PREDIKSI RISIKO CRASH HARGA SAHAM  BERDASARKAN KETERBUKAAN DATA PUBLIK  MENGGUNAKAN PENDEKATAN MACHINE LEARNING

Aji Susanto; Martanto .; Denni Pratama; Mulyawan .

Authors

Aji Susanto STMIK IKMI Cirebon, Indonesia
Martanto . STMIK IKMI Cirebon, Indonesia
Denni Pratama STMIK IKMI Cirebon, Indonesia
Mulyawan . STMIK IKMI Cirebon, Indonesia

Keywords:

risiko crash saham, keterbukaan data publik, deep learning

Abstract

Penelitian ini mengembangkan dan mengevaluasi kerangka prediktif untuk mengidentifikasi risiko penurunan harga saham tajam dengan menggunakan teknik machine learning dan data publik terbuka. Model Deep Neural Network berbasis LSTM dan beberapa model konvensional (Logistic Regression, Random Forest, XGBoost) digunakan sebagai perbandingan dalam pendekatan yang diuji. Indikator pasar frekuensi harian, variabel makro, dan indikator keterbukaan publik/pengungkapan publik adalah semua komponen yang digunakan dalam dataset yang digunakan. Pra-pemrosesan data (pembersihan, imputasi, normalisasi, framing deret waktu), rekayasa fitur, optimasi hiperparameter (bayesian/optuna untuk DNN), pelatihan dengan validasi rolling-window, dan analisis interpretabilitas menggunakan teknik XAI (SHAP/LIME). Hasil eksperimen menunjukkan fenomena "akurasi paradoks": meskipun semua model memiliki akurasi tinggi pada data uji, ketidakseimbangan recall kelas yang ekstrim dan skor F1 yang hampir nol untuk kelas kecelakaan membuat model tidak mampu menemukan kejadian kecelakaan (kelas minoritas). Dalam kondisi data yang timpang tanpa perlakuan khusus, arsitektur LSTM yang kompleks tidak menunjukkan keuntungan yang signifikan dibandingkan dengan model sederhana. Variabel Disclosure Index memberikan kontribusi marginal, sementara analisis kontribusi fitur menunjukkan dominasi sinyal teknikal, seperti return harian. Hasil menunjukkan bahwa penggunaan metrik evaluasi yang menekankan sensitivitas terhadap kejadian ekstrim (recall, F1, PR-AUC), penanganan imbalance (seperti oversampling/SMOTE, pembelajaran yang sensitif terhadap biaya, dan pengayaan variabel keterbukaan dengan sinyal kualitatif (NLP/sentimen) sangat penting. Studi ini menyarankan agar investor dan regulator berhati-hati saat menggunakan model ML tanpa memverifikasi sensitivitas terhadap peristiwa ekstrim. Mereka juga harus mendorong penelitian lanjutan yang meningkatkan kemampuan sistem peringatan dini dengan menggabungkan data alternatif dan strategi penyeimbangan kelas.

References

Borisov, V., Broelemann, K., Kasneci, E., & Kasneci, G. (2023). DeepTLF : robust deep neural networks for heterogeneous tabular data. International Journal of Data Science and Analytics, 16(1), 85–100. https://doi.org/10.1007/s41060-022-00350-z

Chang, V., Ariel, Q., Shola, X., Akinloye, H., Benson, V., & Hall, K. (2025). Prediction of bank credit worthiness through credit risk analysis : an explainable machine learning study. 247–271.

Chen, Z. (2025). ESG Rating Divergence and the Risk of Corporate Share Price Crashes : Facilitation or Suppression. 0, 178–186. https://doi.org/10.54254/2754-1169/138/2024.19232

Chlebus, M., & Kopczewska, K. (2023). Financial Time Series Models — Comprehensive Review of Deep Learning Approaches and Practical Recommendations †. 1–13.

Emmanuel, I., Sun, Y., & Wang, Z. (2024). A machine learning ‑ based credit risk prediction engine system using a stacked classifier and a filter ‑ based feature selection method. Journal of Big Data. https://doi.org/10.1186/s40537-024-00882-0

Ge, Y., Chen, Q., Qiu, S., & Kong, X. (2023). Environmental information disclosure and stock price crash risk : Evidence from China. February, 1–14. https://doi.org/10.3389/fenvs.2023.1108508

Hansen, J. (2023). Opening industry data : The private sector ’ s role in addressing societal challenges. https://doi.org/10.1017/dap.2023.15

Hjelkrem, L. O. (2023). Explaining Deep Learning Models for Credit Scoring with SHAP : A Case Study Using Open Banking Data.

Htun, H. H., Biehl, M., & Petkov, N. (2023). Survey of feature selection and extraction techniques for stock market prediction. Financial Innovation. https://doi.org/10.1186/s40854-022-00441-7

Jiang, F., Ma, T., & Zhu, F. (2024). Fundamental characteristics, machine learning, and stock price crash risk. Journal of Financial Markets, 69, 100908. https://doi.org/https://doi.org/10.1016/j.finmar.2024.100908

Kayit, A. D., & Ismail, M. T. (2025). Advancing stock price prediction through the development of hybrid ensembles : a comprehensive comparative analysis of machine learning approaches.

Kristanti, F. T., Febrianta, M. Y., Salim, D. F., & Riyadh, H. A. (2024). Advancing financial analytics : Integrating XGBoost , LSTM , and Random Forest Algorithms for precision forecasting of corporate financial distress. 8(8), 1–28.

Levchenko, D., Rappos, E., Ataee, S., & Nigro, B. (2025). Chain-structured neural architecture search for financial time series forecasting. International Journal of Data Science and Analytics, 20(4), 3727–3736. https://doi.org/10.1007/s41060-024-00690-y

Li, Y., Xue, H., Wei, S., Wang, R., & Liu, F. (2024). A Machine Learning Approach for Investigating the Determinants of Stock Price Crash Risk : Exploiting Firm and.

Lin, W., Yu, H., & Wang, L. (2025). A data-driven deep learning approach incorporating investor sentiment and government interventions to predict post-crash stock return in China ’ s A-share market. 10(November 2024).

Lin, Y., Zheng, L., & Zhong, Q. (2025). Public data openness and stock price crash risk : evidence from a quasi-natural experiment of government data platforms. Data Science and Management, 8(3), 284–295. https://doi.org/10.1016/j.dsm.2024.11.002

Mienye, E., Jere, N., Obaido, G., & Mienye, I. D. (2024). Deep Learning in Finance : A Survey of Applications and Techniques. 2066–2091.

Olorunnimbe, K., & Viktor, H. (2023). of practice , backtesting , and applications. In Artificial Intelligence Review (Vol. 56, Issue 3). Springer Netherlands. https://doi.org/10.1007/s10462-022-10226-0

Omar, A. Bin, Huang, S., Salameh, A. A., & Khurram, H. (2022). Stock Market Forecasting Using the Random Forest and Deep Neural Network Models Before and During the COVID-19 Period. 10(July), 1–10. https://doi.org/10.3389/fenvs.2022.917047

Reichmann, D., Möller, R., & Hertel, T. (2022). and stock price crash risk. In Journal of Business Economics (Vol. 92, Issue 9). Springer Berlin Heidelberg. https://doi.org/10.1007/s11573-022-01085-w

Risk, P. C. (2024). Research on the Impact of Climate Risk on Corporate Stock. 0, 183–189. https://doi.org/10.54254/2754-1169/132/2024.18342

Su, F., Zhai, L., & Liu, J. (2023). Do MD & A Risk Disclosures Reduce Stock Price Crash Risk ? Evidence from China.

Theodorakopoulos, L., & Theodoropoulou, A. (2025). Big data in financial risk management : evidence , advances , and open questions : a systematic review. October. https://doi.org/10.3389/frai.2025.1658375

Vancsura, L., & Tatay, T. (2025). Navigating AI-Driven Financial Forecasting : A Systematic Review of Current Status and Critical Research Gaps. 1–49.

Wirtz, B. W., Weyerer, J. C., Becker, M., & Müller, W. M. (2022). Open government data : A systematic literature review of empirical research. Electronic Markets, 2381–2404. https://doi.org/10.1007/s12525-022-00582-8

Xu, F., Ji, Q., & Yang, M. (2021). The Pitfall of Selective Environmental Information Disclosure on Stock Price Crash Risk : Evidence From Polluting Listed Companies in China. 9(June), 1–14. https://doi.org/10.3389/fenvs.2021.622345

Xu, N., Liu, J., & Dou, H. (2022). governance information disclosure and stock price crash risk : Evidence from Chinese listed companies. September, 1–15. https://doi.org/10.3389/fpsyg.2022.977369

Zhan, Z., & Kyoo, S. (2024). Versatile time ‑ window sliding machine learning techniques for stock market forecasting. Artificial Intelligence Review, 57(8), 1–25. https://doi.org/10.1007/s10462-024-10851-x

Zhang, L., & Sok, S. (2025). The impact of government open data platform construction on corporate capital market performance: Evidence from stock liquidity. Pacific-Basin Finance Journal, 90, 102667. https://doi.org/https://doi.org/10.1016/j.pacfin.2025.102667

Zhong, Q., Song, Q., & Lee, C. C. (2024). Managing crash risks through supply chain transparency : evidence from China. Financial Innovation. https://doi.org/10.1186/s40854-024-00633-3

Zhou, M., Wang, Y., Jiang, H., Li, M., & Li, G. (2023). How Leadership Influences Open Government Data ( OGD ) -Driven Innovation : The Mediating Role of Organizational Commitment.

Zong, H., & Wang, Y. (2025). Impacts of government public data openness and digital financial development on corporate information reliability. International Review of Economics & Finance, 103, 104495. https://doi.org/https://doi.org/10.1016/j.iref.2025.104495