No Detail Found!

Assessment of credit losses based on arima- wavelet method

Author(s)
JAMIL J. JABER School of Mathematical Sciences, Faculty of Science and Technology,UniversitiKebangsaan Malaysia, Malaysia
NORISZURA ISMAIL School of Mathematical Sciences, Faculty of Science and Technology,UniversitiKebangsaan Malaysia, Malaysia
SITINORAFIDAHMOHD RAMLI School of Mathematical Sciences, Faculty of Science and Technology,UniversitiKebangsaan Malaysia, Malaysia
S. AL WADI School of Mathematical Sciences, Faculty of Science and Technology,UniversitiKebangsaan Malaysia, Malaysia
DALILA. BOUGHACI School of Mathematical Sciences, Faculty of Science and Technology,UniversitiKebangsaan Malaysia, Malaysia
Abstract
The aim of this paper is to estimate and forecast the loss-given defaults (LGD) using a sample data of credit portfolio loan collected from a bank in Jordan for the period up from January 2010 to December 2014. We use a wavelet-inspired analysis to convert the original observations into a time-scale domain. Then, we combine the wavelettransform with the ARIMA (Auto-Regressive Integrated Moving Average) model to get an ARIMA-WT new model to forecast the LGD data time series.We evaluate four wavelet functions, which are Haar (Haar), Daubechies (d4), least Asymmetric (La8), and Coiflet (C6). The numerical results show that the ARIMA-WT is more accurate than the pure ARIMA and the other considered ARIMAWavelet transform based models. We consider several metrics (MAPE, MASE, RMSE, AIC, AICs and BIC) to measure the performance of our proposed model. The combination between ARIMA-WT and La8 function improves highly the forecasting accuracy. According to our findings, we can say that the resulting forecast model is able to produce a high quality result.
Publication Details
Page(s) 1379-1392
DOI DOI not available
Published Journal: Journal of Theoretical and Applied Information Technology, Volume: 98, Issue: 9, Year: 2020

Request the full text from PASTIC

Keywords
LGD, CREDIT RISK, forecasting
0

Citations

0

Downloads

39

Views