Analysis of Large Language Model-Based Stock Investment Recommendations for LQ45 Stocks

Dian Anggraeni; Muryani Arsal; Muchriana Muchran

doi:10.24256/kharaj.v8i3.12313

Authors

Dian Anggraeni Master of Management Postgraduate Program, Muhammadiyah University of Makassar, Indonesia
Muryani Arsal Master of Management Postgraduate Program, Muhammadiyah University of Makassar, Indonesia
Muchriana Muchran Master of Management Postgraduate Program, Muhammadiyah University of Makassar, Indonesia

DOI:

https://doi.org/10.24256/kharaj.v8i3.12313

Keywords:

Large Language Models; stock investment; investment recommendation; LQ45 Index; artificial intelligence; decision support system.

Abstract

This study aims to examine the capability of Large Language Models (LLMs) to generate stock investment recommendations in the form of buy, hold, and sell for stocks included in the LQ45 Index of the Indonesia Stock Exchange. A quantitative experimental approach was employed using three generative artificial intelligence models: ChatGPT, Gemini, and DeepSeek. Each model received an identical prompt incorporating stock price information, trading volume, candlestick patterns, and macroeconomic and microeconomic sentiment. The research sample comprised 45 LQ45 constituent stocks observed over 20 trading days in December 2025, resulting in 900 observations for each model and 2,700 observations in total. Model performance was evaluated based on three primary dimensions: recommendation accuracy relative to actual stock price movements, investment returns, and downside risk. The findings indicate discrepancies between LLM-generated recommendations and actual stock price movements, suggesting that none of the three models can predict stock price direction with complete accuracy. Nevertheless, all models generated positive cumulative returns. DeepSeek achieved the highest Compounded Cumulative Return (CCR) at 24.01%, followed by Gemini at 23.58% and ChatGPT at 1.90%. In terms of risk, ChatGPT demonstrated the most favorable performance, recording the lowest total risk at 16.35%, compared with DeepSeek at 22.34% and Gemini at 24.85%. The loss rates were also below 50% across all models, namely 34.11% for DeepSeek, 37.88% for Gemini, and 27.33% for ChatGPT. These findings suggest that LLMs can function as Decision Support Systems for investment decision-making by generating recommendations with the potential to produce positive returns while maintaining relatively moderate risk. However, LLM-generated recommendations should not be interpreted as definitive market forecasts because their performance remains sensitive to market volatility, information dynamics, and rapidly changing financial conditions.

References

Andersen, T. G., & Bollerslev, T. (2021). Volatility and trading volume: A persistent relationship. Journal of Finance, 76(2), 1043–1091. https://doi.org/10.1111/jofi.12975

Borges, M. R. (2014). Efficiency testing of Asian stock markets. International Journal of Finance & Economics, 19(2), 140–155. https://doi.org/10.1002/ijfe.1489

Cardillo, A. (2025). 65 Most Popular AI Tools Ranked (October 2025)tle. Exploding Topics.

Chawla, D., Bhutada, A., Anh, D. D., Raghunathan, A., SP, V., Guo, C., Liew, D. W., Gupta, P., Bhardwaj, R., Bhardwaj, R., & Poria, S. (2025). Evaluating {AI} for Finance: Is {AI} Credible at Assessing Investment Risk Appetite?

Chen J., W. Y., & Chen, X. (2023). Can ChatGPT forecast stock price movements? Return predictability and large language models. Finance Research Letters, 56, 103664. https://doi.org/10.1016/j.frl.2023.103664

Chen, J., Wang, Y., & Chen, X. (2023). Can ChatGPT forecast stock price movements? Return predictability and large language models. Finance Research Letters, 56, 103664. https://doi.org/10.1016/j.frl.2023.103664

Chen, X., Wang, J., & Zhong, X. (2024). Return Predictability of Prospect Theory: Evidence from the Thailand Stock Market. Pacific-Basin Finance Journal, 83, 102199. https://doi.org/10.1016/j.pacfin.2023.102199

Chordia, T., Roll, R., & Subrahmanyam, A. (2020). Market liquidity and trading activity. Journal of Financial Economics, 135(3), 644–670. https://doi.org/10.1016/j.jfineco.2019.06.008

Dowling, M., & Lucey, B. (2023a). ChatGPT for (finance) research: The Bananarama conjecture. Finance Research Letters, 53, 103662. https://doi.org/10.1016/j.frl.2023.103662

Eka, M. (2024). Gemini AI :Kecerdasan Buatan Canggih Dari Google. Direktorat Pusat Teknologi Informasi.

Fama, E. F. (1970). Efficient capital markets: A review of theory and empirical work. The Journal of Finance, 25(2), 383–417. https://doi.org/10.2307/2325486

Gartner. (2023). Gartner Predicts More Than 80% of Enterprises Will Use Generative {AI} {API}s or Deploy Generative {AI}-Enabled Applications by 2026.

Hangyan. (2025). 印尼股票市场风险收益比 [{Rasio} risiko-return pasar saham {Indonesia}].

Jaglal, N. U., & Kose, U. (2024). Descriptive Statistics for Stock Risk Assessment. In Data Analytics for Smart Cities. CRC Press (Taylor & Francis).

Kahneman, D., & Tversky, A. (1979). Prospect theory: An analysis of decision under risk. Econometrica, 47(2), 263–291. https://doi.org/10.2307/1914185

Ko, H., Lee, J., & Park, S. (2024). {ChatGPT} and portfolio diversification. Finance Research Letters, 59, 104780. https://doi.org/10.1016/j.frl.2023.104780

Kumbure, M. M., Lohrmann, C., Luukka, P., & Porras, J. (2022a). Machine learning techniques and data for stock market forecasting: A literature review. Expert Systems with Applications, 197, 116659. https://doi.org/10.1016/j.eswa.2022.116659

Kustodian Sentral Efek Indonesia. (2025). Jumlah investor pasar modal {RI} capai 20,32 juta {SID} di akhir 2025.

Li, X., Zhang, R., & Zhou, N. (2022). A hybrid deep learning model integrating LSTM and Transformers for stock trend prediction. Applied Intelligence, 52, 9800–9817. https://doi.org/10.1007/s10489-022-03371-1

Liu, Y., Jiang, Z., & Wang, H. (2023). Price--volume interaction and short-term stock return predictability. International Review of Financial Analysis, 86, 102511. https://doi.org/10.1016/j.irfa.2023.102511

Marr, B. (2023). A Short History Of ChatGPT: How We Got To Where We Are Today. FORBES.

Marshall, B. R., Nguyen, N. H., & Visaltanachoti, N. (2021). Trading volume and technical analysis signals. Journal of Banking & Finance, 122, 105987. https://doi.org/10.1016/j.jbankfin.2020.105987

Martinez, R., Kumar, A., Singh, P., Chatterjee, S., Das, M., Roy, S., & Ghosh, D. (2026). Evaluating the efficacy of large language models in stock market decision-making: A decision-focused, price-only, multi-country analysis using historical price data. Machine Learning and Knowledge Extraction, 8(4), 104. https://doi.org/10.3390/make8040104

Nazlioglu, S., Pazarci, S., Kara, A., & Varol, O. (2024). Efficient Market Hypothesis in Emerging Stock Markets: Gradual Shifts and Common Factors in Panel Data. Applied Economics Letters, 31(18), 1773–1779. https://doi.org/10.1080/13504851.2023.2206613

Nelson, D. M. Q., Pereira, A. C. M., & de Oliveira, R. A. (2017). Stock market’s price movement prediction with LSTM neural networks. Proceedings of the International Joint Conference on Neural Networks (IJCNN), 1419–1426. https://doi.org/10.1109/IJCNN.2017.7966019

Nison, S. (2001). Japanese Candlestick Charting Techniques. Penguin Publishing Group.

Oehler, A., & Horn, M. (2023b). Financial advice from robo-advisors versus large language models: Evidence from risk profiling and suitability. Finance Research Letters, 58, 104222. https://doi.org/10.1016/j.frl.2023.104222

Otoritas Jasa Keuangan. (2025). Tingkat melek & akses pasar modal {RI} masih rendah, begini hasil surveinya.

Pelster, M., & Val, J. (2024a). Can ChatGPT Assist in Picking Stocks? Finance Research Letters, 59, 104786. https://doi.org/10.1016/j.frl.2023.104786

Pring, M. J. (2014). Technical Analysis Explained: The Successful Investor’s Guide to Spotting Investment Trends and Turning Points (5th ed.). McGraw-Hill.

Rao, A., & Patel, M. (2023). Stock market prediction using GPT-based sentiment extraction: Evidence from high-frequency tweets. Expert Systems with Applications, 228, 120310. https://doi.org/10.1016/j.eswa.2023.120310

Schneider, J., & Yilmaz, F. (2025). Multifaceted variability in {LLM}-driven stock recommendations. Journal of Banking & Finance, 169, 107320. https://doi.org/10.1016/j.jbankfin.2025.107320

Setiawati, S. (2025). Sosok di Balik DeepSeek yang Bikin AI Amerika Ketar-Ketir. CNN Indonesia.

Sezer, O. B., Gudelek, M. U., & Ozbayoglu, A. M. (2020). Financial time series forecasting with deep learning: A systematic literature review: 2005–2019. Applied Soft Computing, 90, 106181. https://doi.org/10.1016/j.asoc.2020.106181

Simon, H. A. (1957). Models of Man: Social and Rational. John Wiley & Sons.

Solares, E. A., Aiello, L. M., & others. (2022). A comprehensive decision support system for stock market investment. Decision Analytics Journal, 4, 100075. https://doi.org/10.1016/j.dajour.2022.100075

Technology, T. (2025). Top 10 Most Used Gen AI Tools In 2025. Titan Technology.

Unesa. (2025). DeepSeek: Terobosan AI Hemat Biaya yang Mengguncang Dunia Teknologi. Fakultas Teknik Universitas Negeri Suraba.

Xu, L., Hu, Z., & Li, C. (2024). Large language models and market efficiency: Evidence from news-based trading signals. Journal of Financial Markets, 67, 100843. https://doi.org/10.1016/j.finmar.2023.100843

Zhang, Y., & Chen, W. (2022). Volume-confirmed breakout strategies and stock market performance. Finance Research Letters, 47, 102785. https://doi.org/10.1016/j.frl.2021.102785