An Investigation of Bank Insolvencies: Can they be predicted before failure?
The banking system has been a backbone for most developed and emerging economies. It provides support to other industry and in turn to the economy as a whole. Over the past couple of years, the banking system has seen significant increase in bank failures these are results of either bad banking or b...
| Main Author: | |
|---|---|
| Format: | Dissertation (University of Nottingham only) |
| Language: | English |
| Published: |
2009
|
| Online Access: | https://eprints.nottingham.ac.uk/23408/ |
| _version_ | 1848792565942321152 |
|---|---|
| author | shankarram, Deeptha |
| author_facet | shankarram, Deeptha |
| author_sort | shankarram, Deeptha |
| building | Nottingham Research Data Repository |
| collection | Online Access |
| description | The banking system has been a backbone for most developed and emerging economies. It provides support to other industry and in turn to the economy as a whole. Over the past couple of years, the banking system has seen significant increase in bank failures these are results of either bad banking or bad management. Bank insolvencies have resulted in mandatory regulations and special attention by governments in order to maintain a stable economy. Early detection of problems in banks is clearly beneficial in order to prevent bank failures. This study attempts to predict bank failures in the U.S. between 2000-2009. It does this by testing two models (consisting of different sets of variables) on a sample of failed banks, bailed banks and a combination of both for up to three years before failure. Binary logistic regression is the statistical analytical tool used to examine the predictive ability of a continuous/categorical variable on a categorical variable (this variable has only two possible numbers: 0 for a non-failed bank and 1 for a failed bank). After testing the predictive power of both models, it was observed that the predictive ability of model 2 was greater than that of model one for the three samples and over the three year period. It was also observed that the predictive power of the model 2 diminished slightly from the first year before failure to the third year before failure. This study has shown that it is possible to observe some symptoms of the possibility of failure as early as three years before possible failure providing a tool for investors, managers and regulators to assess the health of banks early enough for corrective action to be taken. The variables representing the size of the bank in terms of its total assets, total deposits, profits, etc indicating that larger and more profitable banks are less likely to fail or be allowed to fail whatever the case might be. It is suggested that these variables be included in subsequent models attempting to predict the probability of bank failures. |
| first_indexed | 2025-11-14T18:46:26Z |
| format | Dissertation (University of Nottingham only) |
| id | nottingham-23408 |
| institution | University of Nottingham Malaysia Campus |
| institution_category | Local University |
| language | English |
| last_indexed | 2025-11-14T18:46:26Z |
| publishDate | 2009 |
| recordtype | eprints |
| repository_type | Digital Repository |
| spelling | nottingham-234082018-02-16T05:59:58Z https://eprints.nottingham.ac.uk/23408/ An Investigation of Bank Insolvencies: Can they be predicted before failure? shankarram, Deeptha The banking system has been a backbone for most developed and emerging economies. It provides support to other industry and in turn to the economy as a whole. Over the past couple of years, the banking system has seen significant increase in bank failures these are results of either bad banking or bad management. Bank insolvencies have resulted in mandatory regulations and special attention by governments in order to maintain a stable economy. Early detection of problems in banks is clearly beneficial in order to prevent bank failures. This study attempts to predict bank failures in the U.S. between 2000-2009. It does this by testing two models (consisting of different sets of variables) on a sample of failed banks, bailed banks and a combination of both for up to three years before failure. Binary logistic regression is the statistical analytical tool used to examine the predictive ability of a continuous/categorical variable on a categorical variable (this variable has only two possible numbers: 0 for a non-failed bank and 1 for a failed bank). After testing the predictive power of both models, it was observed that the predictive ability of model 2 was greater than that of model one for the three samples and over the three year period. It was also observed that the predictive power of the model 2 diminished slightly from the first year before failure to the third year before failure. This study has shown that it is possible to observe some symptoms of the possibility of failure as early as three years before possible failure providing a tool for investors, managers and regulators to assess the health of banks early enough for corrective action to be taken. The variables representing the size of the bank in terms of its total assets, total deposits, profits, etc indicating that larger and more profitable banks are less likely to fail or be allowed to fail whatever the case might be. It is suggested that these variables be included in subsequent models attempting to predict the probability of bank failures. 2009-10-02 Dissertation (University of Nottingham only) NonPeerReviewed application/pdf en https://eprints.nottingham.ac.uk/23408/1/An_Investigation_of_Bank_Insolvencies_Can_they_be_predicted_before_failure_by_Deeptha_Shankarram.pdf shankarram, Deeptha (2009) An Investigation of Bank Insolvencies: Can they be predicted before failure? [Dissertation (University of Nottingham only)] (Unpublished) |
| spellingShingle | shankarram, Deeptha An Investigation of Bank Insolvencies: Can they be predicted before failure? |
| title | An Investigation of Bank Insolvencies: Can they be predicted before failure? |
| title_full | An Investigation of Bank Insolvencies: Can they be predicted before failure? |
| title_fullStr | An Investigation of Bank Insolvencies: Can they be predicted before failure? |
| title_full_unstemmed | An Investigation of Bank Insolvencies: Can they be predicted before failure? |
| title_short | An Investigation of Bank Insolvencies: Can they be predicted before failure? |
| title_sort | investigation of bank insolvencies: can they be predicted before failure? |
| url | https://eprints.nottingham.ac.uk/23408/ |