Modeling the variables affecting stock prices with c4.5 decision tree algorithm

Investors want to earn above-average returns by using various financial analysis methods. However, which financial analysis methods will be used depends on the investor profile and the type of stock market to be traded. Market efficiency studies, related to BIST Stock Exchange, demonstrate that the Fundamental Analysis Method for BIST can be used for generating above-average returns. Fundamental Analysis consists of a process starting from macro variables to lasting firm analysis. In this respect, it is crucial for investors to identify and examine macroeconomic and microeconomic variables affecting the shares which they are interested in. However, making this analysis for each stock may be difficult and time consuming for investors. In this regard, producing set of shortcut rules for stocks including the relevant analysis is particularly beneficial for the investor profile preferring irrational behavior. In this respect, within the scope of the study, it was tried to establish set of shortcut rules for stocks traded in BIST 100. At the beginning, a literature study was performed to determine the macroeconomic and microeconomic variables to be used in the study. After that, the selected variables were evaluated by finance experts with the survey. Analysis was performed with C4.5 Decision Tree Algorithm, which is one of the data mining methods, for the 69 stocks in BIST 100 whose data set is complete for 2006Q1 - 2017Q3 period. According to the results of the analysis, model's correctly classifying rate of the successful periods for the 69 stocks is 92.51%. As a result of this research; sets of rules were created for the 69 stocks, the results were shared according to the sectors, macroeconomic and microeconomic variables were compared with relevant literature and the results of the opinion surveys of finance experts.