Efficiency versus robustness of markets : why improving market efficiency should not be the only objective of market regulation

The efficiency of capital markets has been questioned almost as long as the efficient market hypothesis had been worked out. Numerous critics have been formulated against this hypothesis, questioning notably the behavioural assumptions underlying the efficient market hypothesis. The present contribution does not focus on the behavioural assumptions but rather looks at the implications of focusing purely on the objective of market efficiency when considering market design questions. Hence it aims at discussing the following, possibly rather fundamental issue: Is the objective of efficiency, which has guided most of the market reforms in the last decades, sufficient? Or has it to be complemented by the objective of robustness? Mathematical and engineering control theory has developed the concept of robust control (e.g. Zhou and Doyle, 1998) and it has been shown that there is always a trade-off between the efficiency of a control system and its robustness (cf. e.g. Safonov, 1981, Doyle et al., 1988). The efficiency of the system describes its reactions to disturbance signals. The lower the integral loss function over the so-called transfer or sensitivity function, the less a system is affected by disturbances such as demand fluctuations, and the more efficient is the control. The economic equivalent clearly is the maximisation of welfare, which results in an efficient economic system. Robustness by contrast is defined as stability of the control system in the presence of model uncertainty (deviations in the model parameters or misperceptions of the underlying system). These concepts are applied to the financial markets in their interaction with the real economy. The financial markets being understood as the controllers of real world activity through investments, the implications of misperceptions in the financial sphere are analysed both theoretically and in an application example. From the theory it may readily derived that financial markets providing efficient, i.e. welfare-optimal solutions, must have limitations with respect to robustness. Also in the application example it turns out that in the presence of potential misperception a reduction of irreversible cost shares in investments may lead to an increase in overall expected system costs. Hence improvements in (conventional) market efficiency may be counter-productive by facilitating misallocation of capital as a consequence of misperceptions in the financial markets. This leads to the conclusion that a sole focus on the efficiency objective in market design is problematic and some of the recent turmoil in financial markets may be explained by the lack of consideration given to robustness issues.