| Summary: | 碩士 === 國立臺灣大學 === 財務金融學研究所 === 94 === The purpose of this thesis is to empirically investigate which financial ratios should be used as the triggers of PCA for Taiwan. For the financial ratios examined, in addition to RBC ratios and non-performing loan ratio (NPL), I define a NPL-adjusted capital ratio (NCR), which contains information about the bank’s NPL, net worth, and reserves for NPL. I undertake two tests to find out which financial ratios may be the better triggers. First, I check which financial ratios can identify problem banks more precisely at the end of 2000, 2001, and 2002. Second, for the financial ratios that are good in predicting problem banks, I investigate the optimal thresholds that minimize total costs of misclassification.
The main results of the thesis can be summarized as follows.
1. Compared with the NPL ratio and NCR, the RBC ratio is much worse in distinguishing between healthy and troubled banks.
2. In general, the prediction precision of NCR and the NPL ratio are about the same. Which of them is better in terms of the trigger for PCA depends on the timing for when the prediction is made.
3. If the prediction of problem banks is made in the early stage of banking crisis (that is, in 2000), the prediction ability of NCR with higher LGD rate is better. This result may reflect the fact that, in the early stage of the banking crisis, NPL figures usually understate their non-performing loan problems. When a higher LGD is assumed, the bias caused by the underestimation of NPL is reduced by increasing the banks’ losses in assets.
4. Although I cannot determine the superiority of NCR and the NPL ratio from the empirical results, I conjecture that NCR will be a better trigger of PCA. Since NCR contains more information about a bank, it can reflect more aspects of banks’ financial problems. Besides, NCR is less subject to the manipulation problem because it is more difficult for banks to window dress NPL, net worth, and reserves for NPL simultaneously. Moreover, regardless of the source of a bank’s financial problems, the consequences of the problems will ultimately appear in the bank''s capital account. These arguments lead me to the conclusion that NCR should be a better trigger of PCA than NPL.
|
|---|
