中国市场可转债定价研究

发布时间：2018-01-13 21:09

本文关键词：中国市场可转债定价研究　出处：《浙江大学》2013年博士论文　论文类型：学位论文

【摘要】：可转债是一种融合了股票和债券特征的复杂金融衍生品,发行人通过不同条款的组合来适应自身的融资需求,同时为投资者提供具有差别化的产品。这种高度的灵活性在丰富资本市场的同时,也给合理定价提出了更大的挑战。目前,还没有一个统一的定价模型能够完整地包含所有可能的条款假设。因此,在具体的定价实践中,我们需要细致地审定目标转债的市场特征,对涉及条款作出必要的取舍,突出具有结果显著性的产品结构,并在此基础上选择合理的模型假设。本文基于这样的思路对中国市场上的可转债定价进行研究。我们认为,我国可转债条款最重要的特征是赎回、回售具有一定的强制性,且赎回、回售触发条件具有路径依赖性。由第一个特征,本文将可转债定价问题从一个投资者、发行人的双人博弈过程转化为投资者在发行人策略确定情况下的最优决策问题；而第二个特征是本文与国内其他可转债定价文章在模型假设上的主要区别,我们通过具体的实证数据表明,忽略赎回、回售触发条件的路径依赖性会使理论结果显著偏低。在模型实现方面,本文采用的方法更偏向于我国金融市场的实际情况。鉴于国内市场上没有可参考的隐含波动率数据,本文对实际测度下的波动率过程利用GARCH方法建模,并通过风险的市场价格参数将其与风险中性测度下的股票价格动态过程关联起来,从而使得条件波动率的估计值变为一个由风险中性测度下的布朗运动驱动的适应过程。区别于国内其他可转债定价文章将波动率看成是固定常数的做法,本文的建模方式能够更好地体现波动率集聚的特征,并反映市场近期的价格特征。本文利用约化模型方法对可转债的信用风险进行建模,考虑到我国信用衍生品市场的发展现状,我们通过同级企业债到期收益曲线与国债到期收益率曲线剥离出可违约零息债,从而确定违约强度的估计值。通过适当简化的约化模型方法,本文将带违约风险的可转债模型与无违约风险的可转债模型很好地统一在一个定价框架中。本文还考虑了当无风险利率随机时对定价结果的影响。我们利用Ait-Sahalia(1996)提出的非参数化方法对CIR模型的参数进行估计,并比较在不同初始股价以及不同股价、利率相关系数下,利率随机化对定价的影响,数据结果显示,在我国的金融市场环境下,这种影响并不显著。最后,我们对国内一小部分流动性较好的可转债进行了定价效果检验,结果表明当股票价格在转股价附近变动时,模型理论价值与市场价格有很高的拟合度。
[Abstract]:Convertible bond is a kind of complex financial derivatives which combines the characteristics of stocks and bonds. Issuers adapt to their own financing needs through the combination of different terms. This high flexibility not only enriches capital markets but also poses greater challenges to reasonable pricing. At present, there is not a unified pricing model that can fully include all possible clause assumptions. Therefore, in the specific pricing practice, we need to carefully examine the market characteristics of target debt conversion. To make the necessary choice of the terms involved, highlight the product structure with significant results, and on the basis of this, select reasonable model assumptions. Based on this idea, this paper studies the pricing of convertible bonds in Chinese market. We believe that the most important feature of the terms of convertible bonds in China is redemption, and the repurchase is mandatory, and the trigger conditions of redemption and resale are path-dependent. In this paper, the pricing problem of convertible bonds is transformed from the two-game process of an investor and issuer to the optimal decision of the investor under the condition of the decision of the issuer's strategy. The second feature is the main differences between this paper and other domestic convertible bond pricing articles in the model hypothesis, we show through the specific empirical data, ignore redemption. The path-dependence of the trigger condition makes the theoretical results significantly lower. In the realization of the model, the method adopted in this paper is more inclined to the actual situation of our financial market. In view of the fact that there is no implied volatility data for reference in the domestic market, this paper uses GARCH method to model the volatility process under the actual measurement. And through the risk of the market price parameters will be linked to the risk neutral measure of the stock price dynamic process. Therefore, the estimate of conditional volatility becomes an adaptive process driven by Brownian motion under risk-neutral measure. Different from other domestic convertible bond pricing articles, volatility is regarded as a fixed constant. The modeling method of this paper can better reflect the characteristics of volatility agglomeration and the price characteristics of the market in the near future. In this paper, the credit risk of convertible bonds is modeled by reduction model, considering the development of credit derivatives market in China. Through the curve of maturity yield of enterprise bond and the curve of maturity yield of national debt, we can get the estimated value of defaultable zero interest debt by means of the curve of maturity yield of enterprise bonds of the same class, and then we can make sure the estimated value of default intensity. In this paper, the convertible bond model with default risk and the convertible bond model without default risk are well unified in a pricing framework. In this paper, we also consider the influence of risk-free interest rate on pricing results. We use the non-parametric method proposed by Ait-Sahalia 1996 to estimate the parameters of CIR model. The effects of interest rate randomization on pricing under different initial stock price and different stock price and interest rate correlation coefficient are compared. The results show that this effect is not significant in the financial market environment of our country. Finally, we test the pricing effect of a small number of domestic convertible bonds with good liquidity. The results show that the theoretical value of the model has a high fit with the market price when the stock price changes near the stock price.
【学位授予单位】：浙江大学
【学位级别】：博士
【学位授予年份】：2013
【分类号】：F224;F832.51

【参考文献】