Klasifikasi Financial Distress Perusahaan Sektor Energi Di Bursa Efek Indonesia Menggunakan LASSO Gradient Boosting Dan LASSO Support Vector Machine

Asyiraf, Fauzan Dwi (2025) Klasifikasi Financial Distress Perusahaan Sektor Energi Di Bursa Efek Indonesia Menggunakan LASSO Gradient Boosting Dan LASSO Support Vector Machine. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 5006211018-Undergraduate_Thesis.pdf Download (3MB)

Abstract

Financial distress adalah kondisi keuangan tidak stabil yang terjadi sebelum kebangkrutan, disebabkan oleh kerugian signifikan, penurunan nilai aset, dan ketidakmampuan perusahaan dalam memenuhi kewajibannya. Penelitian ini bertujuan untuk menganalisis kondisi financial distress perusahaan sektor energi yang terdaftar di Bursa Efek Indonesia selama periode kuartal III tahun 2017 hingga kuartal III tahun 2023. Dengan meningkatnya minat investasi di sektor energi Indonesia dan tantangan yang dihadapi oleh beberapa perusahaan yang berisiko delisting, penting untuk mengidentifikasi indikator financial distress secara akurat. Penelitian ini menggunakan dua metode klasifikasi, yaitu LASSO Gradient Boosting (GBoost) dan LASSO Support Vector Machine (SVM), untuk mengevaluasi dan membandingkan efektivitas masing-masing metode dalam memprediksi kondisi financial distress. Penentuan status awal financial distress perusahaan ditinjau berdasarkan nilai laba bersih yang negatif selama dua periode berturut-turut. Data yang digunakan mencakup laporan keuangan kuartalan dari sembilan perusahaan sektor energi yang terdaftar di Bursa Efek Indonesia, dan dipilih secara acak menggunakan stratified random sampling. Penentuan hyperparameter optimal untuk setiap model dilakukan dengan melatih data menggunakan berbagai kombinasi hyperparameter yang telah ditentukan, proses ini dilakukan menggunakan bantuan GridSearchCV pada python. Nilai AUC dan akurasi kedua model selanjutnya dibandingkan untuk menentukan model klasifikasi terbaik. Hasil penelitian menunjukkan model LASSO GBoost lebih akurat dari model LASSO SVM dalam mengklasifikasikan kondisi financial distress perusahaan sektor energi yang terdaftar di Bursa Efek Indonesia. Nilai AUC dan akurasi yang dihasilkan model LASSO GBoost masing-masing sebesar 88,28% dan 94,12%, dengan variabel rasio keuangan yang berkontribusi terhadap model meliputi net profit margin, retained earning to total assets, operating profit margin, return on equity, total assets turnover, earning before interest to total assets, debt to assets, market value of equity to total liabilities, return on assets, cash ratio, working capital to total assets, working capital to sales, dan current ratio.
======================================================================================================================================
Financial distress is an unstable financial condition that occurs before bankruptcy, caused by significant losses, decreased asset values, and the company's inability to fulfill its obligations. This study aims to analyze the financial distress condition of energy sector companies listed on the Indonesia Stock Exchange during the period from the third quarter of 2017 to the third quarter of 2023. With increasing investment interest in Indonesia's energy sector and the challenges faced by some companies at risk of delisting, it is important to accurately identify financial distress indicators. This study uses two classification methods, namely LASSO Gradient Boosting (GBoost) and LASSO Support Vector Machine (SVM), to evaluate and compare the effectiveness of each method in predicting financial distress. Determination of the company's initial financial distress status is based on negative net income for two consecutive periods. The data used includes quarterly financial statements of nine energy sector companies listed on the Indonesia Stock Exchange, and randomly selected using stratified random sampling. Determining the optimal hyperparameters for each model is done by training the data using various combinations of hyperparameters that have been determined, this process is done using the help of GridSearchCV in python. The AUC and accuracy values of the two models were then compared to determine the best classification model. The results showed that the LASSO GBoost model was more accurate than the LASSO SVM model in classifying the financial distress condition of energy sector companies listed on the Indonesia Stock Exchange. The AUC value and accuracy produced by the LASSO GBoost model are 88,28% and 94,12%, respectively, with financial ratio variables that are contribute to the model including net profit margin, retained earnings to total assets, operating profit margin, return on equity, total assets turnover, earnings before interest to total assets, debt to assets, market value of equity to total liabilities, return on assets, cash ratio, working capital to total assets, working capital to sales, and current ratio.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Energy Sector, Financial Distress, LASSO-GBoost, LASSO-SVM, Financial Distress, LASSO-GBoost, LASSO-SVM, Sektor Energi
Subjects:	A General Works > AI Indexes (General) A General Works > AI Indexes (General) H Social Sciences > HD Industries. Land use. Labor > HD108 Classification (Theory. Method. Relation to other subjects ) H Social Sciences > HD Industries. Land use. Labor > HD30.23 Decision making. Business requirements analysis. H Social Sciences > HD Industries. Land use. Labor > HD61 Risk Management H Social Sciences > HG Finance > HG4028.V3 Valuation. Economic value H Social Sciences > HG Finance > HG4910 Investments H Social Sciences > HJ Public Finance
Divisions:	Faculty of Mathematics, Computation, and Data Science > Actuaria > 94203-(S1) Undergraduate Thesis
Depositing User:	Fauzan Dwi Asyiraf
Date Deposited:	28 Jul 2025 08:42
Last Modified:	28 Jul 2025 08:42
URI:	http://repository.its.ac.id/id/eprint/122229

Actions (login required)

View Item