Examination of Relationship Between Stock Returns and Factors Affecting Capital Structure and Financial Ratios with ANFIS Method: An Application on ISE 100 Index

In this study, the stock return rate of the company shares was estimated by means of the adaptive neuro-fuzzy inference system (ANFIS). For this aim, the data of financial ratios prepared for the input and output variables of the ANFIS was obtained by using the proportion parameters from the items of balance sheet and income statement of the ISE 100 companies in the industry between the years of 2003 and 2007 announced by 40 companies, and the obtained data was applied into the ANFIS. In the implemented experiment, the values estimated by the ANFIS were compared with the real values of the stock return rate, and significant 4 input variables among 14 input variables were determined to apply into the ANFIS. When the same experiment was implemented by using these determined significant input variables, the ANFIS model reached to approximately same estimation success. As a result, we showed that this model provides the same success by using less number of input variables and it can be used for decreasing the risk of companies and investors.