Jundy, Muhammad Ghozi Al (2026) Penilaian Risiko Ekstrem Investasi Sektor Prioritas Danantara Sebelum, Selama, dan Sesudah Krisis 2020 Menggunakan Model GEV dan GPD. Other thesis, Institut Teknologi Sepuluh Nopember.
![[thumbnail of 5006221061_Undergraduate-Thesis.pdf]](http://repository.its.ac.id/style/images/fileicons/text.png) |
Text
5006221061_Undergraduate-Thesis.pdf - Accepted Version Restricted to Repository staff only Download (10MB) | Request a copy |
Abstract
Krisis pasar tahun 2020 akibat pandemi COVID-19 menunjukkan keterbatasan ukuran risiko tradisional seperti Value-at-Risk (VaR) dalam merepresentasikan kerugian ekstrem pada ekor distribusi return. Penelitian ini bertujuan untuk menilai risiko ekstrem serta ketahanan portofolio saham sektor prioritas di pasar modal Indonesia dengan menerapkan pendekatan Extreme Value Theory (EVT) melalui model Generalized Extreme Value (GEV) dan Generalized Pareto Distribution (GPD). Analisis dilakukan menggunakan data return harian yang dibagi ke dalam tiga sub-periode, yaitu pra-krisis, krisis, dan pasca-krisis, guna menangkap dinamika perubahan risiko antar fase pasar. Estimasi risiko dilakukan menggunakan VaR dan Conditional Value-at-Risk (CVaR) dengan pendekatan block maxima dan peaks over threshold, kemudian dievaluasi melalui uji Kupiec, uji kesesuaian distribusi, serta kriteria informasi AIC dan BIC. Hasil penelitian menunjukkan bahwa karakteristik risiko ekstrem berbeda antar aset dan periode, dengan peningkatan volatilitas dan risiko ekor yang lebih kuat selama krisis. Secara umum, model GPD memberikan kinerja yang lebih stabil dan konsisten dalam menangkap perilaku ekor distribusi dibandingkan GEV. Selanjutnya, hasil optimisasi mean–CVaR menunjukkan perbedaan struktur portofolio antara pendekatan EVT-based dan Empirical CVaR, di mana portofolio EVT-based cenderung lebih sensitif terhadap kejadian ekstrem dan menghasilkan konsentrasi aset yang lebih tinggi pada periode tertentu, sementara Empirical CVaR membentuk portofolio yang relatif lebih stabil dan defensif melalui distribusi bobot yang lebih merata. Hal ini mengindikasikan bahwa integrasi EVT, khususnya melalui model GEV, memberikan estimasi risiko ekstrem yang lebih representatif dan dapat mendukung perancangan portofolio yang lebih adaptif terhadap guncangan pasar ekstrem.
======================================================================================================================================
The market disruption caused by the COVID-19 pandemic in 2020 exposed the weaknesses of conventional risk measures, particularly Value at Risk, in capturing extreme losses located in the tail of return distributions. This study examines extreme risk exposure and portfolio resilience of priority sector stocks in the Indonesian capital market by employing Extreme Value Theory through the Generalized Extreme Value and Generalized Pareto Distribution models. Daily return data are analyzed across three market phases consisting of the pre crisis, crisis, and post crisis periods in order to observe changes in extreme risk characteristics under different market conditions. Risk is estimated using Value at Risk and Conditional Value at Risk based on the block maxima and peaks over threshold approaches, while model performance is evaluated through backtesting, distributional goodness of fit tests, and information criteria including AIC and BIC. The findings reveal considerable variation in extreme return behavior across assets and periods, with a sharp escalation in tail risk during the crisis phase. The results further indicate that the Generalized Pareto Distribution provides more stable and consistent tail risk estimates compared to the Generalized Extreme Value model. In addition, mean Conditional Value at Risk portfolio optimization produces distinct portfolio structures between the EVT based and empirical approaches, where EVT based portfolios tend to exhibit higher concentration and greater exposure to extreme market movements, while empirical Conditional Value at Risk portfolios show more defensive characteristics through broader diversification. Overall, the study highlights that the integration of Extreme Value Theory, particularly the Generalized Extreme Value, offers a more reliable framework for measuring extreme risk and supports the construction of portfolios that are more resilient to severe market shocks.
Actions (login required)
 |
View Item |