Summary: | 碩士 === 銘傳大學 === 財務金融學系碩士班 === 99 === While prior studies have used the financial ratios, corporation governance and macroeconomic variables to predict the financial crises, little research examines the influence of the public news on the financial crises. Following Vega (2006) and Demers and Vega (2008), this study investigates the impact of news contents on the financial crises. The news contents included in this study are MEDIA, Net Optimism (NO), and Intensity of Default-Corpus (ITDC). To reduce the potential problem of outliers (which is not considered by previous research), this paper uses the robust logistic regression suggested by Akinson and Riani (2001) and Shen and Lin (2005) to analyzing the corporate distresses. The empirical results show that the ITDC extracted from the text mining of Chinese financial news will improve the prediction performance of the corporate default warning model.
|