Dual long memory models: An application on some macroeconomic variables

One of the important indicators showing general achievement of national economies is the stock exchange market. By monitoring changes in stock exchange indexes, it is possible to understand the effects of inflation, growth, investments and economic crisis on the economy. Within this context, the concept of efficiency is also important in terms of financial markets. Efficient Market Hypothesis asserts that a new information received in the market reaches over to all investors at the same time and accordingly the prices of securities cannot be estimated with reference to the previous prices. Hyperbolic decrease tendency in autocorrelation functions of return and volatility of stock markets which means slow mean reversion is defined as long memory. In case the existance of long term dependency among price movements, there is positive autocorrelation among price movements. In contrast to the Efficient Market Hypothesis, in case of having Long Memory property, stock market prices will have a predictable structure and retrospective tendency of prices can be used for price estimations in future. In this study, the long memory properties in return and volatilities of stock exchange markets in Turkey, Malaysia, Indonesia and Pakistan which are selected among developing countries (D-8) as the countries having gradually increasing importance in global economy have been examined, and the results of the Efficient market Hypothesis in developing real markets are investigated. Furthermore, symmetric and asymmetric ARFIMA-FIGARCH model types for the presence of long memory properties are estimated under different distribution assumptions such as Normal, Student-t, GED and Skewed Student-t. The results indicate that there is long memory property in volatilities of stock market return, and the effects of shocks are aymmetric. According to obtained findings, the predictable structure of volatility represents that all of the stock markets are weak form inefficient. Consequently, it is possible to say that technical analysis related to these stock market may be valid.