碩士 / 國立臺灣科技大學 / 財務金融研究所 / 105 / Constructing an effective measure of the credit risk model to alert the business of a potential financial crisis has long been an important topic in the finance literature. Traditionally, most of the relevant research start to build the model from the financial ratio variables and ignore the non-financial variables which are more difficult to quantify. Moreover, the techniques of earning vanishing are becoming more and more innovative. They led to a series of endless check bounces, false accounts, breach of contract, hollowing out assets and fake transactions. These accidents resulted in the owner’s or major shareholders’ improper financial operation or illegal chaos, or an attempt to disguise the poor business. Therefore, in addition to focusing on financial information, investors should also pay more attention to "non-financial information" that can reveal more message. It is able to provide a full view of company operating and increase the probability of excavation of fraud.
This study introduces the KMV model with the advantages of market information and immediacy, and the establishment of default distance, default probability estimation and default warning mode. In view of the market-based KMV model with real-time detection and crisis warning characteristics, it reflects the company's financial distress in advance, in a more timely manner than credit rating. In the prediction of financial distress, it will not be overreacting or too conservative, and will not cause serious errors. Therefore, it is hoped that the KMV model is combined with the model based on traditional financial variables, and the characteristics of timely risk information will be combined with the credit model of traditional accounting financial variables. The validity of the model can be improved, and more credit information is made available to help enterprises to strengthen the credit risk management.
In this study, we establish a model based on the relationship between the ratios in the financial statements, and use the logistic regression method to address the company's solvency, profitability, operating efficiency, growth capability, cash flow, corporate governance variables and KMV model (DD) and so on for significance and forecast rate analysis, hoping to establish an effective financial advance warning mechanism. It is hoped that the mechanism can provide useful information for investors in public, government agencies or financial industry when they are in the assessment of financial decision-making or audit financial credit operations. Consequently, the corporate fraud messages can be identified effectively and the impact of false financial statements can be reduced.
It is found that when the default distance (DD) of KMV variable is added to the Logistic model, the default distance (DD) is not significant which also wakes shareholder return rate and cash flow insignificant. This shows that the default distance (DD) which can explain the part of the default originally has been explained by other factors in the model. As a result, it appears to the phenomenon why the ability of explanation rises rather than falls, and other variables become insignificant. Therefore, it is concluded that for the company's default forecast, the default distance (DD) of KMV variable does not add value to the prediction capability of a model based on financial variables and corporate governance variables. This study supports the combination of financial variables and corporate governance variables.
In order to understand whether the KMV model provides warning signal for financial crisis, we provide a case analysis. We selected three defaulted companies Wintek, Powerchip and Everskill companies from the empirical sample set, and their contrasting counterpart of similar size in the same industry, Giantplus, the VIS and Avy companies to proceeded with the case analysis and comparison. Our conclusion shows that the KMV model does not have the ability to alert the financial crisis in advance, and it cannot reflect the financial situation of the company before the default occurs.
| Identifer | oai:union.ndltd.org:TW/105NTUS5304056 |
| Date | January 2017 |
| Creators | Chia-Ling Su, 蘇佳翎 |
| Contributors | Wei-Chung Miao, 繆維中 |
| Source Sets | National Digital Library of Theses and Dissertations in Taiwan |
| Language | zh-TW |
| Detected Language | English |
| Type | 學位論文 ; thesis |
| Format | 172 |
Page generated in 0.0021 seconds