Abstract:

Forecasting of gold prices has been an attracted issue in literature and practice for a long time. For this reason, different methods for forecasting of the price of gold have been applied and compared with each other. In this study, in order to determine the price of gold based on its lagged value, simple exponential smoothing method, Holt's linear trend method, ARIMA (Autoregressive Integrated Moving Average) method, which are time series methods and the Artificial Neural Networks (ANN), which is an artificial intelligence method, have been applied and made comparison of these methods. The monthly weighted average of the gold price ($/ons) received from the Istanbul Gold Exchange has been used in the analysis. The data covered the period from January 1996 to December 2013. As a result of the analysis, the ARIMA model has been found more successful than the ANN model but it has been determined that the ANN illustrated more successful forecasting performance with comparison of the simple exponential smoothing method and Holt's linear trend method

Keywords: