Scholarlink

<p>Following the establishment of Indonesia’s sovereign wealth fund, a move poised to reshape the nation’s investment environment, there has been an increasing need to interpret market sentiment and its influence on stock performance. This study bridges that gap by applying Support Vector Regression (SVR) to forecast the stock prices of the nation’s three largest banks under Danantara, utilizing macroeconomic factors, historical prices, directional inter-stock relations, and numerical public sentiment from Twitter (X). While macroeconomic and historical factors have immediate effects, negative sentiment has a delayed but statistically significant effect. Directional dependence testing reveals powerful same-day dependencies among stocks, with banks of similar market roles being particularly prominent. The inclusion of sentiment variables alongside optimized SVR parameters has been found to significantly improve predictive accuracy, highlighting the critical, context-specific role of public sentiment in guiding financial markets amid structural economic change.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Directional Dependence-Based SVR Modeling for Sentiment-Driven Forecasting of Indonesian Banking Stocks

  • 期刊论文

  • 发表日期：

    2025年12月9日

查看全文
  • Allun Samudra Harum,
  • Hilal Azhar Fawaz,
  • Atina Ahdika,
  • Muhammad Fakhruddin

摘要

Following the establishment of Indonesia’s sovereign wealth fund, a move poised to reshape the nation’s investment environment, there has been an increasing need to interpret market sentiment and its influence on stock performance. This study bridges that gap by applying Support Vector Regression (SVR) to forecast the stock prices of the nation’s three largest banks under Danantara, utilizing macroeconomic factors, historical prices, directional inter-stock relations, and numerical public sentiment from Twitter (X). While macroeconomic and historical factors have immediate effects, negative sentiment has a delayed but statistically significant effect. Directional dependence testing reveals powerful same-day dependencies among stocks, with banks of similar market roles being particularly prominent. The inclusion of sentiment variables alongside optimized SVR parameters has been found to significantly improve predictive accuracy, highlighting the critical, context-specific role of public sentiment in guiding financial markets amid structural economic change.