Enhanced Stock Market Movements towards Economic Sustainability using LSTM Hybrid Algorithm
Abstract
The need for precise stock price evaluation and projection is highlighted by the stock market’s increasing importance in the national economy, a topic that has attracted a lot of scholarly attention. Predictive modelling faces a very dynamic and difficult task when it comes to stock price movements, which are impacted by a variety of elements, including macroeconomic policy, industry news, capital flows, market sentiment, and corporate fundamentals. Rising volatility and increasing complexity are making themselves felt on the global stock markets; hence, there is a need for advanced prediction models that may incorporate financial, macroeconomic, and sustainability variables. This paper has therefore suggested a hybrid Long Short-Term Memory (LSTM) model that combines sentiment analysis of financial news and Principal Component Analysis (PCA) choice of features to predict stock market movements based on the aspect of economic sustainability. Historical stock price data, macroeconomic variables, and ESG (Environmental, Social, and Governance) scores for several global indexes between 2010 to 2024 are fed into the algorithm. The experimental results prove that in consideration of factors such as directional accuracy and predictive accuracy, the suggested LSTM-Hybrid framework performs better than any traditional LSTM or ARIMA model. Informed investment strategies can thus drive investors and policymakers toward economic sustainability in the long run.