Year: 2024
Type: article
Abstract: The NASDAQ Composite Index, a critical barometer of market trends heavily influenced by technology firms, poses significant forecasting challenges due to various external influences, such as economic conditions and geopolitical developments. This study delves into the predictive capabilities of three prominent time series models—Exponential Smoothing (ETS), Autoregressive Integrated Moving Average (ARIMA), and Linear Regression are key forecasting techniques used for time series analysis—utilizing monthly closing data spanning from September 2019 to August 2024. A comprehensive analysis is conducted to compare these models based on Root Mean Squared Error (RMSE) and Akaike Information Criterion (AIC) are important metrics used for evaluating model performance. The findings indicate that ARIMA, with its robust mechanism for handling non-stationary data, provides superior forecasting accuracy compared to the other models. In contrast, while ETS demonstrates effective forecasting capabilities, particularly in capturing level and trend, Linear Regression is found lacking due to its challenges in accurately capturing the intricate dynamics and temporal dependencies of the market. These findings offer important insights for investors, highlighting the need for sophisticated and detailed forecasting methods in the constantly changing financial environment. Additionally, this study paves the way for future research that could incorporate external factors, like macroeconomic indicators and sentiment analysis, while also examining the promise of machine learning techniques to enhance the precision and robustness of financial market forecasts. This holistic approach could significantly improve the understanding of market behavior and inform better investment strategies.
Author: Lin Cao
Institution: New York University
Language: English
Field: Decision Sciences
Domain: Social Sciences
Open Access status: hybrid
OpenAlex ID: W4405818448