Stock Recommendation for Students Using Technical Analysis and LSTM-Based Forecasting

Abstract

Stock investment is increasingly popular among Indonesian students, particularly among millennials and Generation Z, who account for over 79% of investors under the age of 30, according to a 2022 report by the Indonesia Stock Exchange (IDX) and FEB UI. However, despite growing interest, many students remain reluctant to begin investing due to limited financial literacy and a lack of risk analysis skills. Investment decisions are often influenced by the Fear of Missing Out (FOMO), leading to trend-following behavior without proper analytical assessment, under the mistaken belief that following the majority guarantees profit and safety. This study aims to address these challenges by developing a website-based stock recommendation system utilizing the Long Short-Term Memory (LSTM) algorithm to assist beginner investors. The system analyzes historical stock data from IDX80 comprising liquid stocks suitable for novices retrieved from Yahoo Finance. Technical indicators such as Moving Average, Relative Strength Index (RSI), and Bollinger Bands are used to support the predictive model. The LSTM algorithm forecasts stock price movements based on temporal market anomalies, specifically the Turn-of-the-Month (ToM) and Day-of-the-Week (DoW) effects. Predictions are classified into uptrends and downtrends and presented via a user-friendly web interface, allowing users to make more informed investment decisions. Evaluation results show that 63 out of 80 stock tickers achieved an R² value above 0.8, indicating strong predictive performance. Unlike prior studies that primarily focused on general market forecasting or professional investor tools, this research uniquely integrates behavioral finance aspects (such as FOMO) with algorithmic predictions targeted at novice investors. The incorporation of calendar effects into LSTM-based forecasting also fills a methodological gap rarely addressed in existing literature. These findings suggest that personalized, accessible forecasting tools can bridge the knowledge gap for young investors and support more rational, data-driven investment behavior.