我国商品期货定价：跳跃—扩散模型研究

发布时间：2018-01-24 21:20

  本文关键词： 商品期货 跳跃-扩散模型 MCMC算法　出处：《北方工业大学》2013年硕士论文　论文类型：学位论文

【摘要】：期货市场具有价格发现和套期保值的功能,可有效规避价格风险。而对于期货价格变化的研究,有助于市场参与者了解市场方向。我国商品期货市场不仅具有期货市场的一般性质,而且还拥有自己的独特性。本文的主要目的是提出符合我国商品期货价格变化规律的定价模型,并验证模型的定价能力。 文章首先分析了国内外关于商品期货定价的相关理论,主要模型的产生及发展过程,为建立我国商品期货定价模型奠定了基础。 然后重点分析了我国商品期货市场价格变化的两个特点：一是长期受国际商品期货价格的影响,二是具有跳跃特征,在此基础上提出了体现我国商品期货价格变化特点的跳跃-扩散模型。该模型将我国商品期货的价格变化视为可观测变量的影响和不可观测变量之和。可观测变量为国际商品期货价格,不可观测变量为短期偏离,跳跃特征包含在短期偏离中。 随后的实证部分以我国棉花、铜和豆粕期货为例,各选取3个完整期货合约(时间跨度从2010年1月至2013年1月)用MCMC方法进行参数估计。数据选取每日收盘价序列,运用R软件和WinBUGS软件得出参数的估计结果和跳跃成分的拟合效果。结果显示：国际商品期货的价格变化对我国棉花、铜和豆粕期货的价格变化有显著的影响；跳跃成分是我国商品期货价格变化的重要组成部分；本文提出的定价模型能较准确地反映跳跃强度,不论跳跃现象发生次数的多寡。 除此之外,文中将此次提出的定价模型与前人提出的定价模型做了对比分析,发现包含国际商品期货价格影响因素的跳跃-扩散模型更加符合我国商品期货价格变化的实际情况,模型的整体拟合效果也较稳定。 总而言之,研究我国商品期货定价问题时,应当考虑到国际商品期货价格的影响和跳跃特征,这不仅可以较好地解决我国商品期货的定价问题,而且能够为下一步提出套期保值策略奠定良好基础。
[Abstract]:Futures market has the function of price discovery and hedging, which can effectively avoid the price risk. It helps market participants to understand the direction of the market. China's commodity futures market not only has the general nature of the futures market. The main purpose of this paper is to put forward a pricing model which conforms to the law of commodity futures price change in China, and to verify the pricing ability of the model. This paper first analyzes the relevant theories of commodity futures pricing at home and abroad, the emergence and development process of the main models, which lays a foundation for the establishment of commodity futures pricing models in China. Then it focuses on the two characteristics of the price change of the commodity futures market in China: one is influenced by the international commodity futures price for a long time, the other is the characteristic of jumping. On this basis, a jump-diffusion model is proposed, which reflects the characteristics of commodity futures price change in China. The model regards the price change of commodity futures in China as the influence of observable variables and the sum of non-observable variables. The quantity is the international commodity futures price. Non-observable variables are short-term deviations and jump features are included in short-term deviations. Then the empirical part takes China cotton, copper and soybean meal futures as an example. Each of the three complete futures contracts (time span from January 2010 to January 2013) was estimated by MCMC method. The data were selected as the daily closing price sequence. Using R software and WinBUGS software to obtain the parameter estimation results and jump component fitting effect. The results show that: international commodity futures price changes on China's cotton. The price changes of copper and soybean meal futures have significant influence; Jump component is an important part of commodity futures price change in China. The pricing model proposed in this paper can accurately reflect the jump intensity, regardless of the number of jumps. In addition, the paper makes a comparative analysis between the proposed pricing model and the previous pricing model. It is found that the jump-diffusion model, which includes the influencing factors of international commodity futures price, is more in line with the actual situation of commodity futures price change in China, and the overall fitting effect of the model is more stable. In a word, when studying the pricing problem of commodity futures in our country, we should take into account the influence and jump characteristics of international commodity futures prices, which can not only solve the pricing problem of commodity futures in China. And can lay a good foundation for the next step to put forward hedging strategy.
【学位授予单位】：北方工业大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：F224;F724.5

【相似文献】

中国期刊全文数据库 前10条

1 秦学志,吴冲锋;可赎回的可转换债券的博弈定价方法[J];系统工程;2000年05期

2 彭作和;田澎;;供应链临时价格折扣研究[J];工业工程;2006年02期

3 汪航;李飞;;我国股票挂钩型银行理财产品定价分析[J];海南金融;2009年07期

4 赵洋;赵立臣;;基于蒙特卡罗模拟的可转换债券定价研究[J];系统工程学报;2009年05期

5 刘卫国;李乾文;;企业知识产权交易定价的博弈分析[J];统计与决策;2010年22期

6 卢子芳,金毅;适应市场的综合生产计划方法研究[J];管理工程学报;1999年03期

7 陈波，

本文编号：1461048