Loading...
Publication
Simulation Applications in Company Default Prediction
| Name: | Description: | Size: | Format: | |
|---|---|---|---|---|
| 3.24 MB | Adobe PDF |
Authors
Advisor(s)
Abstract(s)
This study applies a simulation methodology, Monte Carlo, to the field of corporate default
prediction, where its presence is only superficial. It attempts to augment a famous model
from a methodology already highly supported in traditional literature – the Z”-score of
Altman (1983), created through multiple discriminant analysis – and transform it
stochastically without the use of the highly complex intelligent models already available in
literature. A sample of 20 000 Portuguese companies from the Agriculture, Forestry,
Fishing, Mining and Construction sectors is analyzed, yielding results that support the Monte
Carlo method as a strong competitor for simple approaches like the logit transformation.
This helps to build the foundation for what may possibly be a path towards models easier to
apply in practice for the average Micro, Small and Medium enterprise (MSME) and beyond.
The evaluation to the model on this study also takes inspiration off the innovative points of
view of Mitton (2021) and Zhang (2022) to scrutinize results through empirical tests of the
model under a high number of parameter conditions, instead of relying heavily on statistical
significance, which is often overrepresented and overvalued in literature, and easily
manipulatable.
Description
Keywords
Simulation Default risk Prediction Portugal Monte Carlo Z”-score