在交易雜訊下估計具違約邊界之結構化信用風險模型

Abstract

在2007年的全球金融風暴過後，不僅學術界對企業的違約風險非常的重視，實務界亦然，因此，如何能夠更準確的預測企業違約風險成為一個重要的研究課題。本篇研究根據 Duan and Fulop (2009) 所提出的平滑局部化取樣/重要性重新取樣粒子濾波器(smoothed localized sampling/importance resampling particle filter)架構去處理在有交易雜訊(trading noise)下之結構式模型估計。我們的模型在障礙選擇權的架構下以結構式方法進行公司有價證券訂價，本研究結果指出交易雜訊在流動性差的股票上會有顯著的影響，而且可能對於波動度與破產機率的估計產生影響。

After the worldwide financial crisis in 2007, credit risk of a company is getting vast attention not only from academic but also from practitioners. It is of interest for researchers to more accurately model and estimate the default risk of a firm. In this paper, we apply the method proposed by Duan and Fulop (2009), the smoothed localized sampling/importance resampling (SL-SIR) particle filter, to deal with the structural model estimation in the presence of trading noise. Our model employs the structural approach for valuing corporate securities under the barrier option framework. Our results suggest that trading noise can be substantial for the less liquid stocks and may potentially affect volatility and default probability estimation.