Implementation of arima and exponential smoothing methods in stock price forecasting

Throughout history, people have made various evaluations to protect their wealth and invest in different areas. One of these areas is stocks, which are among the primary investment instruments traded on the stock exchange. As a result of research in mathematics and statistics, numerous methods have been developed to understand and model stock price behavior. Investors, particularly those focused on future price predictions, have employed a wide range of forecasting methods and algorithms. Among these, ARIMA and Exponential Smoothing are commonly used techniques in time series forecasting. This study aims to build statistical models within a defined confidence interval using historical datasets of financially sound and actively traded stocks. Monthly closing prices of EREGL, PGSUS, and TUPRS from 2017 to 2023 are used as the data source. Forecasting will first be conducted using the ARIMA method, followed by Exponential Smoothing. The goal is to estimate stock prices for the year 2024 based on past data. The performance of the models will be evaluated through graphical comparisons, and the most suitable model will be proposed for practical use. Ultimately, the study seeks to introduce a tested method that can benefit professionals in economics, statistics, and finance. By performing two different analyses, the level of accuracy between the models will be evaluated; this is important for investor confidence. Keywords: ARIMA, BIST100, Stocks, Exponentıal Smoothing