Regar, Haniya Harum Pekerti (2026) Analisis Risiko Sistemik Saham Subsektor Logam dan Mineral dengan Conditional Value-at-Risk Pendekatan Quantile Regression Forest dan Stochastic Search Variable Selection (SSVS). Other thesis, Institut Teknologi Sepuluh Nopember.
![[thumbnail of 5003221002-Undergraduate_Thesis.pdf]](http://repository.its.ac.id/style/images/fileicons/text.png) |
Text
5003221002-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (15MB) | Request a copy |
Abstract
Subsektor logam dan mineral di Bursa Efek Indonesia memiliki volatilitas tinggi serta keterkaitan antaremiten yang kompleks, sehingga berpotensi menimbulkan risiko sistemik yang perlu diukur secara akurat. Penelitian ini bertujuan menganalisis karakteristik return saham, mengestimasi risiko individual menggunakan Value-at-Risk (VaR) dengan pendekatan AR-GARCH, mengukur risiko sistemik dengan Conditional Value-at-Risk (CoVaR) berbasis Quantile Regression Forest (QRF) serta ΔCoVaR, dan menerapkan Stochastic Search Variable Selection (SSVS) untuk menyeleksi prediktor yang signifikan. Data yang digunakan adalah return harian sepuluh saham subsektor logam dan mineral periode Agustus 2023 hingga Desember 2025. Hasil penelitian menunjukkan bahwa seluruh emiten memiliki distribusi return yang leptokurtik dengan ekor tebal, di mana AMMN dan BRMS tercatat sebagai emiten paling berisiko secara individual. Estimasi CoVaR berbasis QRF dan ΔCoVaR menunjukkan bahwa MDKA secara konsisten menjadi emiten dengan kedalaman risiko dan penularan risiko sistemik tertinggi, sedangkan INCO dan NCKL menunjukkan ketahanan relatif paling baik pada kedua ukuran risiko tersebut. Penerapan SSVS berhasil mengidentifikasi prediktor yang paling relevan secara konsisten, dengan ARCI dominan sebagai sumber transmisi risiko pada kondisi distress moderat (kuantil 5%) dan MDKA mengambil peran dominan pada kondisi ekstrem (kuantil 1%), mengindikasikan bahwa struktur transmisi risiko sistemik bersifat asimetris dan bergantung pada tingkat keparahan guncangan, meskipun validitas backtesting model CoVaR-SSVS belum sepenuhnya terpenuhi untuk seluruh emiten. Penelitian ini menegaskan bahwa kombinasi pendekatan QRF dan SSVS efektif untuk memetakan keterkaitan risiko sistemik yang bersifat non-linear dan kondisional di subsektor logam dan mineral.
=======================================================================================================================================
The metal and mineral subsector on the Indonesia Stock Exchange is characterized by high volatility and complex interdependencies among listed companies, potentially giving rise to systemic risk that needs to be measured accurately. This study aims to analyze the return characteristics of stocks, estimate individual risk using Value-at-Risk (VaR) with the AR-GARCH approach, measure systemic risk using Conditional Value-at-Risk (CoVaR) based on Quantile Regression Forest (QRF) and ΔCoVaR, and apply Stochastic Search Variable Selection (SSVS) to select significant predictors. The data used consists of daily returns of ten metal and mineral subsector stocks over the period August 2023 to December 2025. The results show that all stocks exhibit leptokurtic return distributions with heavy tails, with AMMN and BRMS identified as the most individually risky stocks. QRF-based CoVaR and ΔCoVaR estimation show that MDKA consistently exhibits the highest risk depth and risk spillover, while INCO and NCKL demonstrate the strongest resilience across both risk measures. The application of SSVS successfully identified the most relevant predictors consistently, with ARCI emerging as the dominant transmission source under moderate distress conditions (5% quantile) and MDKA assuming the dominant role under extreme conditions (1% quantile), indicating that the systemic risk transmission structure is asymmetric and dependent on shock severity, although the backtesting validity of the CoVaR-SSVS model was not fully satisfied across all stocks. This study confirms that the combined QRF and SSVS approach is effective in mapping non-linear and conditional systemic risk interconnections within the metal and mineral subsector.
| Item Type: | Thesis (Other) |
|---|---|
| Uncontrolled Keywords: | AR-GARCH, Conditional Value-At-Risk, Quantile Regression Forest, Risiko Sistemik, Stochastic Search Variable Selection, Systemic Risk. |
| Subjects: | H Social Sciences > HA Statistics H Social Sciences > HA Statistics > HA30.3 Time-series analysis H Social Sciences > HA Statistics > HA31.7 Estimation H Social Sciences > HG Finance H Social Sciences > HG Finance > HG4915 Stocks--Prices Q Science > QA Mathematics > QA274.2 Stochastic analysis Q Science > QA Mathematics > QA280 Box-Jenkins forecasting Q Science > QA Mathematics > QA402 System analysis. |
| Divisions: | Faculty of Science and Data Analytics (SCIENTICS) > Statistics > 49201-(S1) Undergraduate Thesis |
| Depositing User: | Haniya Harum Pekerti Regar |
| Date Deposited: | 27 Jul 2026 01:11 |
| Last Modified: | 27 Jul 2026 01:11 |
| URI: | http://repository.its.ac.id/id/eprint/137544 |
Actions (login required)
 |
View Item |