Prediksi Financial Distress Perusahaan Asuransi Jiwa Konvensional di Indonesia Menggunakan Additive Generalized Extreme Value Regression

Siahaan, Cicilia Rumanat (2024) Prediksi Financial Distress Perusahaan Asuransi Jiwa Konvensional di Indonesia Menggunakan Additive Generalized Extreme Value Regression. Other thesis, Insititut Teknologi Sepuluh Nopember.

Text 5006201057_Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only until 1 October 2026. Download (4MB) | Request a copy

Abstract

Industri asuransi jiwa konvensional di Indonesia telah mengalami pertumbuhan pesat dalam beberapa tahun terakhir, seiring dengan perkembangan perekonomian nasional. Namun, di balik pertumbuhan tersebut, industri ini juga dihadapkan pada berbagai risiko, terutama risiko keuangan dan potensi terjadinya financial distress atau kesulitan keuangan yang dapat berakhir menjadi kebangkrutan perusahaan. Penelitian ini bertujuan untuk memitigasi risiko financial distress pada perusahaan asuransi jiwa konvensional di Indonesia dengan menggunakan metode Generalized Extreme Value Regression (GEVR) dan Generalized Extreme Value Additive Model (GEVAM). Kedua model ini dipilih untuk mempertimbangkan rare event yang menyebabkan ketidakseimbangan kelas, serta hubungan linear dan non-linear antara variabel prediktor dan respon. Analisis dilakukan dengan menggunakan skema full predictor dan seleksi variabel pada data yang outlier-nya ditangani dan tidak ditangani. Variabel yang digunakan meliputi rasio keuangan, ukuran perusahaan, dan indikator ekonomi makro dari analisis data laporan keuangan tahunan serta laman resmi Bank Indonesia dan Badan Pusat Statistik (BPS). Berdasarkan nilai akurasi dan Area Under Curve (AUC), model terbaik dengan skema full predictor adalah model GEVAM yang menggunakan data winsorization, sedangkan model terbaik dengan skema seleksi variabel backward elimination adalah model GEVAM yang juga menggunakan data winsorization. Secara keseluruhan, model terbaik adalah model GEVAM full predictor dengan data winsorization, dengan nilai akurasi dan AUC data testing secara berurut adalah 91,1% dan 97,9%. Rasio profitabilitas, khususnya Return on Equity (ROE), ditemukan sebagai variabel prediktor yang berpengaruh signifikan. Peningkatan nilai ROE akan menurunkan probabilitas perusahaan asuransi jiwa konvensional di Indonesia untuk mengalami financial distress.
=================================================================================================================================
The conventional life insurance industry in Indonesia has experienced rapid growth in recent years, in line with the development of the national economy. However, behind this growth, the industry is also faced with various risks, especially financial risks and the potential for financial distress or financial difficulties that can end up in company bankruptcy. This study aims to mitigate the risk of financial distress in conventional life insurance companies in Indonesia using the Generalized Extreme Value Regression (GEVR) and Generalized Extreme Value Additive Model (GEVAM) methods. These two models were chosen to consider rare events that cause class imbalance, as well as linear and non-linear relationships between predictor and response variables. The analysis was conducted using the full predictor scheme and variable selection on both treated and untreated outlier data. The variables used include financial ratios, company size, and macroeconomic indicators from data analysis of annual financial reports as well as the official websites of Bank Indonesia and the Central Bureau of Statistics (BPS). Based on the accuracy value and Area Under Curve (AUC), the best model with a full predictor scheme is the GEVAM model that uses winsorization data, while the best model with a backward elimination variable selection scheme is the GEVAM model that also uses winsorization data. Overall, the best model is the GEVAM full predictor model with winsorization data, with the accuracy and AUC values of the testing data being 91.1% and 97.9%, respectively. The profitability ratio, specifically Return on Equity (ROE), is found to be a significant predictor variable. An increase in ROE value will decrease the probability of conventional life insurance companies in Indonesia experiencing financial distress.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Asuransi Jiwa Konvensional, Financial Distress, Generalized Extreme Value Additive Model, Generalized Extreme Value Regression, Winsorization, Conventional Life Insurance
Subjects:	Q Science > QA Mathematics > QA278.2 Regression Analysis. Logistic regression
Divisions:	Faculty of Science and Data Analytics (SCIENTICS) > Actuaria > 94203-(S1) Undergraduate Thesis
Depositing User:	Cicilia Rumanat Siahaan
Date Deposited:	16 Jul 2024 03:52
Last Modified:	16 Jul 2024 03:52
URI:	http://repository.its.ac.id/id/eprint/108345

Actions (login required)

View Item