoretical prediction for assessing this number, observers may have the view that turnover is very high. The increase in turnover has been accompanied by a rise in institutional ownership. A regression of turnover on institutional ownership and real commissions per share shows that institutional ownership is still highlysignificant in explaining turnover. The available evidence is at least suggestive of a causal link between turnover and institutional control.
It seems difficult to explain the level of trading activity purely on the basis of 'rational' motives for trade. The authors argue that the motive stems from a contracting problem between professional traders and their clients or employers. The contracting problem in the authors model is whether the delegated portfolio manager can convince the client/employer that inactivity was his best strategy. The difficulty is that the employer cannot distinguish "actively doing nothing" in this sense from "simply doing nothing." If the contract allows a reward for not trading, portfolio managers may simply do nothing; the contract may either attract incompetent managers or lead competent managers to shirk. If this makes it impossible to reward inactivity, and limited liability prevents punishing ex post incorrect decisions, then the optimal contract may induce trading by the portfolio manager which is simply a gamble to produce a satisfactory outcome by change. The authors call this noise trading or churning and show that the noise trade will occur in equilibrium.
The paper then considers the implication of noise trading for agents welfare. Noise trading would appear to be costly for the employer since it lowers the expected rate of return on the portfolio. It will benefit hedgers; if managed portfolios earn lower rates of return, then uninformed hedgers earn higher returns. The higher return earned by the hedgers effectively reduces the cost of hedging; as a result they will trade larger amounts. In turn, this increase in volume can support a larger amount of investment by an informed fund manager. If the manager earns a smaller (percentage) return on a sufficiently increased investment, then he will be better off.
The model is a general equilibrium model of portfolio management in a security market. The authors conclude that a portfolio manager will frequently find that the best investment policy is simply to hold the existing portfolio. The question is whether, in this situation, he will be able to credibly convince his client or employer that he is 'actively' doing nothing. The client may instead believe that he is simply doing nothing. He may think that the portfolio manager has not spent any effort on producing information or he has no talent. The paper describes a contractualrelationship, and its economic consequences, where actively doing nothing is indistinguishable from simply doing nothing. Ultimately it is an empirical question as to when these are indistinguishable. Designing a contractual relationship for portfolio management is to a large extent a matter of maximizing this distinction.
Noise trade is a manifestation of this agency problem. Because all agents objectives are specified, the authors can examine the welfare implications of this agency problem. The example discussed shows that noise trade, by making the market more liquid, can benefit everyone. This illustrates that welfare effects can be more subtle and more complex than is allowed by standard models with exogenous noise traders.