Forecasting the Stock Indexes of Fragile Five Countries through Box-Jenkins Methods
Sedat Yenice, Mustafa Agah Tekindal

Abstract
The main aim of this study is to forecast the prospective daily closing values of the stock indexes of Turkey, Brazil, Indonesia, South Africa, and India that are referred to as developing countries and are called fragile five in the report prepared by Morgan Stanley in 2013. Another aim is to determine whether or not similar index forecasts will be provided for these countries. To this end, 3-year daily index closing data of these countries were used as dataset, and an attempt was made to create forecasting models through ARIMA method. ARIMA models created for all countries were found to be significant and have quite low margins of error. The closing values forecasted via the created models and the actual closes were compared, and the following was realized: The model created for Turkey forecasted with an accuracy rate of approximately 72%; that created for Brazil forecasted with an accuracy rate of approximately 65%; that created for Indonesia forecasted with an accuracy rate of approximately 74%; that created for South Africa forecasted with an accuracy rate of approximately 66%; and that created for India forecasted with an accuracy rate of approximately 59%.

Full Text: PDF