Pergeseran Feature Importance pada Prediksi Pasar Saham Teknologi Menggunakan Machine Learning: Studi Komparatif Pra dan Pasca Pandemi COVID-19
Abstract views: 7
,
PDF downloads: 4
Abstract
Tujuan: Penelitian ini bertujuan untuk menganalisis apakah model machine learning (XGBoost dan Random Forest) mengalami pergeseran dalam menentukan fitur terpenting untuk memprediksi arah pergerakan saham teknologi sebelum dan sesudah pandemi COVID-19.
Metode: Data yang digunakan adalah time-series harian dari Invesco QQQ Trust (QQQ) sebagai representasi sektor teknologi Amerika Serikat, serta variabel makroekonomi dan volatilitas. Periode penelitian dibagi menjadi dua rezim: pra-pandemi (2018–2019) dan pasca-pandemi (2021–2022). Model dilatih secara terpisah untuk masing-masing rezim, kemudian dilakukan analisis komparatif terhadap feature importance. Evaluasi model menggunakan metrik Accuracy dan F1 Score.
Hasil: Hasil menunjukkan adanya peningkatan prediktabilitas pasar pada periode pasca-pandemi, dengan F1 Score XGBoost meningkat dari 0,346 menjadi 0,556 dan Random Forest dari 0,164 menjadi 0,544. Analisis feature importance menunjukkan pergeseran dominasi faktor: pra-pandemi dipengaruhi secara merata oleh harga, teknikal, dan makroekonomi, sedangkan pasca-pandemi lebih didominasi faktor makroekonomi (FedFundsRate) dan volatilitas (ATR, VIX).
Kesimpulan: Penelitian ini menyimpulkan bahwa pandemi COVID-19 menyebabkan perubahan rezim prediktif di pasar saham teknologi, dengan meningkatnya peran faktor makroekonomi dan volatilitas. Temuan ini menegaskan pentingnya adaptasi model prediksi serta memberikan wawasan praktis bagi investor dalam memahami dinamika pasar pasca-pandemi.
Downloads
References
Fischer, T., & Krauss, C. (2018). Deep learning with long short-term memory networks for financial market predictions. European Journal of Operational Research, 270(2), 654-669.
Baker, S. R., Bloom, N., Davis, S. J., Kost, K., Sammon, M., & Viratyosin, T. (2020). The unprecedented stock market impact of COVID-19. The Review of Asset Pricing Studies, 10(4), 742–758.
Tsay, R. S. (2005). Analysis of financial time series (2nd ed.). John Wiley & Sons.
Thorbecke, W. (2020). The Impact of the COVID-19 Pandemic on the U.S. Economy: Evidence from the Stock Market. Journal of Risk and Financial Management, 13(10), 233.
Henrique, B. M., Sobreiro, V. A., & Kimura, H. (2019). Literature review: Machine learning techniques applied to financial market prediction. Expert Systems with Applications, 124, 226–251.
James, G., Witten, D., Hastie, T., & Tibshirani, R. (2013). An introduction to statistical learning: With applications in R. Springer.
Murphy, J. J. (1999). Technical analysis of the financial markets: A comprehensive guide to trading methods and applications. New York Institute of Finance.
Shiller, R. J. (2003). From efficient markets theory to behavioral finance. Journal of Economic Perspectives, 17(1), 83–104.
Bussmann, N., Giudici, P., Marinelli, D., & Papenbrock, J. (2021). Explainable machine learning in credit risk management. Computational Economics, 57(1), 203-228.
Copyright (c) 2026 Dzaky Rayssa Buntoro
This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.
Authors who publish with this journal agree to the following terms:
1. Copyright on any article is retained by the author(s).
2. The author grants the journal, right of first publication with the work simultaneously licensed under a Creative Commons Attribution License that allows others to share the work with an acknowledgment of the work’s authorship and initial publication in this journal.
3. Authors are able to enter into separate, additional contractual arrangements for the non-exclusive distribution of the journal’s published version of the work (e.g., post it to an institutional repository or publish it in a book), with an acknowledgment of its initial publication in this journal.
4. Authors are permitted and encouraged to post their work online (e.g., in institutional repositories or on their website) prior to and during the submission process, as it can lead to productive exchanges, as well as earlier and greater citation of published work.
5. The article and any associated published material is distributed under the Creative Commons Attribution-ShareAlike 4.0 International License
