金融危机背景下的我国外汇储备汇率风险动态研究

本文利用DCC-MGARCH模型结合风险价值VaR和条件风险价值CVaR描述了次贷危机前后我国外汇储备的汇率风险,发现在次贷危机后,VaR和CVaR均大幅增加,且CVaR作为风险衡量工具更加谨慎,同时得出:提高美元比重,可以降低外汇储备的汇率风险;欧元、日元、英镑出现剧烈波动是次贷危机以后外汇储备汇率风险大幅增加的主要原因。     

中国外汇储备资产的币种结构、收益和风险分析

文章选取2001～2010年月度数据,首先利用计量模型估计了我国外汇储备币种结构及收益率,在此基础上采用VaR方法分析了外汇储备的风险变化,结果表明:在我国外汇储备币种构成上,欧元资产比重上升,美元、日元资产的比重下降,英镑资产的比重基本维持不变,美元资产仍处于主导地位;外汇储备资产的平均收益率在下降,面临的风险在加大。文章建议:适当降低美元资产在外汇储备中的比重,适当增加欧元资产比重;成立专门的外汇储备资产管理公司,对外汇储备进行投资和管理;加强外汇储备资产的风险管理。     

欧元流通对我国的影响及对策

诺贝尔经济学奖获得者、被誉为“欧元之父”的罗伯特·孟德尔教授认为 ,欧元的诞生是二战以后美元取代英镑成为全球第一大货币之后国际金融领域出现的最重大进展 ,它必将改变国际金融体系的力量格局。本文将详细分析欧元全面流通对我国经济在生产和贸易、投融资、外汇储备、金融业发展以及货币区域一体化等方面的影响 ,并提出我国应采取的对策。     

基于ARCH类模型的人民币汇率波动特征比较

文章基于ARCH类模型,利用对2005年7月20日至2011年5月31日的人民币高频日汇率数据进行实证研究,分析了人民币兑美元、欧元、日元、港币与英镑汇率的波动率特征。实证结果表明,人民币兑美元、港币的波动性较其它货币小一个数量级;各汇率序列具有明显的尖峰厚尾与波动率聚集特征;除港币外,人民币汇率波动主要由过去波动引起,并且冲击的影响会持续较长时间,外部效应较弱;日元与其它货币的杠杆效应不同,好消息比坏消息冲击效应更大;风险溢价效应只在人民币兑美元与港币的汇率中显著。这些特征表明人民币汇率应更加市场化。     

汇率预测及其经济基本面：基于多元自适应可变窗算法的构建

在人民币国际化不断推进、人民币汇率双向波动加强的背景下，构建具有优良预测能力的人民币汇率预测模型意义重大。参数模型对汇率预测的能力不仅取决于模型设定是否正确，同时还取决于能否迅速探测参数的结构性变化以使用最佳信息估计模型参数。本文构建了多元自适应可变窗算法以及时监测模型参数的时变特征，探测最大化参数同质区间。结果显示：①在中长期（3至24个月）的美元、欧元、英镑和日元兑人民币汇率的样本外推预测中，多元自适应可变窗算法能显著优于随机游走模型、购买力平价模型、弹性货币模型、利率平价模型、泰勒规则模型与偏移型泰勒规则模型这六种汇率预测主流模型，其预测能力也显著优于实时窗宽选择算法与自回归模型；在美元兑人民币汇率中长期（3至24个月）预测中，其预测误差MAE度量相比于次优模型能降低 25%～50%。②多元自适应可变窗算法能迅速捕捉美元、欧元、英镑和日元兑人民币汇率的拐点，预测人民币汇率走向并刻画人民币汇率的周期性变化，其长期（向前9个月）方向性趋势样本外推预测精度比次优模型提高了16%～40%。③断点前后的汇率动态结构性变化显示“811”汇改促进了经济基本面对汇率预期重要性的显著提升与市场风险偏好的转变。“811”汇改之后，人民币汇率预期更易受外部冲击影响。加速利率市场化建设、提高国内收入、稳定物价、坚持带管制的浮动汇率制度与有效的资本管制相结合等措施对促进汇率市场化、防止汇率风险具有重要意义。     

我国外汇储备资产的币种结构、资产收益与投资风险的实证分析

文章根据外汇储备资产币种选择理论,以我国金融数据(1978~2010)为考据,首先分析了对外贸易、外商直接投资和经济实力与我国外汇储备的相关关系,其次对外汇储备的收益和风险进行了分析,得出我国外汇储备的币种结构应该以美元为主、增持港元、适当增持其他主要国际币种的多样化构成方式,以降低美元贬值所造成的收益损失。并据此对我国外汇储备资产的构成提出相关政策建议,对优化我国外汇储备币种结构的优化有一定借鉴意义。     

人民币与国外主要货币的尾部相依和联动

文章选择1994年1月到2014年11月的人民币、美元、英镑、日元、欧元和港币等六种货币实际有效汇率的月度数据,首先采用AR-GARCH-t过程对收益率进行过滤,然后运用规则藤Copula函数对标准化残差序列进行建模,探讨人民币“第一次汇改”前后六种汇率之间的波动溢出效应,研究汇率之间的尾部相依和联动现象。研究发现：“第一次汇改”前后,人民币与其它五种货币实际有效汇率之间的相依结构发生了变化,且联动中的主导货币也由美元变为以人民币,但仍可用同一类型的规则藤Copula结构来描述;人民币与其它货币之间都存在对称的或非对称的尾部相依,其中人民币与日元之间存在对称的负尾部相依,而与欧元之间则存在负的上尾部相依。“第一次汇改”后,人民币与美元实际有效汇率之间正的尾部相依程度有所增加,且相依性最大,而港币与美元之间正的尾部相依性有较大幅度的下降。此外,人民币与日元之间负的尾部相依关系也有较大幅度的下降。     

人民币汇率风险的测度

文章引入极值理论对欧元/人民币和日元/人民币的收益序列尾部进行估计,发现收益序列尾部与广义Pareto分布拟舍得非常好;实证结果表明:期望损失(ES)在某些情况下能弥补在险价值(VaR)所具有的尾部风险.由返回检验的结果知,与历史模拟法和方差-协方差法计算的VaR相比,基于极值理论的VaR能更准确地度量欧元/人民币和日元/人民币的风险.     

数字货币助力人民币国际化

从2009年7月央行启动跨境贸易人民币结算试点算起,人民币国际化战略已经推进了10余年。中国人民银行发布的《2021年人民币国际化报告》指出:2021年一季度,在国际货币基金组织官方外汇储备货币构成中人民币排在第五位,人民币在全球外汇储备中的占比为2.5%,较2016年人民币刚加入特别提款权货币篮子时上升1.4个百分点;环球银行金融电信协会发布的数据显示,2021年6月,在主要国际支付货币中人民币排在第五位,人民币支付金额占所有货币支付金额的2.5%,较上年同期上升0.7个百分点。尽管人民币在这些指标中的排名靠前,但与美元、欧元差距还很大。     

基于藤结构Copula-SV模型的外汇投资组合风险分析

文章针对现有时间序列统计模型对高维金融资产收益率特征描述不充分导致的风险度量不准确问题,构建了藤结构Copula-SV模型,用以对单个外汇资产收益率的波动异方差性和外汇资产间的复杂相关性进行描述.基于四种外汇资产(美元、欧元、日元和港币)的实证结果表明藤结构Copula-SV模型能较好地描述外汇资产的波动异方差性和复杂相关性,且基于拟合模型采用对偶变量法的Monte Carlo模拟计算出来的VaR通过了Kupiec失败率检验.     

Modeling Heteroscedasticity in Daily Foreign-Exchange Rates

This article estimates autoregressive conditionally heteroscedastic (ARCH) and generalized ARCH (GARCH) models for five foreign currencies, using 10 years of daily data, a variety of ARCH and GARCH specifications, a number of nonnormal error densities, and a comprehensive set of diagnostic checks. It finds that ARCH and GARCH models can usually remove all heteroscedasticity in price changes in all five currencies. Goodness-of-fit diagnostics indicate that exponential GARCH with certain nonnormal distributions fits the Canadian dollar extremely well and the Swiss franc and the deutsche mark reasonably well. Only one nonnormal distribution fits the Japanese yen reasonably well. None fit the British pound.     

欧元导入后外汇市场的特征变化

我们使用改进的EGARCH模型考察了外汇市场在欧元导入前后的特征变化。我们发现几个主要货币对美元汇率的波动幅度较之前增加了,各汇率之间的相关程度也加深了。这些现象的发现具有实际的操作意义。     

Particle Learning for Fat-Tailed Distributions

It is well known that parameter estimates and forecasts are sensitive to assumptions about the tail behavior of the error distribution. In this article, we develop an approach to sequential inference that also simultaneously estimates the tail of the accompanying error distribution. Our simulation-based approach models errors with a t_ν-distribution and, as new data arrives, we sequentially compute the marginal posterior distribution of the tail thickness. Our method naturally incorporates fat-tailed error distributions and can be extended to other data features such as stochastic volatility. We show that the sequential Bayes factor provides an optimal test of fat-tails versus normality. We provide an empirical and theoretical analysis of the rate of learning of tail thickness under a default Jeffreys prior. We illustrate our sequential methodology on the British pound/U.S. dollar daily exchange rate data and on data from the 2008–2009 credit crisis using daily S&P500 returns. Our method naturally extends to multivariate and dynamic panel data.     

新汇改以来人民币汇率参考一篮子货币权重的估计及经验验证

本文首先构建人民币对美元汇率指数和篮子货币汇率指数,根据人民币对美元汇率指数和篮子货币相关性指数最大化可以确定篮子货币的权重和数量,并对新汇改后篮子货币的权重和数量进行测度。我们的研究发现：在参考一篮子货币中,美元的权重最高,英镑的权重最小,但是其他货币权重都占有一定比例,我们的货币篮子和央行的货币篮子线性相关性较高。根据每一种货币的权重大小,我们可以计算出篮子中每一种货币的数量。根据最优货币权重,我们研究发现人民币对美元汇率和篮子货币之间存在长期协整关系,误差修正模型显示人民币汇率参考篮子货币调整的短期弹性和长期弹性比较接近,说明人民币汇率参考篮子货币调整相对比较稳定。本文检验了模型的预测效果,模型拟合度较高,说明我们选取的货币篮子具有较好的代表性。     

Testing for Nonlinear Autoregression

This article considers consistent testing the null hypothesis that the conditional mean of an economic time series is linear in past values. Two specific tests are discussed, the Cramér–von Mises and the Kolmogorov–Smirnov tests. The particular feature of the proposed tests is that the bootstrap is used to estimate the nonstandard asymptotic distributions of the test statistics considered. The tests are justified theoretically by asymptotics, and their finite-sample behaviors are studied by means of Monte Carlo experiments. The tests are applied to five U.S. monthly series, and evidence of nonlinearity is found for the first difference of the logarithm of the personal income and for the first difference of the unemployment rate. No evidence of nonlinearity is found for the first difference of the logarithm of the U.S. dollar/Japanese Yen exchange rate, for the first difference of the 3-month T-bill interest rate and for the first difference of the logarithm of the M2 money stock. Contrary to typically used tests, the proposed testing procedures are robust to the presence of conditional heteroscedasticity. This may explain the results for the exchange rate and the interest rate.     

The use of the ARDL approach in estimating virtual exchange rates in India

This paper applies the autoregressive distributed lag approach to cointegration analysis in estimating the 'virtual exchange rate' (VER) in India. The VER would have prevailed if the unconstrained import demand were equal to the constraint imposed due to foreign exchange rationing and the VER is used to approximate the 'price' of rationed foreign exchange reserves. We highlight the shortcomings of the existing literature in approximating equilibrium exchange rates in a less developed country such as India and propose the VER approach for equilibrium rates, which uses information from an estimated structural model. In this relationship, black market real exchange rate (E U ) is a dependent variable and real official exchange rates (E O ), the ratio of the foreign (r*) to the domestic (r) interest rate (I), and official forex reserves (Q) are explanatory variables. In our estimation, the VERs are higher than E O by about 10% in the short-run and 16% in the long-run.     

Semiparametric ARCH Models

This article introduces a semiparametric autoregressive conditional heteroscedasticity (ARCH) model that has conditional first and second moments given by autoregressive moving average and ARCH parametric formulations but a conditional density that is assumed only to be sufficiently smooth to be approximated by a nonparametric density estimator. For several particular conditional densities, the relative efficiency of the quasi-maximum likelihood estimator is compared with maximum likelihood under correct specification. These potential efficiency gains for a fully adaptive procedure are compared in a Monte Carlo experiment with the observed gains from using the proposed semiparametric procedure, and it is found that the estimator captures a substantial proportion of the potential. The estimator is applied to daily stock returns from small firms that are found to exhibit conditional skewness and kurtosis and to the British pound to dollar exchange rate.     

Periodic Autoregressive Conditional Heteroscedasticity

Most high-frequency asset returns exhibit seasonal volatility patterns. This article proposes a new class of models featuring periodicity in conditional heteroscedasticity explicitly designed to capture the repetitive seasonal time variation in the second-order moments. This new class of periodic autoregressive conditional heteroscedasticity, or P-ARCH, models is directly related to the class of periodic autoregressive moving average (ARMA) models for the mean. The implicit relation between periodic generalized ARCH (P-GARCH) structures and time-invariant seasonal weak GARCH processes documents how neglected autoregressive conditional heteroscedastic periodicity may give rise to a loss in forecast efficiency. The importance and magnitude of this informational loss are quantified for a variety of loss functions through the use of Monte Carlo simulation methods. Two empirical examples with daily bilateral Deutschemark/British pound and intraday Deutschemark/U.S. dollar spot exchange rates highlight the practical relevance of the new P-GARCH class of models. Extensions to discrete-time periodic representations of stochastic volatility models subject to time deformation are briefly discussed.     

Bayesian Inference in Generalized Error and Generalized Student-t Regression Models

This study takes up inference in linear models with generalized error and generalized t distributions. For the generalized error distribution, two computational algorithms are proposed. The first is based on indirect Bayesian inference using an approximating finite scale mixture of normal distributions. The second is based on Gibbs sampling. The Gibbs sampler involves only drawing random numbers from standard distributions. This is important because previously the impression has been that an exact analysis of the generalized error regression model using Gibbs sampling is not possible. Next, we describe computational Bayesian inference for linear models with generalized t disturbances based on Gibbs sampling, and exploiting the fact that the model is a mixture of generalized error distributions with inverse generalized gamma distributions for the scale parameter. The linear model with this specification has also been thought not to be amenable to exact Bayesian analysis. All computational methods are applied to actual data involving the exchange rates of the British pound, the French franc, and the German mark relative to the U.S. dollar.