The Research of Dynamic Financial Predicting Model－An Empirical Study of Securities Brokers

• Main Authors: Hung - Hui Lee, 李洪慧
• Other Authors: Hsiao - Ling Shen
• Format: Others
• Language: zh-TW
• Published: 1998
• Online Access: http://ndltd.ncl.edu.tw/handle/92014110455745073750

碩士 === 東吳大學 === 企業管理學系 === 86 === Detection of a firm’s slide toward failure at an early stage has long interested practitioners in the financial sector. Lots of models have been developed with this purpose, but they are static in nature and ignore useful information from the past financial condition. Therefore, it is necessary to develop business failure prediction models that assess the financial condition of firms sequentially over time. This thesis provides a dynamic financial prediction model and show how it can be used to predict a firm’s tendency towards failure. Because financial variables for firms are largely serially correlated, a time series methodology is adopted. The model is based on Theodossiou’s (1993) multivariate cumulative sum (CUSUM) and vector autoregressive moving average (Vector ARMA) frameworks. The sample consists of 63 securities brokers in Taiwan, 21distress and 42 non-distress in 1996. From previous studies, current ratio, leverage ratio, fixed assets to total assets, total expense to total revenue, and rate of net income before taxes were used in this study. At the beginning, the characteristics of distress securities brokers are compared with those of non-distress ones. Then utilize Vector ARMA to identify the random process of independent variables. Third, set up the dynamic financial predicting model by CUSUM model. Finally, demonstrate how the CUSUM model can be used to assess a firm’s financial condition and compared with Z model. The major results are as follows: 1.In contract to the healthy securities brokers, the distress ones have higher operation leverage risk, and their abilities of management, operation efficiency, profits are worse. 2.In this study, rate of net income before taxes is the most sensitive to the model, and current ratio is the worst. 3.Analyze the financial condition of three distress securities brokers and three healthy ones. It’s difficult for Z model to identify the securities brokers as potentially failing. But the CUSUM model would have classified three securities brokers as failing many months before their distress date. So CUSUM model is better than Z model.