公司債殖利率指標正確性的探討

Abstract

本研究以具馬可夫過程(Markov process)的簡化模型(reduced form model)，檢驗臺灣經濟新報中取得的公司債殖利率指標之正確性。本研究主要是依據Jarrow及Turnbull在1995年發表的簡化模型，其假設破產過程為一服從離散的公司信用評等的轉變，而此信用評等的轉變為一馬可夫鏈。此模型最大好處是可以直接使用可觀察到的資料，例如透過違約機率的計算以求得信用風險債券的價格，無需計算公司資產。本研究使用臺灣經濟新報中過去六年臺灣上市上櫃公司的信用評等改變資料及由臺灣經濟新報得到公司債殖利率指標，導入簡化模型中，可以預測未來公司債的殖利率指標。本研究稱之為殖利率指標理論值，再以此理論值與由臺灣經濟新報中取得的公司債殖利率指標比較，討論是否有產生錯詁的可能。實證分析之後發現實際殖利率指標皆高於理論值，推測有兩個原因：一、模型假設與資料使用的限制；二、經濟因素。 經由實證研究後，在現今的臺灣債券市場中，本研究雖不認為JLT模型可以直接應用於預測上，但冀望未來當市場漸趨成熟，而相關資料也較完整時，JLT模型會對臺灣公司債市場的變化，會有更佳的預測能力。

This article provides a Markov process for the reduced form model to test the accuracy of the risky bond yield index gotten from TEJ (Taiwan Economic Journal) data base. The model is based on Jarrow and Turnbull (1995), with the bankruptcy process following a discrete state space Markov chain in credit ratings. The parameters of this model are easily estimated by observable data. For example, through calculating the probability of default, the risky bond price could be applied without evaluating the company’s asset. This article uses two kinds of data from TEJ data base. One is the credit rating change data of Taiwan Publicly traded companies in the past six years, and the other is risky bond yield index. Using these data in reduced form model, we get the theoretical risky bond yield index. To ompare the difference between actual and theoretical risky bond yield index, we discuss the probability of the misestimating. After the real diagnosis analysis, the actual risky bond yield index is found to be larger than theoretical index. Two master factors are predicted: First, the constraints from the hypothesis of model and data source and second, some economic factors.