FORECASTING STOCK INDEX RETURN WITH NEURO FUZZY NETWORK STRUCTURE: BORSA İSTANBUL (BIST) 100 CASE
BULANIK SİNİR AĞI YAPISI İLE BORSA ENDEKS GETİRİSİ TAHMİNİ: BORSA İSTANBUL (BİST) 100 ÖRNEĞİ

Author : Fatih GÜZEL -- Melek ACAR BOYACIOĞLU,
Number of pages : 452-465

Abstract

The subject of this research is forecasting the main indicator of Borsa İstanbul stock market, BIST 100 index return, with using Fuzzy neural network. Although it has not achieved sufficient efficiency and depth in Turkey, stock market index is assumed as an alternative investment tool. Thus main field of this study is determined to be stock market index. Also, excessive return rate observed in some years makes index more interesting. Stock indexes are the parameters that may be indicators of the overall economic trend. Macroeconomic factors, were selected as independent variables in forecasting the return of the index. Ounce gold price, currency basket, the consumer price index, interest rate on deposits, the Dow Jones index, the current account deficit and gross domestic product are independent variables of the study. The model was used to forecast is Artificial Neural Networks backed by fuzzy inference systems, which is Adaptive Neuro-Fuzzy Inference System-ANFIS approach. Analysis

Keywords

Stock exchange index, artificial intelligence, artificial neural network, ANFIS

Read: 740

Download: 212