Parameter Vector Research Articles

Mode combinability: Exploring convex combinations of permutation aligned models

We explore element-wise convex combinations of two permutation-aligned neural network parameter vectors ΘA and ΘB of size d. We conduct extensive experiments by examining various distributions of such model combinations parametrized by elements of the hypercube [0,1]d and its vicinity. Our findings reveal that broad regions of the hypercube form surfaces of low loss values, indicating that the notion of linear mode connectivity extends to a more general phenomenon which we call mode combinability. We also make several novel observations regarding linear mode connectivity and model re-basin. We demonstrate a transitivity property: two models re-based to a common third model are also linear mode connected, and a robustness property: even with significant perturbations of the neuron matchings the resulting combinations continue to form a working model. Moreover, we analyze the functional and weight similarity of model combinations and show that such combinations are non-vacuous in the sense that there are significant functional differences between the resulting models.

Dynamic interlinkages between carbon risk and volatility of green and renewable energy: A TVP-VAR analysis

Our paper applies a time-varying parameter vector autoregression (TVP-VAR) in combination with an extended joint connectedness approach to investigate interlinkages among carbon emissions futures and the volatility of the renewable energy sector. The findings show that the system-wide dynamic connectedness realized a peak in early 2020 in the wake of the COVID-19 crisis. Net total directional connectedness proves that carbon emissions futures and wind energy play the roles of both net transmitters and net receivers of shocks in both periods – before and after the pandemic. The findings of this paper can support policy formulations to avoid rapid fluctuations in carbon prices, make the carbon price table, and limit the negative effect of carbon risk on the energy market, while promoting the protection of systemic financial risks in the renewable energy sector and ensuring a green energy supply.

Market volatility and crisis dynamics: a comprehensive analysis of U.S., China, India, and Pakistan stock markets with oil and gold interconnections during COVID-19 and Russia–Ukraine war periods

The objective of this paper is to explore the interconnectedness of volatility among the stock markets of U.S., China, India, and Pakistan in conjunction with oil and gold markets. Employing the novel Time-Varying Parameter Vector Autoregression (TVP-VAR) model for assessing connectedness, the study scrutinizes key patterns of dependency and interrelation between these markets. Furthermore, this study investigates the dynamic connectedness during the global health crisis due to COVID-19 and the geopolitical crisis due to Russia–Ukraine war periods to identify the changes in their relationship following the two crises episodes. The findings underscore the significance of volatility transmissions emanating from the U.S., a developed market, in shaping these dynamic linkages. It is observed that oil and gold returns play a limited role as sources of shocks for market returns in China, India, and Pakistan, suggesting a relatively lower contribution of oil and gold to equity market volatility. The results also emphasize the safe-haven characteristics of gold during periods of crisis such as the COVID-19 pandemic and the Russia–Ukraine war. Moreover, the study indicates that the volatility transmissions during the COVID-19 pandemic are more pronounced compared to the Russia–Ukraine war crisis. These findings hold notable implications for both investors and policymakers, emphasizing the need for a nuanced understanding of market dynamics and the development of risk-averse strategies, particularly in times of crisis.

Interlinkages across US sectoral returns: time-varying interconnectedness and hedging effectiveness

This study examines the time-varying asymmetric interlinkages between nine US sectoral returns from January 2020 to January 2023. To this end, we used the time-varying parameter vector autoregression (TVP-VAR) asymmetric connectedness approach of Adekoya et al. (Resour Policy 77:102728, 2022a, Resour Policy 78:102877, 2022b) and analyzed the time-varying transmitting/receiving roles of sectors, considering the positive and negative impacts of the spillovers. We further estimate negative spillovers networks at two burst times (the declaration of the COVID-19 pandemic by the World Health Organization on 11 March 2020 and the start of Russian-Ukrainian war on 24 February 2022, respectively). Moreover, we performed a portfolio back-testing analysis to determine the time-varying portfolio allocations and hedging the effectiveness of different portfolio construction techniques. Our results reveal that (i) the sectoral return series are strongly interconnected, and negative spillovers dominate the study period; (ii) US sectoral returns are more sensitive to negative shocks, particularly during the burst times; (iii) the overall, positive, and negative connectedness indices reached their maximums on March 16, 2020; (iv) the industry sector is the largest transmitter/recipient of return shocks on average; and (v) the minimum correlation and connectedness portfolio approaches robustly capture asymmetries. Our findings provide suggestions for investors, portfolio managers, and policymakers regarding optimal portfolio strategies and risk supervision.

The time‐varying volatility spillover effects between China's coal and metal market

AbstractThis study employs a time‐varying parameter vector autoregression methodology with the Diebold and Yilmaz spillover index to scrutinize the temporal fluctuations in volatility spillovers between the Chinese coal and metal markets. The analysis is conducted from the dual perspectives of security indices and futures prices. The findings reveal a robust correlation between the coal and metal markets, with the coal market serving as a primary conduit for volatility spillover into the metal market. Furthermore, this study investigates the time‐specific impacts of coal decommissioning policies, the COVID‐19 pandemic, and the coal supply crisis on the coal–metal market volatility spillovers. The findings indicate that these three unique shocks significantly increase the overall risk spillover index between the coal and metal markets. Moreover, during these exceptional events, the extent or role of risk spillover in the coal–metal market undergoes varying degrees of change. On the basis of these findings, this article presents pertinent policy recommendations.

On the propagation mechanism of international real interest rate spillovers: evidence from more than 200 years of data

ABSTRACT This article analyzes the real interest rate transmission mechanism across the United States, Japan, France, Germany, Holland, Italy, Spain and the United Kingdom over a period of more than 200 years. Based on a time-varying parameter vector autoregressive (TVP-VAR) connectedness methodology, the empirical results suggest that the size of the international real interest rate spillovers ranges between 30% and 75% across the sample period. Furthermore, it is shown that these spillovers increase during crisis periods, such as the two World Wars, the Great Depression of 1929, the 1980 and 1990 recessions, or the Great Financial Crisis of 2009. More interestingly, our findings illustrate the position of each of these eight countries as net transmitters or receivers of monetary policy shocks over time. Our analysis contributes to the debate on whether the conduct of monetary policy in a country should consider its international spillovers, suggesting the need for more coordinated monetary policies especially during periods of economic crises.

Simulation model parameter optimization method for multidimensional signals

The paper describes an approach to minimize the number of simulation experiments of multidimensional signals by means of a regression neural network. Multivariate signal simulation systems are in demand for testing real-time computing systems, but they mostly have a wide vector of model parameters. The formation of parameter vector for the simulation model of multidimensional signals, which ensures an adequate solution to the problem of signal processing obtained as a result of the model operation, is an urgent task. The authors propose a method of heuristic optimization of input parameters, implemented by means of machine learning, which reduces the searching of values for the given optimization function. In this study the ability to perform real-time signal simulations is the optimization feature. The research outcomes are the description of the neural network, the rationale for its configuration, and the training and validation results.

Continuity of the solution set mappings to parametric unified weak vector equilibrium problems via free-disposal sets

In this paper, the lower and upper semicontinuity of the solution mappings for parametric unified weak vector equilibrium problem (PUWVEP) are established, via freedisposal set and nonlinear scalarization function. Moreover, example is given to illustrate the obtained results. The results improve the corresponding ones in the literature [12-14].

A vibration-based Machine Learning type Structural Health Monitoring methodology for populations of composite aerostructures under uncertainty

A robust to uncertainty Machine Learning (ML) based Structural Health Monitoring methodology for populations of composite aerostructures is postulated. The methodology is founded upon a number of unsupervised ML algorithms for damage detection and a supervised counterpart for damage characterization. Damage detection is specifically based on two types of Healthy Subspace representations: A Multiple Model (MM) and a varying radii Hyper-Sphere (HS) type. Both are built upon response-only vibration acceleration and/or strain signals at properly selected sensor locations. Based on them, Multiple Input Single Output (MISO) Transmittance Function AutoRegressive with eXogenous (TF-ARX) excitation data driven models representing the partial structural dynamics are obtained. Decision making is then based on the model parameter vector that may be transformed and reduced via Principal Component Analysis (PCA). Damage detection is achieved via multi-level information fusion using acceleration and/or strain sensors. Damage characterization, referring to damage type, location, and level determination, is achieved via a hierarchical cosine similarity based algorithm. The methodology is successfully assessed via hundreds of experiments using a population of small-scale composite coupons for the detection and characterization of Delamination and Impact damage under material/manufacturing, temperature, excitation, and experimental uncertainty.

Parameter estimation for Gegenbaeur Arfisma processes with infinite variance innovations

. In this study, we consider the Gegenbauer ARFISMA process with α-stable innovations which belong to the class of infinite variance time series. This is a finite parameter model that exhibits long-range dependence, high variability, and Seasonal and /or Cyclical Long Memory (SCLM) variations. The Gegenbauer ARFISMA time series with α-stable innovations can be rewritten as linear processes (infinite order moving averages) with coefficients that decay slowly to zero and with innovations that are in the domain of attraction of a α-stable distribution with stability parameter ( 1 < α ≤ 2 ) . We assume that the coefficients of the linear process depend on an unknown parameter vector ξ which should be estimated from samples of length n. We prove that under some assumptions, a Whittle-type estimator ξ n for ξ is consistent (ξ n converges to the true value ξ 0 in probability as n ⟶ + ∞ ), and, we derive the limiting distribution of the resealed differences ( n / log ⁡ ( n ) ) 1 / α ( ξ n − ξ 0 ) . We perform some simulations to illustrate the consistency of the Whittle-type estimator ξ n for ξ.

Return and volatility transmission among economic policy uncertainty, geopolitical risk and precious metals

Purpose This paper aims to examine the return and volatility transmission among economic policy uncertainty (EPU), geopolitical risk (GPR), their interaction (EPGR) and five tradable precious metals: gold, silver, platinum, palladium and rhodium. Design/methodology/approach Applying time-varying parameter vector autoregression (TVP-VAR) frequency-based connectedness approach to a data set spanning from January 1997 to February 2023, the study analyzes return and volatility connectedness separately, providing insights into how the data, in return and volatility forms, differ across time and frequency. Findings The results of the return connectedness show that gold, palladium and silver are affected more by EPU in the short term, while all precious metals are influenced by GPR in the short term. EPGR exhibits strong contributions to the system due to its elevated levels of policy uncertainty and extreme global risks. Palladium shows the highest reaction to EPGR, while silver shows the lowest. Return spillovers are generally time-varying and spike during critical global events. The volatility connectedness is long-term driven, suggesting that uncertainty and risk factors influence market participants’ long-term expectations. Notable peaks in total connectedness occurred during the Global Financial Crisis and the COVID-19 pandemic, with the latter being the highest. Originality/value Using the recently updated news-based uncertainty indicators, the study examines the time and frequency connectedness between key uncertainty measures and precious metals in their returns and volatility forms using the TVP-VAR frequency-based connectedness approach.

Parameter estimation of multivariable Wiener nonlinear systems by the improved particle swarm optimization and coupling identification

This paper investigates the parameter estimation of multivariable Wiener nonlinear systems. To solve the inconsistency problem of the parameter vector and the parameter matrix, the coupling identification concept is applied. Combined with particle swarm optimization (PSO) and an auxiliary model, the partially coupled improved particle swarm optimization (PC-IPSO) method is proposed. In this algorithm, the adaptive feedback inertia weight is improved to accelerate the convergence speed, and the retirement update mechanism is introduced to improve the optimization ability of the basic PSO algorithm. To verify the performance of PC-IPSO, we also derive a multivariable improved PSO (M-IPSO) method for comparison. The computational complexity analysis shows that the PC-IPSO algorithm requires less computational resources than the M-IPSO algorithm. Then, the convergence of the improved PSO method is analyzed. The simulation results indicate that the PC-IPSO method has a faster convergence speed and higher identification accuracy than the M-IPSO and several existing state-of-the-art methods for multivariable Wiener system identification.

Fiscal policy and dimensions of inequality in South Africa: A time‐varying coefficient approach

AbstractSouth Africa continues to face high inequality levels despite its progressive tax and extensive social protection systems. We compare the dynamic impact of fiscal policy on the distribution of incomes, wages and wealth in South Africa from 1993 to 2019. For this purpose, we use a time‐varying parameter vector autoregression to estimate the impact of direct tax revenue and total transfer spending on three distinct inequality datasets. The analysis of various dimensions of inequality is the main contribution of the paper as the literature typically focuses on income inequality. A second contribution lies in the incorporation of time‐varying effects, which enables the analysis of the changing relationship between fiscal policy and inequality. The results suggest that this relationship is indeed time‐varying and that the impact of direct taxes and transfers differs markedly across the inequality dimensions, both in terms of magnitude and sign. Overall, we find that both transfers and direct taxes have not significantly reduced income, wage or wealth inequality in South Africa.

Spillover effects between fossil energy and green markets: Evidence from informational inefficiency

This paper builds an indicator of the market inefficiency degree by quantifying the extent of the market deviation from an efficient status. From a novel perspective of informational inefficiency, we further model the dynamic spillover effects between fossil energy and green markets by using a time-varying parameter vector autoregressive (TVP-VAR) model. Our findings indicate that the market inefficiency degree and its spillovers exhibit evident time-varying dynamics and are sensitive to extreme events. The international crude oil benchmarks serve as the most significant information transmitter and receiver in the spillover network of market inefficiency. Within green-related markets, carbon market demonstrates the strongest correlation with the fossil energy markets. Additional analyses involving consideration of economic policy uncertainty reassure the robustness of our findings.

Impact of Geopolitical Risk on G7 Financial Markets: A Comparative Wavelet Analysis between 2014 and 2022

This study investigates co-movements between the GPR generated by the Crimean Peninsula’s annexation in 2014, the Russia–Ukraine war in 2022, and the volatility of stock markets in the G7 states. Using wavelet analysis, concentrated co-movement was found for all indices in both periods. Contrary to the general perception, we find that the G7 financial market response in 2014 was robust. Using a time-varying parameter vector autoregression (TVP-VAR) test, we found a larger reaction in the amplitude of the G7 financial markets in 2022 than in 2014. The financial markets in France, Germany, and the UK showed a similar reaction in 2022. We have identified some common aspects, even if the political and military contexts of the two studied events were completely different. Our findings offer new and interesting implications for understanding how geopolitical risk affects financial assets for market participants with multiple investment horizons and strategies.

Unveiling hidden connections: Spillover among BRICS' cryptocurrency-implied exchange rate discounts and US financial markets

This study employs connectedness approach based on Time-Varying Parameter Vector Autoregression (TVP-VAR) and Dynamic Conditional Correlation Generalized Autoregressive Conditional Heteroskedasticity (DCC-GARCH) methods, to scrutinize the connectedness between the cryptocurrency implied exchange rate discount in BRICS countries (Brazil, Russia, India, China, and South Africa) and US financial markets. Empirical evidence suggests that Bitcoin market volatility substantially influences other markets. Moreover, given the intimate connections between BRICS countries and developed nations such as the US in today's globalized context, fluctuations in the Bitcoin market within BRICS countries are apt to impact US financial markets. Our empirical findings specifically indicate the following: (i) Among the total spillover effects of the cryptocurrency implied exchange rate discount in BRICS countries, Russia exhibits the most potent total spillover effect and contributes the most to market connectedness. (ii) The Brazil directional spillover index is elevated in both spillover relationship categories, and the spillover effect of the cryptocurrency implied exchange rate discount in Brazil on US financial variables is notably pronounced. (iii) Of the US financial variables, the US Financial Stress Index demonstrates the most robust connectedness with the cryptocurrency implied exchange rate discount in BRICS countries and exhibits heightened sensitivity to uncertainty.

Tail risk connectedness in G7 stock markets: Understanding the impact of COVID-19 and related variants

This study uses a time-varying parameter vector autoregression (TVP-VAR) approach to examine the transmission of tail risk among G7 stock markets from January 2000 to September 2022, focusing on major financial episodes like the dot-com collapse, the 2008 Global Financial Crisis, and the European debt crisis, as well as the COVID-19 pandemic and its variants. Our analysis shows fluctuating tail risk connectedness across G7 markets during the pandemic, influenced by lockdowns, supply chain issues, interventions, and investor sentiment. Notably, the UK and Italy were major tail risk transmitters, whereas Japan predominantly absorbed risk, highlighting its distinct vulnerability. The findings stress the need for a comprehensive understanding of tail risk dynamics to inform decisions by investors, governments, and regulators, ensuring financial system resilience during heightened economic stress.

Time-varying relationship between international monetary policy and energy markets

This paper examines the spillover dynamics and connectedness between international monetary policies of four developed economies (i.e., UK, US, Japan and Eurozone) and energy markets (i.e., crude oil, heating oil, gasoline and propane) while accounting for the impact of global volatility, economic uncertainty and geopolitical risk factors. This paper uses the time-varying parameter vector auto-regression (TVP-VAR) and TVP-VAR-based extended joint connectedness models to explore the connectedness of monetary policy and energy markets using daily data from June 1, 2007, to March 31, 2022. The results reveal a time-varying connectedness between the energy markets and international monetary policy, and global events affect the magnitude of connectedness. In all, energy markets were found to be the major shock transmitters and monetary policy the receiver. Additionally, we found that energy dependence explains why the Euro Area shadow short rate is more related to energy markets than other monetary policies. The findings also show “Oil to monetary policy” risk spillover, which suggests oil price controls monetary policy. Finally, economic uncertainty positively affects monetary policy and energy price connectedness.

Adaptive time-varying state safety constraint control for discrete-time nonlinear systems with random actuator failures

This work investigates the adaptive time-varying state safety constraint control problem for a class of discrete-time nonlinear systems subject to large parameter uncertainties and stochastic actuator failures for the first time. The existence of time-varying full-state constraints, large parameter uncertainties and stochastic actuator failures leads to a series of difficulties for realizing all control objectives. Due to the absence of the linearity property of the derivative in discrete-time log type barrier Lyapunov functions, they are not used to deal with the state constraint problem of nonlinear discrete-time systems. The discrete quadratic-fraction time-varying barrier Lyapunov function is employed to guarantee that all states are constrained in the time-varying predefined regions. By virtue of the constructed multiple-model set, an adaptive identification algorithm is developed to deal with large parameter uncertainties and it has a clear advantage that this method does not requires cn (where n is the dimension of the parameter vector and c is an integer) identification models and only relies on n+1 identification models, thereby heavily reducing the computational burden. By employing the discrete-time filter and backstepping techniques, the adaptive failure compensation controller is designed to ensure the boundedness in probability of the considered systems with respect to stochastic actuator failures. Finally, the effectiveness of the proposed scheme is proved by two simulation examples.

INFERENCE IN PARTIALLY IDENTIFIED PANEL DATA MODELS WITH INTERACTIVE FIXED EFFECTS

In this paper, we develop methods for statistical inferences in a partially identified nonparametric panel data model with endogeneity and interactive fixed effects. Under some normalization rules, we can concentrate out the large-dimensional parameter vector of factor loadings and specify a set of conditional moment restrictions that are involved with only the finite-dimensional factor parameters along with the infinite-dimensional nonparametric component. For a conjectured restriction on the parameter, we consider testing the null hypothesis that the restriction is satisfied by at least one element in the identified set and propose a test statistic based on a novel martingale difference divergence measure for the distance between a conditional expectation object and zero. We derive a tight asymptotic distributional upper bound for the resultant test statistic under the null and show that it is divergent at rate-N under the global alternative. To obtain the critical values for our test, we propose a version of multiplier bootstrap and establish its asymptotic validity. Simulations demonstrate the finite sample properties of our inference procedure. We apply our method to study Engel curves for major nondurable expenditures in China by using a panel dataset from the China Family Panel Studies.