BANKRUPTCY HAZARD MODEL ANALYSIS FOR FINANCIAL FIRM IN INDONESIA
Predict the probability of firm bankruptcy considering the change factor of firm is one of the evolving resarch study. Most commonly used probability prediction model is discerete-time forward hazard model (DFHM) based on historical data. Annual financial report data for bank and non-bank institutio...
Saved in:
| Main Author: | |
|---|---|
| Format: | Final Project |
| Language: | Indonesia |
| Online Access: | https://digilib.itb.ac.id/gdl/view/24655 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Institution: | Institut Teknologi Bandung |
| Language: | Indonesia |
| Summary: | Predict the probability of firm bankruptcy considering the change factor of firm is one of the evolving resarch study. Most commonly used probability prediction model is discerete-time forward hazard model (DFHM) based on historical data. Annual financial report data for bank and non-bank institution in Indonesia will be used to construct DFHM model. This final project proposes a combination DFHM with PCA for predicting the probablity of firm bankruptcy at time 0 and 1. To simplify data processing PCA is provided in reducing Altmans predictor variables into some principal components. The linier combination of principal components can be written as hazard function in the form of logistic regression. This regression coefficient estimated by maximum likelihood method. The focus of model validation is the comparation between foward default probability and the actual realized of default. In general the result of forward default probability for bank and nonbank institution is very small. In fact there is only one bankcrupt firm of nonbank institution in 2012. The result of forward default probability for each firm depend on the number observation used to construct DFHM model. |
|---|