在Hull-White 隨機利率下信用風險之衡量-運用創新的數值方法DFPM-HWT

Abstract

本文提出創新的二維度立體樹狀結構評價模型DFPM-HWT，當公司資產與利率服從相依的隨機變動過程下，此模型可評價出公司債券的價值，也可調整其數值結構來配適具有隨機性質的違約門檻和公司價值具有離散跳躍等性質，以解決數狀結構常有的非線性誤差。此評價模型不僅可以衡量信用風險，更可以對具有美式性質 (American-style features) 的公司債做評價。Briys and Varenne (1997)結構式評價模型(對有違約風險的債券定價) 和Hull-White (1990) 利率延伸性商品評價模型(對可贖回買權定價) 都可視為DFPM-HWT評價模型的特例。本文利用敏感度分析觀察贖回價格與債券利差之間的關係，並加以解釋與說明。

This paper suggests a novel two-dimensional lattice tree model “DFPM-HWT” that can evaluate the corporate bond when firm value and interest rates follow correlated stochastic processes. This lattice model can adjust its structure to fit stochastic default barriers and discrete jumps of the firm value to alleviate the oscillation problem. It can not only measure the credit risk, but price some American-style features like callable, putable bonds. Both Briys and Varenne (1997) evaluation model (for pricing defaultable bonds) and Hull and White (1990) evaluation model (for pricing callable call option) can be regarded as special cases of DFPM-HWT. Detail sensitive analysis are given to illustrate the relations between credit spreads, redemption premium, and etc.