Purba, Sumaniata Angelica (2026) Analisis Risiko Sistemik Saham Perusahaan Sektor Teknologi Menggunakan Conditional Value-at-Risk Berbasis AR-GARCH dan QRNN dengan Stochastics Search Variable Selection. Other thesis, Institut Teknologi Sepuluh Nopember.
![[thumbnail of 5003221064-Undergraduate_Thesis.pdf]](http://repository.its.ac.id/style/images/fileicons/text.png) |
Text
5003221064-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (10MB) | Request a copy |
Abstract
Sektor teknologi di Bursa Efek Indonesia (BEI) mengalami pertumbuhan pesat seiring meningkatnya pemanfaatan teknologi digital pascapandemi COVID-19. Di sisi lain, saham sektor ini memiliki volatilitas yang tinggi sehingga berpotensi menimbulkan risiko sistemik melalui efek penularan antarperusahaan. Penelitian ini bertujuan untuk mendeskripsikan karakteristik return saham sektor teknologi, mengukur risiko individual menggunakan Value-at-Risk (VaR), mengukur risiko sistemik menggunakan Conditional Value-at-Risk (CoVaR) dan ΔCoVaR, serta mengidentifikasi variabel prediktor yang berpengaruh dalam pemodelan CoVaR menggunakan Stochastic Search Variable Selection (SSVS). Data yang digunakan adalah harga penutupan harian dari delapan emiten sektor teknologi dengan market share tertinggi sebesar 95%, yaitu DCII, MLPT, EMTK, GOTO, BELI, BUKA, WIFI, dan CYBR, yang diperoleh dari Yahoo Finance dengan periode pengamatan 8 Agustus 2023 hingga 30 Desember 2025. Hasil penelitian menunjukkan bahwa return harian seluruh saham bersifat heterogen, leptokurtik, dan mengandung heteroskedastisitas sehingga memerlukan pendekatan pemodelan risiko yang adaptif. Estimasi VaR berbasis AR-GARCH dinyatakan valid pada tujuh dari delapan saham berdasarkan Kupiec Test, sedangkan estimasi CoVaR dan ΔCoVaR berbasis regresi kuantil maupun Quantile Regression Neural Network (QRNN) valid pada seluruh saham. Risiko sistemik antar emiten bersifat heterogen dan cenderung meningkat pada kondisi pasar yang lebih ekstrem. Saham MLPT, DCII, WIFI, dan CYBR menunjukkan kontribusi risiko sistemik yang relatif lebih tinggi, sedangkan BELI secara konsisten memiliki kontribusi terendah. Implementasi SSVS menghasilkan model yang lebih parsimonious. Secara keseluruhan, model CoVaR QRNN dan CoVaR QRNN - SSVS menunjukkan validitas yang paling konsisten, sedangkan CoVaR QRNN memberikan performa terbaik berdasarkan kombinasi Kupiec Test dan Proportion of Shortfall. Dengan demikian, kombinasi AR-GARCH, QRNN, dan SSVS efektif dalam mengukur risiko individual dan risiko sistemik pada saham sektor teknologi di Bursa Efek Indonesia.
===============================================================================================================================
The technology sector on the Indonesia Stock Exchange (IDX) has experienced rapid growth following the increasing adoption of digital technology in the post COVID-19 pandemic period. However, technology stocks are characterized by high volatility, which may give rise to systemic risk through contagion effects across firms. This study aims to describe the characteristics of technology stock returns, measure individual risk using Value-at-Risk (VaR), measure systemic risk using Conditional Value-at-Risk (CoVaR) and ΔCoVaR, and identify significant predictor variables in the CoVaR model using Stochastic Search Variable Selection (SSVS). The data consist of the daily closing prices of the eight largest technology companies representing 95% of the sector's market share, namely DCII, MLPT, EMTK, GOTO, BELI, BUKA, WIFI, and CYBR, obtained from Yahoo Finance over the period from August 8, 2023, to December 30, 2025. The results indicate that the daily returns of all stocks are heterogeneous, leptokurtic, and exhibit heteroscedasticity, highlighting the need for an adaptive risk modeling approach. The AR-GARCH-based VaR model is validated for seven of the eight stocks based on the Kupiec Test, whereas the quantile regression-based and Quantile Regression Neural Network (QRNN)-based CoVaR and ΔCoVaR models are validated for all stocks. Systemic risk varies across firms and tends to increase under more extreme market conditions. MLPT, DCII, WIFI, and CYBR exhibit relatively higher contributions to systemic risk, while BELI consistently shows the lowest contribution. The implementation of SSVS produces a more parsimonious model. Overall, the QRNN-based CoVaR and QRNN-SSVS-based CoVaR models demonstrate the most consistent validity, while the QRNN-based CoVaR model achieves the best performance based on the combined results of the Kupiec Test and the Proportion of Shortfall. Therefore, the integration of AR-GARCH, QRNN, and SSVS provides an effective approach for measuring both individual risk and systemic risk in technology stocks listed on the Indonesia Stock Exchange.
| Item Type: | Thesis (Other) |
|---|---|
| Uncontrolled Keywords: | AR-GARCH, CoVaR, QRNN, SSVS, Sektor Teknologi, AR-GARCH, CoVaR, QRNN, SSVS, Technology Sector |
| Subjects: | H Social Sciences > HA Statistics > HA30.3 Time-series analysis H Social Sciences > HG Finance H Social Sciences > HG Finance > HG4529 Investment analysis H Social Sciences > HG Finance > HG4915 Stocks--Prices H Social Sciences > HG Finance > HG8054.5 Risk (Insurance) |
| Divisions: | Faculty of Science and Data Analytics (SCIENTICS) > Statistics > 49201-(S1) Undergraduate Thesis |
| Depositing User: | Sumaniata Angelica Purba |
| Date Deposited: | 27 Jul 2026 06:43 |
| Last Modified: | 27 Jul 2026 06:43 |
| URI: | http://repository.its.ac.id/id/eprint/137817 |
Actions (login required)
 |
View Item |