| As an essential part of China’s capital market, bond market plays an important role in economy development. Since the establishment of inter bank bond market in 2002, the custody of the bond market in China grew by 10 times, the amount of the issue grew by 7 times, the volume of transactions grew by 17 times. According to the statistics on the bond market analysis report of Central Clearing Corp in 2014, the total amount of the bond market in China reached 35.64 billions, the total number of bonds is 11075.In 2014 that year to complete the issuance of 11.87 billions, the issue of bond number is 6773. With the development of the interest rate market, scientific management of the risk of bonds become the most important subject.In recent years, quantitative investment is accepted by the capital market, and the investment strategy based on quantitative trading is widely used in the stock market and futures market. In this paper, we focus on the pricing of bond and pricing error of bond portfolio transaction. According to the conclusion of this paper, we put forward policy recommendations for the current problems of China’s bond market. Based on this paper, this paper is divided into six chapters:Chapter 1 is the introduction. This chapter mainly introduces the background and significance of this study, reviews the main research object of the research status and literature review, and finally summarizes the innovations and shortcomings of this paper.Chapter 2 is Inter-bank bond pricing based on interest rate term structure models. This chapter is divided into five sections. Section 1 briefly reviews the development process of China’s bond market and to analyze the current development. Taking into account the current status of the bond market segmentation, and more than 97% of the inter-bank bond market trading accounts for the entire bond market trading volume, so this paper mainly use inter-bank bonds as research subjects; Section 2 and Section 3, describes static interest rate term structure model and dynamic term structure model, and analyze the advantages and disadvantages of each model; Section 4, Use McCulloch cubic spline model, Svensson model, Vasicek model and Cox-Ingersoll-Ross model estimated theoretical price of inter-bank bonds, compare pricing power by four models, the Vasicek model is good at pricing Inter-bank bond. So in this paper, the Vasicek model is chosed to be the theoretical model as a reference model pricing model.In order to influence the choice of bonds verify the validity of the pricing error and model error in the pricing of bonds,we use Bliss (1997) method, the McCulloch cubic spline model is used to be the premise of the reference model, by the pricing error is divided into three intervals two models to calculate the estimate of the probability of falling into the same debt pricing error interval, eventually found Svensson model, Vasicek model and CIR model and cubic spline model to estimate the debt pricing error occurred while falling into the same interval probabilities were 64% 49% and 57%, although this result shows that the model selection does affect the pricing of bonds, but also reflects the national debt market effective information contained pricing errors, transaction-based pricing strategy will get error based on this information to build a certain excess returns; Section 5 is the summarition of this chapter.Chapter 3 is the Exchange corporate bond pricing research based on structural model. This chapter is divided into four sections.Section 1 briefly describes different varieties of default bond, taking into account the strong degree of market exchanges in corporate bonds, trading relatively frequent, so this paper mainly corporate bonds Exchange as a main object of study; Section 2 describes the structure of the model and simple model to analyze the differences and advantages and disadvantages of each model; Section 3, the choice of structural model estimation theory herein Exchange price of corporate bonds, including Merton model, Longstaff-Schwartz model and Leland-Toft model estimation theory Exchange corporate bond prices. The empirical results show that with the introduction of stochastic interest rates and default boundary structure model of corporate bond pricing power has increased, which Leland-Toft model for pricing of corporate bonds strongest in three models, this paper select the model as a reference model for pricing theory. Overall the structural model of the high-grade bond pricing ability, while low-grade bond pricing power is weak. Explanation structure model risk factors do not fully explain the composition of the corporate bond spreads, its causes may be due to the corporate bond market of the Exchange bond market compared to developed countries, caused by lack of liquidity. Select the company in order to verify the validity of the debt pricing error and a model of the impact on the corporate bond pricing errors, and bonds similar paper, Bliss (1997) approach to Merton model seen as the premise of the reference model, by the pricing error points three intervals to calculate the two models of corporate bond pricing error is estimated to fall in the same range of probability, eventually found Longstaff-Schwartz model and Leland-Toft model and Merton pricing model to estimate the company’s debt to fall while the probability of errors occurring in the same range They were 49% and 38%. Although this result shows that the model selection does affect bond pricing, but also reflects the market of corporate bonds effective information contained pricing errors; Section 4 of this chapter summarized.Chapter 4 is Liquidity effect impact on the bond pricing. This chapter is divided into five:Section 1 introduces the definition and characteristics of liquidity, although currently there is no theoretically correct measure of liquidity, but overall liquidity consists of four basic characteristics:market breadth, market depth, timeliness and flexibility; Section 2 introduces several major measure of liquidity, and to summarize the advantages and disadvantages of these methods; Section 3, Elton and Green (1998) research methods to study the mobility Effect Bond Pricing between banks, the October 1,2008 to October 1, 2014 is divided into three observation period:a subsample of the observation period is October 1,2008 to October 1,2010 During the two sub-samples for observation January 1, 2011 to January 1,2013, during the observation of three sub-samples is March 1,2013 to October 1,2014. Empirical studies have shown that the degree of impact on the liquidity bank bonds priced during the three sub-samples were 40bp,73bp and 60bp. Except for all the other factors that can affect the value of Treasury bonds basic pricing, the pricing of the debt contribution liquidity effect was 91.91%, respectively, during the three sub-samples,95.19% and 97.08%. This shows that our country is currently in the bond market, all factors that affect the pricing of bonds, liquidity is the most important effect of the pricing or even the only determining factor, and a share of 91.91% from sample to sample three accounted for 97.08% of the high, also reflects the development of China’s bond market as the impact of the liquidity effect of increasing national debt pricing; Section 4, from Amihud illiquid metric method, bid-ask spread and turnover three angles studied the effect of liquidity Effect of Exchange corporate bonds pricing, empirical results show that the contribution of the liquidity effect on corporate bonds priced at 39bp to 64bp. Wherein Amihud illiquidity measure approach to corporate debt risk premium and significant explanatory power in three of the strongest measure of liquidity, so the current illiquidity measure Amihud method is more suitable to measure liquidity effect of the Exchange corporate bonds. This paper compares the company’s debt further liquidity variable characteristic variable, market variables, the degree of credit risk affect the risk premium on corporate bonds, through the analysis of four types of variable risk premium on corporate bonds discovery liquidity variable degree of influence on corporate debt risk premium explain the strongest and most significant; Section 5 of this chapter summarized.Chapter 5 is Strategies bond portfolio transaction-based pricing error mean reversion principle. This chapter is divided into four:Section 1 briefly describes the status of development and the current definition of quantitative trading transactions of quantification; Section 2, according to the second chapter empirical research on inter-bank bonds pricing, based on regression to the mean pricing error principle to build trading strategies using Nettekoven (2005) method, the inter-bank bonds were constructed transaction pricing error signal based on the moving average model and the time series model. By analyzing the information Treasury trading week 1 January 2010 to 31 December 2014, and eventually won in this period among 119 relatively frequent trading bank debt. In this 119 bonds as samples to build trading strategies, build trading signals moving average model, in the proportion of funds multiplier 0.5,1.0,1.5 and 2.0 and 2.5 m, respectively, and each investment bonds only limited to 2%,3% and 4% and k lag period of 10 weeks,20 weeks,30 weeks and 40 weeks, the gains of more than buy and hold bonds to maturity composition about 6-8bp, more than inter-bank bond index in debt about 7-9bp, wherein when the lag period of 40 weeks, the excess return obtained almost zero, mainly due to the lag period is too long, so that the trading signals insensitive to price changes; Construction trades based on time series model, the ride the proportion of sub-funds 0.5,1.0,1.5 and 2.0 and 2.5 m, respectively, and each investment bonds is limited to only 2%,3% and 4% of the cases, the gains than buy and hold bonds to maturity composition 3-5bp, inter-bank bond index over the debt 3-4bp. Wherein when the multiplier of 2.5 m, the excess return obtained two trading strategy is almost zero, this may be due to the volatility of bond prices is small, multiplier m set too large, making it difficult issue trading signals; first 3, in accordance with Chapter III empirical research on the Exchange corporate bonds pricing, pricing errors mean reversion principle to build trading strategies using Nettekoven (2005) based methods for Exchange corporate bond pricing errors were based on the moving average model and time Sequence Model trading signals. By analyzing corporate bond trading week Information December 31,2007 to December 31,2014, and eventually won the 83 listed companies issued A shares listed in the 115 corporate bonds. In this sample of 115 corporate bonds to build trading strategies, build trading signals based on the moving average model, when the ratio of funds multipliers 1,2,2.5 and 3 m, respectively, and each investment corporate bonds is limited to 2%,3% and 4%, and the lag period of 20 weeks, earnings at 30 weeks,40 weeks and 50 weeks in the case received in excess of buy and hold a combination of corporate debt maturity about 40-50bp, exceeding the SSE Corporate Bond Index about 30-40bp; based on time series model to build trading signals, when multipliers 1,2,2.5 and 3 m, respectively, and the proportion of funds invested in each company’s only debt is limited to 2%,3% and 4%, the gains over buy the combination of corporate bonds held to maturity of approximately 35-46bp, exceeding the SSE Corporate Bond Index about 30-45bp. By comparing the two trading strategies excess returns found little difference between the two, and when the multiplier of 2 and 2.5 m, the excess rate of return based on these two trading strategies acquired up 40-50bp, and when the multiplier m as 3, the excess return obtained two trading strategies are less, this is mainly due to multiplier m set too large, making it difficult to issue trading signals. According to the latest debt credit rating to 115 corporate bonds are divided into three groups to calculate excess returns in both trading strategies available, AAA grade corporate bonds in combination than buy and hold bonds to maturity composition about 20-38bp, more than in SSE Corporate Bond Index 18-28bp; AA+grade corporate bonds in combination than buy and hold bonds to maturity composition about 38-52bp, exceeding the SSE Corporate Bond Index about 35-50bp; AA, AA-, A grade corporate bonds a combination of more than buy and hold bonds to maturity composition about 50-80bp, exceeding the SSE Corporate Bond Index 50-72bp. Over the gains as credit rating fall significantly increased, indicating that the impact of the credit risk premium trading strategies more significant. Overall, since the corporate bond pricing error fluctuations, based on the principles of the pricing error mean reversion trading strategies more likely to get excess returns, inter-bank bond market compared to this trading strategy is more suitable for the corporate bond market of the Exchange; Section 4 of this chapter summarized.Chapter 6 is conclusions and policy recommendations. This chapter reviews the recent progress on the conclusions, and conclusions based on the future development of China’s inter-bank bond market, the corporate bond market and the exchange gives the corresponding policy recommendations.Innovation of this paper is mainly reflected in the following three areas:(1) Innovation study, the current domestic research on quantitative trading strategies focused on the stock market and the futures market, trading strategies less research on the bond market, and this paper trading strategies for conduct in China’s bond market research; innovation (2) research methods, we will apply to China’s bond market based on the principles of the pricing error mean reversion trading strategy, and on the basis of previous strong abilities to describe selected The model of inter-bank bonds and corporate bonds priced exchanges, and ultimately select the most suitable for China’s bond market pricing model as the theoretical basis for building transaction pricing strategy; and (3) innovation research perspective, we use several models of the theoretical price of the bond estimate, changing the results of previous research analyzed only a single model, so this article is more persuasive.The inadequacies of the paper is:(1) conducting trading strategy does not take into account the transaction costs of bonds, if taking into account the transaction costs will reduce the effectiveness of trading strategies; (2) Due to the lack of liquidity of the bond market, in particular, the inter-bank bond market liquidity is worse, the number of transactions resulting in too little in the absence of actual transactions can only be in debt valuation data as the actual transactions, valuation of the data in the sample period is too smooth, the pricing error is small, This also leads to one of the main trading strategies in the inter-bank bond market is not good effect; Reason (3) Based on the discussion of the pricing error generated will focus on the analysis of liquidity factors, and no other factors that may cause pricing error on for further analysis.
|
PDF Full Text Request