Linear and nonlinear dependence in Turkish equity returns and its consequences for financial risk management

This paper investigates the dynamic behaviour of daily aggregate returns of one of Europe's largest and fastest growing emerging equity markets, the Istanbul Stock Exchange (ISE). It is found that ISE returns exhibit significant linear and nonlinear dependence. We investigate the nature and source of the nonlinear dependence and find that it is due primarily to linear dependence in the conditional variance of returns, rather than nonlinear dependence in the conditional mean. We analyse the implications of our findings for financial risk management. We show that by exploiting the nonlinear dependence in ISE returns, the average capital required to cover against unexpected portfolio losses can be considerably reduced. In contrast, exploiting the linear dependence in ISE returns has only a negligible impact on capital requirements.