我国外汇储备中美元资产的风险度量及组合投资

发布时间：2018-04-27 12:04

本文选题：外汇储备 + 熵式风险　；参考：《东北财经大学》2012年硕士论文

【摘要】：随着经济全球化和我国金融市场改革与对外开放的不断深化,我国庞大的外汇储备一方面可以为进口贸易提供强有力的支撑,另一方面也面临着越来越复杂的金融风险。同一般商业银行外汇资金一样,中央银行的外汇储备在经营中也面临着各式各样的风险。长期以来,我国以美元资产为主的外汇储备结构面临极大的汇率风险,尤其是在美元贬值的大背景下,美元汇率的走低会直接引起我国外汇储备账面资产的缩水。2008年金融危机严重打击了美国经济,导致美元疲软。从2007年8月美国次贷危机全面爆发到2008年底,美元对人民币汇率从7.5600降到6.8346,贬值高达9.6%。若以我国1.9万亿美元的外汇储备测算,汇率风险造成的损失就高达13782.6亿元。我国大量的美元资产以美国国债形式存在,资产结构较为单一,导致风险较为集中。为摆脱金融危机,美国政府通过的7000亿美元的救市计划需要通过发行大量的国债来实现,这导致已发行的国债市场价格下跌,而且美联储为刺激经济所采取的连续降息政策,也导致国债收益率下降。这无疑进一步给我国外汇储备造成损失。我国的外汇储备结构存在期限错配问题,期限错配带来了流动性风险。从我国投资于美元证券的结构可以看出,长期证券投资所占比重较大,相比较而言短期投资所占比重不大。中国持有的美元资产的投资工具和期限结构在不断调整。截止至2011年6月中国共持有美元资产1.61万亿。其中,长期债券14793亿,短期债券49亿,股权投资1265亿。长期债中国债、机构债和公司债分别约为11081亿、3600亿和111亿,短期债中国债、机构债和公司债分别约为40亿、0.75亿和8.47亿。根据美国财政部公布的各国持有美国证券的资产组合显示中国的长期债券投资占证券投资总额的90%以上,而短期债券投资仅占证券投资总额的7%。而人民币升值引发了大量的国际资本流入,这些资本主要投资于国内的债券、股票等流动性较强的金融工具上,也就是说中国把长期投资投到国外,却引入了大量的短期投资。期限错配把中国置于巨大的流动性风险面前,一旦资本外逃,东南亚金融危机就可能重演。短期内,我国的外汇储备还会继续主要投资于美元,美元资产内部的结构调整更适合中国当前外汇储备管理现状。综上所述,巨额美元资产的有效管理对于研究如何有效地察觉和防范我国外汇储备资产面临的风险,得到我国外汇储备的最优投资组合意义重大。金融风险管理的基础与前提是风险度量。只有科学准确地度量风险的大小,才可以为风险管理奠定良好的基础。资产管理者及监管者对风险进行管理时,需要知道风险的大小,如果不能对风险进行科学计量,就不能对风险进行有效管理,所以,对于巨额美元资产的有效管理离不开对其可能带来的风险的科学计量。 Shannon在研究数学通讯理论时发现了信息熵理论,它是研究信息系统不确定性测度的指标。由于金融投资风险是金融投资收益不确定性的体现,所以,这一理论也被用于金融投资风险的计量中。信息熵作为金融投资风险(不确定性)的计量指标概念清晰、意思明了,将系统的不确定性用统一的数字来反映,为不同系统不确定性之间的比较提供了客观的标准,除此之外,熵式风险计量的结果与投资者对资产的预期收益率有关,因此,它具有风险事先计量的特征。采用风险的熵式度量原理,本文度量了我国外汇储备中的短期和中期美国国债的熵式风险。但是,该指标没有突出损失与收益的差别,这与投资者的心理感受不同,而且,没有考虑损失的大小,而仅考虑各种状态分布的概率。为了使得金融资产的风险能够具体量化为一个与收益相配比的数字,突出损失与收益的差别,从而符合投资者的心理感受,引入风险计量的VaR方法。 VaR方法计量金融风险含义简洁、价值判断比较直观,使得金融资产的风险能够具体量化为一个与收益相配比的数字,突出了损失与收益的差别,这与投资者的心理感受相符,从而有利于资产管理目标的实现。并且,VaR方法计量金融风险时,考虑了投资决策者所处的具体环境,使风险度量及决策更具有可操作性。在完成金融风险计量的基础上我们开始研究金融风险管理。马柯维兹的资产组合理论是研究期望收益和方差的组合,马柯维兹的资产组合理论一个最基本的假设就是投资者期望在给定的风险水平下获得最大的投资收益,或者在相同的收益水平下,能够承担最小的风险。资产组合理论假设投资者都是风险厌恶的,也就是说,在收益相同的两种资产之间他们会选择风险水平较低的。马柯维兹的经典理论的思路可以表述为如下四个阶段：首先,区分有效的资产组合和无效的资产组合；其次,描述有效资产组合的收益性与不确定性(风险),即用期望收益和方差来描述有效资产组合；再次,选择最能满足投资者期望效用的收益和不确定性的组合：最后,确定满足投资者效用最大化的这种投资组合的资产构成。但是,均值-方差的资产组合的分析方法只是提供收益率相对于均值对称性离散的程度,马柯维兹自己也指出应该用其他方法来代替和补充。针对模型中以方差或标准差作为风险计量的指标,许多学者提出批评：其中显而易见的一点是,方差是用来反应收益率的易变性或不确定性的,正负偏差之间具有对称性,然而损失和利得对于风险的贡献度是不同的,投资者对于偏离均值的上下偏差具有明显不同的看法,所以,收益率的方差计量风险违背了投资者对风险的真实心理感受。因此,有必要将“均值-方差”资产组合管理中的风险计量指标替换为VaR,从而引出如下的“均值-VaR”资产组合管理模型。我们运用“均值-VaR”资产组合管理模型,利用LING012实证研究了不同收益率要求下的非长期国债、长期国债、抵押债、公司债、股权等美元资产的最优投资组合。
[Abstract]:With the economic globalization and the deepening of China's financial market reform and opening to the outside world, China's huge foreign exchange reserves can provide strong support for import trade on the one hand, and more and more complex financial risks on the other hand. The foreign exchange reserve of the central bank, like the foreign exchange funds of the general commercial banks, is also in operation. There are all kinds of risks.
For a long time, the foreign exchange reserve structure based on US dollar assets is facing great exchange rate risk, especially in the background of the depreciation of the dollar. The decline of the dollar exchange rate will directly cause the shrinking of the foreign exchange reserve account assets of our country in.2008 years. The financial crisis has severely hit the United States economy and led to the weakening of the dollar. From August 2007, the subprime mortgage in the United States was subprime. The crisis broke out by the end of 2008, the dollar fell from 7.5600 to 6.8346 of the RMB exchange rate from 7.5600 to 9.6%., as high as $1 trillion and 900 billion of foreign exchange reserves in China, and the loss caused by exchange rate risk was up to 1 trillion and 378 billion 260 million yuan.
In order to get rid of the financial crisis, the US government's $700 billion bailout plan needs to be realized by issuing a large amount of treasury bonds to get rid of the financial crisis. This leads to the decline in the price of the issued treasury bond market and the fed to stimulate the economy. The continuous interest rate reduction policy has also led to a decline in the yield of treasury bonds, which will undoubtedly further cause losses to China's foreign exchange reserves.
China's foreign exchange reserve structure has a period of mismatch, which brings liquidity risk. From the structure of our investment in US dollar securities, it can be seen that the proportion of long-term securities investment is larger than that of short-term investment. The investment tools and term structure of China's dollar assets are constantly adjusted. As of June 2011, China held a total of 1 trillion and 610 billion dollar assets, including long-term bonds of 14793 billion, short-term bonds 4 billion 900 million, equity investment 126 billion 500 million. Long term debt of China, institutional bonds and corporate bonds were about 11081 billion, 360 billion and 11 billion 100 million, short term debt of China, institutional debt and corporate debt of 4 billion, 75 million and 847 million, respectively.
According to the United States Treasury's portfolio of assets held by the US securities, China's long-term bond investment accounts for more than 90% of the total stock investment, while the short-term bond investment accounts for only 7%. of the total amount of securities investment, while the appreciation of the RMB has triggered a large amount of international capital inflows, which are mainly invested in domestic bonds, stocks and other flows. On the more dynamic financial instruments, that is to say, China has invested long term investment abroad, but has introduced a large amount of short-term investment. The mismatch put China in the face of huge liquidity risk. Once capital escaping, the financial crisis in Southeast Asia may repeat itself.
In the short term, China's foreign exchange reserves will continue to be mainly invested in US dollars, and the structural adjustment within the US dollar assets is more suitable for the current situation of China's current foreign exchange reserve management. In summary, the effective management of the huge dollar assets will be used to study how to effectively detect and prevent the risks faced by our foreign exchange reserve assets. The basis and premise of the financial risk management is the risk measurement. Only the scientific and accurate measurement of the size of the risk can lay a good foundation for the risk management. When the asset manager and the supervisor are managing the risk, it needs to know the size of the risk, and if the risk is not measured scientifically, it will not be possible. It can effectively manage risks, so the effective management of huge dollar assets is inseparable from the scientific measurement of the risks it may bring.
Shannon found the theory of information entropy in the study of the theory of mathematical communication. It is an indicator of the uncertainty measurement of the information system. Because the risk of financial investment is the embodiment of the uncertainty of the financial investment income, this theory is also used in the measurement of the risk of financial investment.
The concept of information entropy as a measurement index of financial investment risk (uncertainty) is clear. It is clear that the uncertainty of the system is reflected by a unified number, which provides an objective standard for the comparison of the uncertainty of different systems. In addition, the results of entropy risk measurement are related to the expected rate of return on assets by the investors, so the results of the entropy risk measurement are related to the expected returns of the assets. It has the characteristics of risk beforehand measurement. By using the entropy measure principle of risk, this paper measures the entropy risk of the short-term and medium-term treasury bonds in China's foreign exchange reserve.
However, the index does not highlight the difference between the loss and the income, which is different from the investor's psychological feeling, and does not consider the size of the loss, but only considers the probability of a variety of state distribution. In order to make the risk of the financial asset specific as a number that matches the income ratio, the difference between the loss and the income is highlighted, thus conforming to the investment. The psychological experience of investors and the VaR method of risk measurement are introduced.
The VaR method is simple to measure financial risk and is more intuitionistic in value judgment. The risk of financial assets can be quantified as a figure matched with income, which highlights the difference between loss and income. This is in accordance with the psychological feelings of investors and is conducive to the realization of the target of asset management. And when the VaR method is to measure the financial risk, Considering the specific environment of investment decision-makers, risk measurement and decision making are more operational.
On the basis of financial risk measurement, we begin to study financial risk management. Makavitz's portfolio theory is a combination of expected returns and variance. The most basic hypothesis of makovitz's portfolio theory is that investors expect to obtain maximum investment returns at a given risk level, or in the same way. The portfolio theory assumes that investors are all risk aversion, that is, they will choose lower risk levels between the two assets with the same income.
The thinking of makecviz's classic theory can be expressed as four stages as follows: first, distinguish effective portfolio and invalid portfolio; secondly, describe the income and uncertainty (risk) of the effective portfolio, that is, the effective asset portfolio is described with the expected return and variance; and again, the choice is best to meet the investor expectations. The combination of utility's income and uncertainty: finally, determine the asset composition of this portfolio that satisfies the maximization of investor utility. However, the mean variance portfolio analysis method only provides the degree of return relative to the mean symmetry, and makovitz has also pointed out that other methods should be used instead of and supplemented.
In view of the variance or standard deviation as an indicator of risk measurement, many scholars have put forward criticism: the obvious one is that the variance is used to respond to the volatility or uncertainty of the rate of return, and the positive and negative deviations are symmetrical. However, the contribution of loss and gain to risk is different, and the investor's deviation is the same. There is a distinct difference between the upper and lower deviations of the value, so the variance measurement risk of the yield is contrary to the real psychological feeling of the investor. Therefore, it is necessary to replace the risk measurement index of "mean variance" portfolio management to VaR, which leads to the following "mean -VaR" portfolio management model. "Mean -VaR" portfolio management model, using LING012 empirical research on the optimal portfolio of non long-term treasury bonds, long-term treasury bonds, mortgage debt, corporate debt, equity and other dollar assets under the requirements of different rates of return.

【学位授予单位】：东北财经大学
【学位级别】：硕士
【学位授予年份】：2012
【分类号】：F832.6;F224

【参考文献】