Penelitian ini bertujuan untuk mencari tahu pengaruh non-linear dari hubungan capital buffer dan risk-taking behaviour bank di Indonesia. Risk-taking behaviour diukur dengan menggunakan rasio non-performing loan terhadap gross loan (NPL). Diperoleh 101 sampel bank di Indonesia dari tahun 2013 hingga 2017. Peneliti menggunakan regresi kuantil data panel untuk menggambarkan hubungan non-linear capital buffer pada risk-taking behaviour bank. Penelitian juga mengelompokkan sampel bank menjadi kelompok bank besar dan kelompok bank kecil berdasarkan ukuran permodalannya untuk menunjukkan pentingnya memperhatikan perbedaan karakteristik bank dalam menganalisis hubungan capital buffer pada risk-taking behaviour bank. Ditemukan bahwa terdapat terhadap hubungan non-linear capital buffer terhadap risk-taking behaviour bank di Indonesia. Peningkatan capital buffer akan menurunkan risk-taking behaviour bank sampai pada suatu titik (turning points), penambahan buffer setelah titik tersebut akan cenderung mendorong bank untuk mengambil risk-taking berlebih. Ditemukan pula bahwa high-risk bank atau kelompok bank besar lebih sensitif terhadap perubahan efek capital buffer dibanding low-risk bank.

---

This study aims to discover the non-linear effect of the relationship between capital buffer and bank risk-taking behaviour in Indonesia. Risk-taking behaviour is measured using the ratio of non-performing loan to gross loans (NPL). 101 bank samples were obtained in Indonesia from 2013 to 2017. The researcher uses panel data quantile regression to capture the non-linear relationship of capital buffer on bank risk-taking behaviour. The study also grouped bank samples into large bank and small bank groups based on their capital size to show the importance of considering different bank characteristics in analysing the relationship of capital buffer on bank risk-taking behaviour. The study found that there is a non-linear relationship between capital buffer and bank risk-taking behaviour in Indonesia. An increase in capital buffer will, at first, reduce bank risk-taking behaviour to a certain point (turning points), whilst adding buffer after that point will encourage bank to take excessive risk-taking. It was also found that high-risk banks or large bank groups are more sensitive to changes in the capital buffer than low-risk banks.