This study attempts to develop new market efficiency tests depending on the spot and future prices, or the differences of them alternative to traditional unit root test build on univariate time series. As a result of the autocorrelation, normality and run tests applied to spot and futures prices or differences of them, and Adopted Purchasing Power Parity test based on a regression the future markets of ISE30, ISE100 index indicators, USD and Euro currencies, all of which have been traded dailly in the Izmir Futures and Options Market for five years, are found inefficient. Autocorrelation, normality and run tests on the differences between spot and futures prices series, and Adopted Purchasing Power Parity test, or autocorrelation, normality and run tests test based on spot series all rejected “the acceptance of efficient market hypothesis” under the existence of unit root in a series. The results of autocrrelation, normality and run tests based on univariate series are found contradictory to the unit root test result. As a result, the acceptance of “efficient market hypothesis” under the existence of unit root is not supported by alternative tests developed in this study. It is suggested that efficiency test shall be stepping on the spot and futures prices; differences of them or Adopted Purchasing Power Parity test developed here rather than unit root test based on univariate series, which is also not appropriate to the definition of futures market efficiency. In addition, one must be sure that the errors disturbances are randomized in deciding whether market is efficient or not.

Keywords:Futures markets, spot and futures prices, alternative efficiency tests, unit root, autocorrelation, normality, run, Adopted PPP test, Turkey