Enhanced predictive credit scoring model for customer's default detection based on new international financial reporting standard (IFRS9)
In July 2014, the International Accounting Standards Board (IASB) delivered a new directive on how to recognize and measure financial instruments as a continuous effort to increase financial stability across the globe. The new International Financial Reporting Standard or IFRS9 which includes requir...
Saved in:
| Main Author: | |
|---|---|
| Format: | Thesis |
| Language: | eng eng eng |
| Published: |
2021
|
| Subjects: | |
| Online Access: | https://etd.uum.edu.my/9519/1/depositpermission-not%20allow_s95131.pdf https://etd.uum.edu.my/9519/2/s95131_01.pdf https://etd.uum.edu.my/9519/3/s95131_02.pdf |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | In July 2014, the International Accounting Standards Board (IASB) delivered a new directive on how to recognize and measure financial instruments as a continuous effort to increase financial stability across the globe. The new International Financial Reporting Standard or IFRS9 which includes requirements for recognition and
measurement, impairment and general hedge accounting will supersede the older version of FRS139 in phases and mandatorily implemented at the periods beginning of 1st January 2018. However, the implementation of IFRS9 depends on the strategy
imposed by the respective bank to secure their business goal. Motivated by this standard, this study constructs a mathematical credit scoring model that aligns and adheres with this new impairment standard outlined by IASB by incorporating a
forward-looking expected credit loss from the initial origination date of financing. This proposed model simplifies and strengthens risk measurement and the reporting of financial instruments by considering the time value of money and cost amortization as required under IFRS9 guidelines. Then, the model computes the
probability of default based on credit risk of individual evaluation attributes and anticipates future credit risk deterioration with the use of available historical, current and forecasted macroeconomic variables. Empirical evidence recorded from the analysis performed on six years financial data shows promising results in respect to
low error rate in distinguishing between default and non-default creditors. In addition, statistical analyses conducted for model adequacy check, model goodness-of-fit test and model validation indicate that the model is adequate. The model is much reliable than the traditional risk profile model and is able to assist the financial institutions in identifying group of future creditors accurately. |
|---|