本論文目的是探討合成型抵押債權受益憑證的發行流程以及分散績分可造成的效果。首先，我們介紹合成型抵押債權受益憑證的演進與其發展。其次，簡述抵押債權受益憑證的評等過程和各種評等模型。接者，我們訂立信用違約交換的挑選準則與本息償還優先順序去建構一個實際的合成型抵押債權受益憑證-絲路。最後，我們模擬分散績分在遞增的情境下，各階層受益憑證所面臨的損益變化。除此之外，我們也使用KMV模型先求算出每家樣本公司的違約機率，再進一步加入評等因子與歐德曼的Z績分，最後找出與信用價差的相關性高低依序為評等因子、違約機率、歐德曼的Z績分。 The purpose of this article is to describe the structuring process of Synthetic Collateralized Debt Obligations (synthetic CDOs) and diversity score effect. We first introduce the evolution of synthetic CDOs and their growth. Following this, we have a short discussion of rating process and rating methodologies. Next, we set up the filter rules of credit default swaps and the “waterfall” to construct a practical synthetic CDO, Silk Road. Finally, under different diversity score scenarios, we explore the influence of the diversity score on tranches’ performance. Additionally, we use KMV model to calculate probabilities of default for our sample and in that case, the sequence of correlation with credit spread is rating factor, probability of default, and then Altman’s Z-score.
