Dynamic Correlation Model Research Articles

Dynamic partial correlation models

We introduce a new scalable model for dynamic conditional correlation matrices based on a recursion of dynamic bivariate partial correlation models. By exploiting the model’s recursive structure and the theory of perturbed stochastic recurrence equations, we establish stationarity, ergodicity, and filter invertibility in the multivariate setting using conditions for bivariate slices of the data only. From this, we establish consistency and asymptotic normality of the maximum likelihood estimator for the model’s static parameters. The new model outperforms benchmarks like the t-cDCC and the multivariate t-GAS, both in simulations and in an in-sample and out-of-sample asset pricing application to US stock returns.

Modelling of wind and photovoltaic power output considering dynamic spatio-temporal correlation

The simulation technology of wind and solar power output can provide data support for the planning of new energy stations and the optimization and scheduling of power systems. In order to solve the problem that the existing output models can not accurately describe the dynamic spatio-temporal dependence between wind and solar output, a dynamic spatiotemporal correlation model of wind and solar output based on Markov stochastic process and Copula function theory is proposed. Firstly, wind power and photovoltaic output are regarded as a stochastic process, and the time autocorrelation models of wind power and photovoltaic output are constructed based on a one-dimensional Markov chain and hybrid Copula function. Next, two one-dimensional Markov chains were coupled to form a two-dimensional Markov chain, and a spatial correlation model between wind and solar output was constructed using the dynamic SJC Copula function. Take the measured data of adjacent wind farms and photovoltaic power stations in Hami, Xinjiang as an example for simulation. The simulation results show that the proposed model can effectively reflect the spatio-temporal correlation of the original data and reflect the dynamic changes in the correlation between wind and solar energy.

Macroeconomic forecasting and variable ordering in multivariate stochastic volatility models

We document five novel empirical findings on the well-known potential ordering drawback associated with the time-varying parameter vector autoregression with stochastic volatility developed by Cogley and Sargent (2005) and Primiceri (2005). First, the ordering does not affect point prediction. Second, the standard deviation of the predictive densities implied by different orderings can differ substantially. Third, the average length of the prediction intervals is also sensitive to the ordering. Fourth, the best ordering for one variable in terms of log-predictive scores does not necessarily imply the best ordering for another variable under the same metric. Fifth, the ordering problem becomes exacerbated in conditional forecasting exercises. Then, we consider three alternative ordering invariant models: a canonical discounted Wishart stochastic volatility model and two dynamic stochastic correlation models. When the forecasting performance of these ordering invariant models is compared to Cogley, Primiceri, and Sargent’s ordering variant model, the former underperforms relative to all orderings and the latter two have an out-of-sample forecasting performance comparable with the median outcomes across orderings.

The dependencies of subindexes of Stoxx 600 during the Covid-19 pandemic

In this index study, the relationships between Stoxx Europe 600 and sector indices are analyzed. This research uses DCoVar and MES as analytical tools developed as a measure of systemic risk and applied to financial institutions, to sectoral subindexes. For the sake of systemic risk assessment we calculate the dynamic correlation model with bivariate t copula distribution. We focus on the impact of sectors on the market. Despite the similarity between the time series plots of both measures, with maximum values on similar days, the compatibility of daily rankings, measured as a percentage of concordant pairs, is equal to about 50%. The rankings of the most and least risky sectors are different and depend on the choice of measure, but in the case of both we observe poor stability. When sectors are ranked in terms of the highest and lowest mean values at specific intervals (designated by the structural break estimation method, which surpisingly detects very similar dates of structural changes) we draw the same conclusions. For both measures we note huge percentage changes in mean values of risk, especially in the period from February 24, 2020 till August 20, 2020 with respect to the previous period. The percentage changes for both intervals indicate the same most risky sectors, but the indications of both measures are not consistent.

Temperature control of vibrating heat-generating hardware using spray evaporative cooling in the nucleate boiling region

A temperature control approach using evaporative spray cooling of vibrating surfaces in the nucleate boiling region is proposed and verified experimentally. This is relevant to temperature control of heat-generating automotive vehicle components. By exploiting an experimentally calibrated dynamic correlation model to represent evaporative spray cooling of a flat test-piece, a PID controller has been adopted with emphasis focused on the choice of gain parameters to ensure both stability of temperature control, and favourable responses in terms of relevant performance measures. Optimum linearisation of the correlation model has been achieved by solving an appropriate Wiener-Hopf equation, mainly to undertake a practical stability assessment of the closed-loop temperature control system. To verify the predicted control system performance, experimental measurements have been obtained from an instrumented, and spray-evaporatively-cooled, flat test-piece exposed to displacement vibration from a shaker. Experimental testing, appropriate to automotive vehicle component applications, includes large-amplitude, low frequency vibration at 12 mm and 1.9 Hz, and low amplitude, high-frequency vibration at 0.02 mm and 400 Hz. To assess the effects of different PID controller gains on the thermal performance of the thermal management system, a coefficient of performance (COP) is used, defined as the ratio of heat power removal to the required pumping power. To achieve a reduction in the settling time, and an increase in the rise time of stable control, a PID controller with a negative proportional gain showed most promising results. A 10.5% increase in COP was achieved in comparison to a PID controller with positive gains. This information is useful for the design and optimization of thermal management systems using evaporative spray cooling.

Modeling dynamic correlation in zero-inflated bivariate count data with applications to single-cell RNA sequencing data.

Interactions between biological molecules in a cell are tightly coordinated and often highly dynamic. As a result of these varying signaling activities, changes in gene coexpression patterns could often be observed. The advancements in next‐generation sequencing technologies bring new statistical challenges for studying these dynamic changes of gene coexpression. In recent years, methods have been developed to examine genomic information from individual cells. Single‐cell RNA sequencing (scRNA‐seq) data are count‐based, and often exhibit characteristics such as overdispersion and zero inflation. To explore the dynamic dependence structure in scRNA‐seq data and other zero‐inflated count data, new approaches are needed. In this paper, we consider overdispersion and zero inflation in count outcomes and propose a ZEro‐inflated negative binomial dynamic COrrelation model (ZENCO). The observed count data are modeled as a mixture of two components: success amplifications and dropout events in ZENCO. A latent variable is incorporated into ZENCO to model the covariate‐dependent correlation structure. We conduct simulation studies to evaluate the performance of our proposed method and to compare it with existing approaches. We also illustrate the implementation of our proposed approach using scRNA‐seq data from a study of minimal residual disease in melanoma.

An experimentally-verified temperature control simulation model for spray evaporative cooling of vibrating powertrain parts

A new simulation capability is presented to enable the performance of a hardware-based temperature control system to be assessed in thermally-managing heat-generating automotive vehicle powertrain parts. Temperature control is assumed to involve spray evaporative cooling of powertrain parts exposed to vibration. Two hypotheses are proposed to enable construction of a practical simulation that is both accurate and computationally efficient. The first is that a dynamic correlation model for single-nozzle spray evaporative cooling of a flat test-piece exposed to vibration, can be used as a reasonable model for multiple-nozzle spray evaporative cooling of component parts with curved cooling surfaces of non-horizontal orientation. The second is that the transient heat diffusion properties of a particular 3-dimensional component can be replaced by a 1-dimensional (1D) equivalence. To test this hypothesis, Finite Element models for two representative parts have been constructed and used to demonstrate the quality of the 1D heat diffusion equivalence, for which a fast Finite Difference solution can be exploited. To test the accuracy of test-piece surface temperature control simulation, an experimental test facility has been built in hardware, in which the temperature of two instrumented test-pieces exposed to vibration (from a shaker) are controlled by spray evaporative cooling. Each test piece is electrically-heated and the hardware control system is configured using PID control, for which appropriate gains are selected. Detailed comparisons of temperature control by hardware and simulation are given for the two test-pieces under static and dynamic conditions. Good agreement is generally obtained between simulated surface temperatures compared with measurements taken from both test-pieces. The paper shows that temperature control of a hardware-based control system using spray evaporative cooling of powertrain parts can be confidently simulated.

Study on the Measurement of Industrial Structure “Sophistication, Rationalization and Ecologicalization” Based on the Dynamic Analysis of Grey Relations — A Case Study of Beijing-Tianjin-Hebei

AbstractThe Beijing-Tianjin-Hebei region is one of the regions with the most dynamic economies, highest degrees of openness, strongest innovation capacities and largest populations, as well as an important engine driving China’s economic development. Currently, the development of the Beijing-Tianjin-Hebei region faces many difficulties, including an inefficient distribution of regional functions, imbalance in industrial structure, low sophistication, and serious overburdening of resources and the environment. Promoting the overall coordinated development of the region has become an urgent challenge. In this paper, according to the conventions of industrial structure optimization, the three core indicators of regional industrial structure rationalization, sophistication and ecologicalization are constructed. With the help of the grey dynamic correlation model, the three indicators are coupled with the economic growth rate, and the industrial structure optimization index of the Beijing-Tianjin-Hebei region is comprehensively measured. The study shows that rationalization of the industrial structure in the Beijing-Tianjin-Hebei region is increasing but that this trend is not obvious; sophistication of the industrial structure is relatively high and has also been operating at a high level; however, the overall ecological level of the industrial structure in the Beijing-Tianjin-Hebei region is not high and presents the characteristics of high in the middle and low at both ends. In terms of the relationship between industrial structure and economic growth in the Beijing-Tianjin-Hebei region, ecologicalization of the industrial structure has the greatest influence on economic growth, followed by sophistication of the industrial structure, and rationalization of the industrial structure has the weakest influence.

Multivariate Regime Switching GARCH Model Application to Portfolio and Risk Management

Regime switching dynamic correlation (RSDC) model allows the correlations to be constant with the regimes themselves however, it differs across regimes. RSDC model has retained many good properties from CCC multivariate GARCH and DCC multivariate GARCH. For example, RSDC does not suffer from the curse of dimensional due to the two steps estimation for volatility. Positive semi definite is also easy to achieve from its decomposition from covariances. In this research, the feasibility of the constant correlations assumptions have been tested using Lagrange Multiplier test developed by Tse (2000) on the RSDC model.

Nonlinearities and regimes in conditional correlations with different dynamics

New parameterizations of the dynamic conditional correlation (DCC) model and of the regime-switching dynamic correlation (RSDC) model are introduced, such that these models provide a specific dynamics for each correlation. They imply a nonlinear autoregressive form of dependence on lagged correlations and are based on properties of the Hadamard exponential matrix. The new models are applied to a data set of twenty stock market indices and a data set of the thirty Dow Jones components, comparing them to the classical DCC and RSDC models. The empirical results show that the new models improve their classical versions in terms of several criteria.

Regime switching dynamic correlations for asymmetric and fat-tailed conditional returns

A non-Gaussian multivariate regime switching dynamic correlation model for financial asset returns is proposed. It incorporates the multivariate generalized hyperbolic law for the conditional distribution of returns. All model parameters are estimated consistently using a new two-stage expectation–maximization algorithm that also allows for incorporation of shrinkage estimation via quasi-Bayesian priors. It is shown that use of Markov switching correlation dynamics not only leads to highly accurate risk forecasts, but also potentially reduces the regulatory capital requirements during periods of distress. In terms of portfolio performance, the new regime switching model delivers consistently higher Sharpe ratios and smaller losses than the equally weighted portfolio and all competing models. Finally, the regime forecasts are employed in a new dynamic risk control strategy that avoids most losses during the financial crisis and vastly improves risk-adjusted returns.

Multivariate Stochastic Volatility Model With Realized Volatilities and Pairwise Realized Correlations

Although stochastic volatility and GARCH (generalized autoregressive conditional heteroscedasticity) models have successfully described the volatility dynamics of univariate asset returns, extending them to the multivariate models with dynamic correlations has been difficult due to several major problems. First, there are too many parameters to estimate if available data are only daily returns, which results in unstable estimates. One solution to this problem is to incorporate additional observations based on intraday asset returns, such as realized covariances. Second, since multivariate asset returns are not synchronously traded, we have to use the largest time intervals such that all asset returns are observed to compute the realized covariance matrices. However, in this study, we fail to make full use of the available intraday informations when there are less frequently traded assets. Third, it is not straightforward to guarantee that the estimated (and the realized) covariance matrices are positive definite.Our contributions are the following: (1) we obtain the stable parameter estimates for the dynamic correlation models using the realized measures, (2) we make full use of intraday informations by using pairwise realized correlations, (3) the covariance matrices are guaranteed to be positive definite, (4) we avoid the arbitrariness of the ordering of asset returns, (5) we propose the flexible correlation structure model (e.g., such as setting some correlations to be zero if necessary), and (6) the parsimonious specification for the leverage effect is proposed. Our proposed models are applied to the daily returns of nine U.S. stocks with their realized volatilities and pairwise realized correlations and are shown to outperform the existing models with respect to portfolio performances.

A Flexible Regime Switching Model for Asset Returns

A non-Gaussian multivariate regime switching dynamic correlation model for fi nancial asset returns is proposed. It incorporates the multivariate generalized hyperbolic law for the conditional distribution of returns. All model parameters are estimated consistently using a new two-stage expectation-maximization algorithm that also allows for incorporation of shrinkage estimation via quasi-Bayesian priors. It is shown that use of Markov switching correlation dynamics not only leads to highly accurate risk forecasts, but also potentially reduces the regulatory capital requirements during periods of distress. In terms of portfolio performance, the new regime switching model delivers consistently higher Sharpe ratios and smaller losses than the equally weighted portfolio and all competing models. Finally, the regime forecasts are employed in a dynamic risk control strategy that avoids most losses during the fi nancial crisis and vastly improves risk-adjusted returns.

Relative Efficiency of DCC Estimates via Different Algorithms

Volatility plays an important role in capturing the variability in any series, hence multivariate volatility models helps us in figuring out the spillovers across variables. The observation of such spillovers plays a very crucial role in financial decision making. Starting from the pioneering works of [1] and [2] ARCH/ GARCH type models and their multivariate extensions are widely used for capturing volatility spillovers. Dynamic Conditional Correlation (DCC) model proposed by [3] is one of the most widely used multivariate GARCH model. As GARCH type models are non - linear in nature, hence their solutions vary across the algorithms used for their computation. Different algorithms are used in different software and thus give varying results. In this study, we are going to compute and compare DCC model estimates through different software. In our study SAS results would be considered as baseline estimates and the competing results would be obtained from Stata and R. We will model the DCC estimates between returns on NYSE and PSX indices data for the time period of September 2001 to August 2016.This study would help us to determine the relative efficiency and accuracy of algorithms used for estimating the DCC model.

Evaluation of multivariate GARCH models in an optimal asset allocation framework

This paper analyses plethora of advanced multivariate econometric models, which forecast the mean and variance-covariance of the asset returns to create optimal asset allocation models. Most existing studies use a limited number of Generalized Autoregressive Conditional Heteroscedasticity (GARCH) models. In this study, we provide an in-depth knowledge of large asset modeling and optimization strategies for solving a portfolio selection problem involving the dynamic conditional correlation models (DCC). Specifically, we use symmetric GARCH models and an asymmetric version of it (GJR-GARCH). Several studies have also tried to examine the effectiveness of using parametric copula in estimating portfolio risk measures but their results have been inconclusive. We are interested in evaluating if Copula-GARCH could be an optimal model for assessing the performance of a portfolio. This study, therefore, implemented various Copula-GARCH based models using the static and dynamic estimation of the correlation. By employing different model specifications, we are able to explore the empirical applicability of the multivariate GARCH models when estimating large conditional covariance matrices. In constructing the optimal portfolios, we evaluate the minimum variance, mean-variance, maximising Sharpe ratio, mean-CVaR, and maximization of Sortino ratio. We compare the out-of-sample performance for each of the models based on the risk-adjusted performance for a portfolio with and without short sales, consisting eight stocks and four bond indices of 10 years maturity, in the United States (US), United Kingdom (UK), Germany, Japan, Netherlands, Canada and Hong Kong. Our results suggest that the dynamic models are more capable of delivering better performance gains than the static models. These models reduce portfolio risk and improve the realized return in the out-of-sample period. This paper concludes that by adding copula functions to the model, it does not give a better performance model when compared to the dynamic correlation model.

A Class of Generalized Dynamic Correlation Models

This paper proposes a class of parametric correlation models that apply a two-layer autoregressive-moving-average structure to the dynamics of correlation matrices. The proposed model contains the Dynamic Conditional Correlation model of Engle (2002) and the Varying Correlation model of Tse and Tsui (2002) as special cases and offers greater flexibility in a parsimonious way. Performance of the proposed model is illustrated in a simulation exercise and an application to the U.S. stock indices.

Neglecting structural breaks when estimating and valuing dynamic correlations for asset allocation

ABSTRACTThis paper assesses the econometric and economic value consequences of neglecting structural breaks in dynamic correlation models and in the context of asset allocation framework. It is shown that changes in the parameters of the conditional correlation process can lead to biased estimates of persistence. Monte Carlo simulations reveal that short-run persistence is downward biased while long-run persistence is severely upward biased, leading to spurious high persistence of shocks to conditional correlation. An application to stock returns supports these results and concludes that neglecting such structural shifts could lead to misleading decisions on portfolio diversification, hedging, and risk management.

Determinantes del crédito y la morosidad en México

A short and long run analysis of the factors that influence the credit default of Mexican companies and individuals is carried out. The short-run analysis is performed using a Non-normal Asymmetric Conditional Dynamic Correlation Model that provides empirical evidence on the importance of US manufacturing volatility and the peso- dollar exchange rate on consumption volatility and local default. The paper also provides empirical evidence on the cointegrated or long-run relationship between the levels of these variables, where the US manufacture is the external variable that triggers the changes within the system formed by consumption and Mexican industry, the 91 days interest rate, the exchange rate and current and past due loans.

Testing and comparing the performance of dynamic variance and correlation models in value-at-risk estimation

This study addresses and examines certain advanced approaches for value-at-risk (VaR) estimation. In particular, we employ a multivariate generalized autoregressive conditionally heteroskedastic (MVGARCH) model involving time-varying settings and multivariate Markov switching autoregressive conditionally heteroskedastic (MVSWARCH) model with regime-switching techniques and compare them with a conventional linear regression-based (LRB) model. Our empirical findings are as follows: First, while the LRB VaR model behaves reasonably well in tranquil periods, it significantly underestimates actual risk during unstable periods. Second, in comparison with the LRB VaR model, MVGARCH- and MVSWARCH-based VaR models do better under unusual conditions, whereas better models are needed to estimate VaR. Third, dynamic variance settings improve the accuracy of VaR estimates. However, the effect of dynamic correlation designs on VaR is marginal.This study addresses and examines certain advanced approaches for value-at-risk (VaR) estimation. In particular, we employ a multivariate generalized autoregressive conditionally heteroskedastic (MVGARCH) model involving time-varying settings and multivariate Markov switching autoregressive conditionally heteroskedastic (MVSWARCH) model with regime-switching techniques and compare them with a conventional linear regression-based (LRB) model. Our empirical findings are as follows: First, while the LRB VaR model behaves reasonably well in tranquil periods, it significantly underestimates actual risk during unstable periods. Second, in comparison with the LRB VaR model, MVGARCH- and MVSWARCH-based VaR models do better under unusual conditions, whereas better models are needed to estimate VaR. Third, dynamic variance settings improve the accuracy of VaR estimates. However, the effect of dynamic correlation designs on VaR is marginal.

Correlated Idiosyncratic Volatility Shocks

Commonality in idiosyncratic volatility cannot be completely explained by time-varying volatility. We decompose the common factor in idiosyncratic volatility (CIV) of Herskovic et al. (2016) into two components: idiosyncratic volatility innovations (VIN) and time-varying idiosyncratic volatility (TVV). VIN is priced in the cross section of stock returns, whereas TVV is weakly priced. A long-short strategy based on double-sorted VIN and TVV portfolios earns average returns of 8.0% per year. To capture the commonality in idiosyncratic volatility, we propose the Dynamic Factor Correlation model, which outperforms Engle’s (2002) DCC model in simulations and empirical tests.