Pemodelan Stokastik Transaksi Kliring Harian dan Prefund Kliring

Abstract

Bank Indonesia-Real Time Gross Settlement (BI-RTGS) system is very important in payment transactions. It has huge value and potential systemic risk due to the total value of the transactions. In 2019, the amount of daily RTGS transactions reached 3.5% of Indonesia's Gross Domestic Product. Since the new regulation on the clearing system was issued, BI has set a clearing prefund that is considered high for middle down class bank, it has an impact on banking management because not all bank has the ability to provide the prefund so that bank has to sacrifice potential profits from available cash funds. Meanwhile, there are idle funds that is caused by determination based on nonoutlier data with the lowest daily negative netting from the past year. It is necessary to build stochastic modeling about the characteristics of the clearing system to obtain an optimal prefund which has 1% and/or 5% settlement failure also the lowest potential loss prefund. Forecasting by Seasonal ARIMA helps this research in determining the prefund for the next year. The inventory model is considered to be similar to the clearing model with several changes and assumptions. The results of the stochastic process of daily clearing transactions show that the prefund value with the lowest potential loss is around the average daily clearing value and the prefund with a settlement failure of 1% and 5% has a lower value than the prefund set by BI. Keywords: clearing, modeling, prefund, stochastic, transaction.

Kehadiran sistem Bank Indonesia-Real Time Gross Settlement (BI-RTGS) sangat penting dalam transaksi pembayaran yang bernilai besar namun berpotensi risiko sistemik dikarenakan nilai total transaksinya. Di tahun 2019, total transaksi RTGS harian mencapai 3,5% dari Produk Domestik Bruto Indonesia. Sejak regulasi baru sistem kliring diterbitkan, BI menetapkan prefund kliring yang dianggap tinggi untuk bank kelas menengah ke bawah, hal ini berdampak pada manajemen perbankan karena tidak semua bank memiliki kemampuan untuk menyediakan prefund tersebut sehingga bank harus mengorbankan potensi keuntungan dari dana tunai tersedia. Sedangkan terdapat dana idle yang disebabkan penentuan berdasarkan data nonoutlier dengan netting negatif harian terendah selama setahun terakhir. Perlu pemodelan stokastik terhadap karakteristik sistem transaksi kliring untuk memperoleh prefund optimal yang memiliki kegagalan setelmen 1% dan/atau 5% serta prefund dengan potensi kerugian terendah. Peramalan data dengan menggunakan Seasonal ARIMA membantu dalam menentukan prefund di tahun berikutnya. Model inventori dianggap mirip dengan model transaksi kliring dengan beberapa perubahan dan asumsi baru. Dari proses stokastik transaksi kliring harian menunjukkan bahwa nilai prefund dengan potensi terendah berada di sekitar rata-rata nilai transaksi kliring harian dan prefund dengan kegagalan setelmen 1% dan 5% memiliki nilai yang jauh lebih rendah dari prefund yang ditetapkan oleh BI.