Forecasting stock returns with panel data analysis: A research in İstanbul Stock Exchange

In addition to providing funds for firms, security exchanges which have a significant place in the country?s economy have the function of getting returns for the savings of investors. One of the important investment instruments of the security exchanges is the stocks. Although investors use various methods in order to get the highest return, basically there are two methods: Fundamental analysis and technical analysis. Fundamental analysis estimates the stock prices by using financial ratios and macroeconomic variables. However, technical analysis estimates the stock prices by using only the past prices of stocks.In this study, the stock prices of the firms in industrial, financial and services sectors are forecasted by using the variables constituting the basis of both fundamental and technical analyses, namely financial ratios, macroeconomic variables and the past prices of stocks. For this aim, one-way and two-way fixed effects, Arellano-Bond and Arellano-Bover dynamic panel data regression models are used.From the panel data regression models, for the industrial sector: book-to-market ratio, inflation rate, interest rate and foreign exchange rate; for the financial sector: book-to-market ratio, earnings pers hare and all macroeconomic variables and for the services sector: book-to-market ratio, current asset turnover, inflation rate and interest rate are found the significant variables that determine the stock prices.