The high-quality development of the express delivery industry (EDI) is an important guarantee for the efficient operation of the modern service industry and an important support for economic and social development.The outbreak of the COVID-19 at the end of 2019 has brought an unprecedented impact on economic and social development,and has also had a lot of impact on the EDI,but how and how big is the impact,prevention and control (P&C) policies,and what are the differences in the impact of different regions are questions that need to be answered.Based on the monthly scale,income and complaint data of EDI in 31 provinces and cities across China,the fixed effect model was used to analyze the direction and degree of the impact of COVID-19,its P&C measures,the stage of epidemic P&C,and the degree of epidemic damage on the development speed and quality of EDI.The results show that:1) The COVID-19 did not affect the growth of EDI,but promoted the development of the EDI in the short term,and reduced complaints such as delay and delivery.2) Strict P&C measures are detrimental to the development of scale∈come and service quality of EDI.3) The positive impact of the epidemic on the scale∈come of EDI is mainly reflected in the first stage (January-February) after the outbreak of COVID-19,and the areas that are severely affected by the epidemic are most prominent.The conclusions of this study have important practical significance for the EDI to deal with public emergencies and improve emergency management abilities.

Obviously,the promotion of total factor productivity of enterprises is the driving force for the high-quality development of China's economy.Under the background of national emission reduction strategy,whether the carbon emission trading pilot policy,as a market incentive environmental regulation means,can influence the total factor productivity of enterprises by triggering the"Porter effect "is a topic worthy of in-depth study.At present,there is no empirical evidence related to Chinese practice.On the premise of adopting difference-in-differences to design the quasi-natural experiment of carbon emission trading pilot policy,this paper aims to identify the impact of carbon emission trading pilot policy on total factor productivity and the corresponding mechanism from the macro-region and micro-enterprise dimensions.Based on the foregoing analysis,this study finds that the carbon emission trading pilot policy is significantly beneficial to the improvement of total factor productivity of enterprises.Moreover,after a series of robustness tests including overcoming sample selection biases and alleviating the endogenous problems as well as eliminating other policy interference,the conclusions of this study are still valid.According to the further analysis,it indicates that this policy can improve the total factor productivity of enterprises by stimulating enterprise innovation and optimizing the allocation of resource elements.Meanwhile,the heterogeneity analysis indicates that the pilot policies have a more significant effect in promoting the total factor productivity of enterprises in the western region,state-owned enterprises and industries focused on by the policy as well as areas with high law enforcement.Briefly,this paper provides valuable empirical evidence and policy reference for implementing carbon emission trading policies according to local conditions and promoting the development of economy in the form of" low carbon "and" high quality".

While giving rise to the transformation of the operation patterns of the manufacturing industry,shared manufacturing brings new opportunities as well as challenges to the theories and methods of production operations and management.The existing literature conducts comprehensive reviews on the cloud manufacturing techniques that shared manufacturing relies on and the research status of capacity allocation of participating enterprises.However,there is a lack of surveys on the progress of operation management of shared manufacturing platform.This study first analyzes real cases of shared manufacturing,and summarizes the typical characteristics and classification from the perspective of sharing economy and platform operation.Then,this paper summarizes the evaluation index system of supply-demand matching and gives a comprehensive review on the operational optimization,matching and scheduling models and solution algorithms of two types of supply-demand matching and scheduling research,namely "order-capacity selection" and "order-capacity mutual selection".Finally,we compare the shared manufacturing with the cloud computing and classical production scheduling problems and discuss the future studies.

Variance risk premium reflects the degree of risk aversion of investors and their overestimation of extreme risk loss. This paper decomposes the variance risk premium of China's option market into upside and downside variance risk premium for the first time, discusses its return predictability, and tests the variance risk premium by constructing delta hedging portfolio. The results show that the variance risk premium is significantly positive and the return of delta hedging portfolio is significantly negative, which confirms the fact that Chinese investors are risk averse and willing to pay a premium for hedging variance risk. The results also show that different from the mature market, the variance risk premium of Chinese market lacks the ability to predict the market return. Both downside variance risk premium and upside variance risk premium have no significant predictive power. Our research enriches the understanding of investors' behavior in China's option market.

We aim at reducing the power usage effectiveness (PUE) metric of data centers with machine learning methods.At the present stage,industry such as Google considers only a small number of features and the impact of a single variable on the PUE metric,lacking the analysis of the coupling characteristics between features.In addition,machine learning methods have high requirements on the quality and quantity of data,and it is easy to be interfered by signal noises when implementing machine learning methods in practice.There are few specific optimization cases in both academia and industry at this stage.In this paper,we improve the current method of optimizing the PUE metric using neural networks by increasing the feature dimension,so as to increase the prediction accuracy,which is higher than that of the PUE predicting model built by Google.We use statistical methods to approximate the coupling characteristics between features with historical samples,and integrate them into sensitivity analysis to obtain more accurate results.We propose a cooling system parameter setting method based on sensitivity analysis.Based on the data and infrastructure in a Tencent Data Center located in Tianjin,we implement an experiment of parameter optimization of the cooling system,and the effectiveness of the proposed method is demonstrated by the experimental results.

How could we identify systemic and understand the risk contagion pattern in our financial market? Therefore, to quantify financial systemic risk indicator for systemic risk identification and early warning becomes an important research question. In this paper, the long-run equilibrium, index connectedness and spillover are studied with Chinses stock, bond, fund, commodity, money market and foreign exchange market. Risk contagion pattern is studied to conclude that the stock and fund market plays the key role in affecting other markets, the key focus of controlling systemic risk would be these two markets. Furthermore, we construct systemic risk index by time rolling technique for systemic risk identification and early warning signal. We conclude that the systemic risk is under control. Further analysis is done on employing online lending market, we find that it has little impact on the finical market risk.

The recession of globalization brings about the layout adjustment of global value chains and seriously threatens the security of supply chains.The COVID-19 pandemic aggravates various uncertainties and greatly affects the operations of supply chains.Meanwhile,emerging technologies such as digitalization are propelling business innovations and providing new opportunities for the value creations of supply chains.Operations optimization is committed to facilitating efficient collaborations for achieving supply chain coordination.The interweaving of various variables in great changes further catalyzes the reinvention of supply chains towards platform-based ecosystems.The ecosystem promotes transboundary integrations of multi-entities by opening and sharing resources,builds ecological balances of the system by operations coordination and mutual benefit,energizes risk management and green operations by intelligentization,stimulates non-boundary innovations by structure reorganization and pattern transformation,and realizes the value co-creations of participators.Therefore,it is necessary to strengthen the philosophy of "community of interests",actuate resource sharing,risk sharing,mutual trust and benefit distribution through mechanism constructions,enhance ecological concept and value recognition,improve platform operations management,optimize ecosystem governance,create an open and innovative ecology,upgrade crisis management capability,and build ecosystems for enabling the high-quality development of supply chains.

Based on the 268-dimensional monthly data of China from 2002 to 2019,this paper measures China's policy uncertainty,economic uncertainty and financial uncertainty respectively,and further studies their correlation mechanism and influence dynamics by using dynamic spillover index and STVAR model.The results show that policy uncertainty will not only aggravate economic uncertainty and financial uncertainty,but also be positively affected by economic uncertainty and financial uncertainty.The spillover effect among policy uncertainty,economic uncertainty and financial uncertainty is affected by the degree of uncertainty,the level of economic development and the intensity of financial friction,which shows a significant nonlinear characteristic.For example,the aggravating effect of policy uncertainty on economic uncertainty and financial uncertainty obviously increases with the increase of uncertainty level and financial friction intensity,but it can be effectively alleviated in the period of economic prosperity.These findings provide useful empirical evidence and policy insights for a deeper understanding of the theoretical implications of macro uncertainty,a reasonable control of uncertainty levels in the economy,and a practical improvement of China's macro-control and financial regulatory policy framework.

This paper proposes a new EEMD-LSSVR-K-based decomposition-clustering-ensemble learning approach for foreign exchange rates forecasting by integrating ensemble empirical mode decomposition (EEMD),least square support vector regression (LSSVR) and K-means clustering algorithm.Clustering strategy is used to extend the fixed-weighted meta-synthetic in decomposition-ensemble learning approach to weighted with local data characteristics meta-synthetic.Our proposed approach can effectively solve the shortcoming of fixed-weighted meta-synthetic in decomposition-ensemble learning approach.Meanwhile,our proposed approach is applied to four type exchange rates forecasting.The empirical results show that our proposed approach significantly improves the level and directional accuracy of exchange rates forecasting,and verify the importance of clustering strategy in decomposition-ensemble learning approach.

Since a single-layer network is often difficult to comprehensively measure the complex connectedness of the financial system, we study the asset portfolio optimization from the perspective of multilayer connectedness networks. First, we build multilayer connectedness networks (including linear connectedness layer, nonlinear connectedness layer, partial connectedness layer and tail connectedness layer) of sample stock assets. Then, we analyze the structure of multilayer connectedness networks and examine the correlation between the stock weights in the portfolio and node centralities in the network. On this basis, we propose two stock selection strategies based on the network models: ρ investment strategy and the low centrality investment strategy. Finally, we use the rolling window method to simulate the dynamic investment process, and employ Sharpe ratio, turnover rate, the breakeven transaction cost and alpha indicator to evaluate the portfolio performance. The empirical results reveal that: 1) there exist similarity and uniqueness between each layers, and there is a negative correlation between stock weights and node centralities; 2) the return of the proposed network-based portfolios is higher than that of the benchmark portfolio, and this high return is not offset by the large systematic risk factor exposure or transaction costs; and 3) the equal-weighted portfolio of the low centrality strategy has the best performance, providing investors with a significantly higher Sharpe ratio than the benchmark portfolio.

This paper investigates a multi-channel formation mechanism of systemic risk under the superposition of insolvency and liquidity exhaustion, in which debt defaults, liquidity runs and fire sales penetrate and strengthen each other. The paper dynamically simulates the evolution process of risk contagion under the impact of real estate loan defaults, and carries out an empirical analysis on 38 banks from 2012 to 2020. It's found that the banking system can fully absorb the initial impact of real estate loan defaults, and systemic risks mainly generate in the process of multi-channel risk contagion caused by bank deleveraging. Fire sales have significant impacts on the formation of systemic risk, while the impacts of inter-bank market debt defaults and liquidity runs are limited. Insolvency and liquidity exhaustion are the main forms of risk contagion, and the superposition effects continue to amplify systemic risk. Liquidity exhaustion is more serious, but it has been significantly improved, while insolvency becomes more prominent year by year. The release of systemic risk has "immediate effect" under severe impact, "continuity effect" under moderate impact, and "self-recovery" under mild impact. The types of risks in the banking system tend to be complicated. State-owned commercial banks and joint-stock commercial banks are more prone to the dual risks of liquidity exhaustion and insolvency, and vulnerability and risk contribution are more prominent. Urban commercial banks and rural commercial banks mostly show double risks or single risk. Based on the research conclusions, the paper provides feasible policy suggestions for China's banking systemic risk prevention and financial supervision.

E-commerce platform plays an important role in Chinese economy and industrial structure transformation. Based on characterizations of operations and crisis subsideis of e-commerce platform, from the perspective of financial subsidies, we build a theoretical model to analyze the optimal strategies and benefits of e-commerce platform subsidies, in order to describe the differences and interactions among different platform subsidies. We find that:1) For single subsidy strategy, when the crisis is relatively serious, only the "blood transfusion" type strategy can help the retails, such as provision of loans; by contrast, when the crisis is relatively moderate, the "haemospasia" type strategy is more effective, such as exemotion or delay in collecting usage fees; 2) For the mixed subsidy strategies, the benefit of platform and the scale of crisis subsidies increase (decrease) with the degree of the severity of crisis when the platform has no (has) capital constraints; 3) Compraring to outside financial institutions such as banks, e-commerce platform is more efficient:Considering the symbiosis relationship, platform is willing to sacrifice short-term revenues in exchange for long-term income. Our study provides a framework to analyze strategies and benefits of e-commerce platform subsidies as well as a new perspective to understand how the e-commerce platform provide efficient financial subsidies due to the symbiosis relationship.

China is now facing the double pressure of economic downturn brought by COVID-19 and low-carbon transition. The trade-off between short-term economic recovery and long-term green development makes it necessary to design the economic recovery policies under multiple objectives. This paper constructs a new Keynesian dynamic stochastic general equilibrium model, and analyses the process from the outbreak of the COVID-19 epidemic to its economic impact and then to government intervention. The results show that 1) the short-term economic recovery effects of all three economic stimulus policies are remarkable. Under the green economic stimulus policy, GDP grew by 3.3%、5.3% and 6% in the second, third and fourth quarter of 2020. 2) the green economic stimulus policy reduced economic fluctuations, thus acting as an automatic stabilizer and contributing to a stable economic recovery. 3) in the long run, the green economic stimulus policy is conducive to achieving a green transition of the economy, and avoiding a high carbon lock-in effect in the future. More importantly, it is the preferred path to achieve the 2060 carbon neutrality target. We estimate that each 1 CNY increase in green investment will reduce future abatement costs by 1.5~2.6 CNY.

Gold has the dual property of commodity and currency, which is an important means for investors to maintain and increase the value of assets. Based on the behavioral finance theory, this paper adopts the generalized autoregressive conditional heteroscedasticity mixed data sampling model (GARCH-MIDAS) to explore the predictive ability of Baidu index and Google trend to the Chinese gold price volatility. At the same time, global economic policy uncertainty (GEPU) index and geopolitical risk (GPR) index variables are introduced to test the impact on gold volatility. Furthermore, two evaluation methods, model confidence set (MCS) and direction-of-change (DoC), are used to test the out-of-sample prediction accuracy of each model. This study employs the closing price data of Au (T+D) gold deferred trading contract with large trading volume as the research object. The sample interval from January 4, 2011 to December 31, 2019, and a total of 2,186 daily data are obtained from the CSMAR database. The monthly data of Baidu index is from http://index.baidu.com, and the monthly data of Google trends is from https://trends.google.com/trends/. We obtain the monthly data of GEPU and GPR from http://www.policyuncertainty.com. Based on the empirical results of the MCS and DoC tests, the GARCH-MIDAS-Google model performs the best predictive power than other competing models, which means Google trends contain more useful information for the Chinese gold price volatility. Moreover, our results are robust to different forecasting window. Therefore, our findings provide a new perspective for predicting China's gold volatility, provide a reliable guarantee for stable financial stability, and provide valuable information for policy makers.

Trade liberalization and capital globalization are important reflection of economic globalization. The transnational flow of production factors, especially capital, has becoming more and more frequent, making it difficult for the commonly used accounting method based on trade volume or value-added in trade to distinguish the ownership of trade economic benefits. With the compilation of non-competitive input-output tables which distinguish domestic-invested and foreign-invested enterprises, processing and non-processing trade, and reveal the ownership of production factors, this paper revaluates the national income in China's exports along the time, sector and trading partner dimensions. Further, foreign income in trade is computable and foreign capital penetrance can be established. This study is of great significance for understanding the development and characteristics of China's exports.

This paper constructs an DSGE model with bank run mechanism and monetary policy uncertainty (MPU) shock, and measures China's MPU based on the monetary policy rule with stochastic volatility. We further explore the impacts of MPU shock on commercial banks' risk-taking and loan activities, and its impacts on macro-level credit risk and real activities by SVAR model. We find that that: The increase of MPU will exacerbate banks' non-performing loan ratio, which will push up the risk-taking level of banks. And the rise of MPU will compress the loan activities of banks. MPU shock can negatively affect the real economic activities through raising banks' credit risks. This paper provides some policy implications for our central bankers to optimize monetary policy operation to prevent financial risk and advance the economic works such as "stabilizing finance" and "stabilizing expectation".

In order to effectively solve the dynamic vehicle routing problem of heterogeneous fleets, customers' dynamic requests and real-time variations in travel times between nodes in distribution networks, a two-stage mathematical model with the goal of minimizing distribution costs was established based on the idea of pre-optimization and dynamic adjustment in this paper. In the pre-optimization stage, an improved adaptive genetic algorithm is designed to gain the initial distribution scheme; In the dynamic adjustment stage, comprehensively consider the customers' dynamic requests and the speed of the distribution networks, formulate an optimization strategy that combines continuity and periodicity, and turn the problem into multi-depot vehicle routing problem for solution. The effectiveness of the model and algorithm is verified by example analysis. The research results can enrich the relevant research on vehicle routing problem and provide theoretical basis for logistics enterprises to optimize realistic distribution schemes.

This paper applied network analysis method to the world multi-regional input-output model, and used the shortest path algorithm to describe the transnational long industrial chains map between the sectors of China and Belt and Road countries. Research shows that the key downstream hubs of China include the mining sector the United Arab Emirates, Saudi Arabia, Qatar and Russia, the automobile manufacturing sector in Indonesia and the waterway transport sector in Greece. Upstream hubs include the automotive manufacturing sector in Germany, the mining sector in the United Arab Emirates and the electronics, electrical and machinery manufacturing in Malaysia. These hubs improve the integrity of the production network between China and Belt and Road countries by connecting the sectors of countries and forming transnational long industrial chains. Based on the analysis results, this paper proposed policy suggestions for further implementation of the Belt and Road Initiative and the promotion of the "dual circulation" strategy:1) Realizing the driving and optimizing effect of international circulation on domestic circulation; 2) Promoting service-oriented development of manufacturing sectors; 3) Investment layout of Belt and Road countries to balance external risks.

People have been paying more attention to the relationship between urban innovation ability and urban air quality in recent years. As a result, this paper constructs a spatial panel data model to study the impact of urban innovation capacity on urban air quality and analyses urban PM2.5 from the perspectives of population density, economic growth, urban traffic level, industrial structure, energy consumption, and urban sprawl using panel data from 35 Chinese cities from 2001 to 2016. The findings show that:Urban PM2.5 pollution has a significant spatial spillover effect; improving urban innovation ability and optimizing the urban industrial structure significantly reduces PM2.5 pollution emissions; and improving urban population density, energy consumption, and transportation levels all together aggravate urban PM2.5 pollution emissions.

The prediction of the development trend of internet public opinion is a very significant reference for monitoring and control of the network public opinion by the relevant government departments. On account of small sample characteristics of online public opinion and the needs for both accuracy and stability in the prediction model, in this paper, an improved grey wolf optimization algorithm (IGWO) based on the initialization of the good-point set method, nonlinear parameter control and the weighting of the leading wolf is proposed. Using IGWO to optimize the super parameters of SVM regression model, a network public opinion prediction model based on improved grey wolf optimized support vector machine regression (IGWO-SVR) is established. Empirical research is carried out with Baidu indexes such as COVID-19 as public opinion data samples. The experimental results of 12 test functions show that the improved grey wolf optimization algorithm has relatively strong global search ability, faster convergence speed and better stability. The IGWO-SVR model has relatively outstanding accuracy and stability in the prediction of the development trend of public opinion, which can provide better decision-making support for public opinion supervision department of government.

Global sensitivity analysis aims to identify how the uncertainty propagates within a model and to screen out factors or their interactions that have the most significant influence on the output.When the functional relations between inputs and outputs is a'black box'or the dimension of the inputs is high,the computation of Sobol'indexes using Monte Carlo integration is no longer affordable.Further,the potential correlation among the random inputs makes the existing methods based on independence assumption not valid.By modeling the correlated inputs using a Gaussian copula,we propose an approach for computing Sobol'indexes based on sparse PCE (polynomial chaos expansion).Firstly,the dependent inputs are converted into independent,uniformly distributed variables on[0,1]ⁿ by a series of transformations,so that the Sobol'decomposition can be applied using a shifted Legendre basis.For the adaptive modeling,two criteria are recommended to achieve an appropriate PCE:The optimal number of sampling points are determined by the convergence of variance of first-order expansion terms,and the order of polynomials for first and second-order expansion terms are respectively determined based on the model approximation error.Lastly,the PCE-Based Sobol'indexes are easily calculated by the model parameters.The proposed approach is applied to analytical examples and a practical problem in auto manufacturing.The results demonstrate its effectiveness and reveal the relationships between the magnitude of correlation,Sobol'indexes and the complexity of PCE model.

As an important part of the social network, supply chain network have an important impact on the company's production and operation and strategic decision-making. This paper uses the information of the top five suppliers and customers disclosed by China's Shanghai and Shenzhen A-share listed companies from 2004 to 2019 to build a supply chain network, and uses degree centrality to measure the position of different companies in the supply network. The research finds that the listed companies in the center of the supply network have a higher competitive position. This conclusion is still valid after the robustness test methods such as instrumental variable method, PSM method, Heckman two-stage method and replacement variables. Heterogeneity analysis finds that the closer the average spatial distance between companies and customers or suppliers, or when companies have no contract disputes with customers or suppliers in the current year, the stronger the positive correlation between the central position of supply chain network and company competitive position. Mechanism test shows that the location centrality of supply chain network can help companies obtain more information and technology resources, improve company's investment level and investment efficiency, enhance innovation ability and reduce transaction cost, so as to promote the promotion of company' competitive position. The economic consequence test finds that when the company is in the center of the supply chain network, it can also weaken the operation uncertainty and restrain the management shortsightedness. This paper not only enriches the literature in the two fields of supply network location characteristic and company competitive position, but also has positive practical significance for companies to seek growth power and realize high-quality development from the location level of supply chain network.

It is the first time to construct a multi-dimensional measurement framework of foreign exchange tail risk based on two-way extreme value model and coefficient decomposition, and deeply analyze the change law of foreign exchange tail risk exposure. It is found that under the competitiveness effect and conversion risk effect, the tail risk exposure and its decomposition components have obvious foreign exchange up and down heterogeneity. The regression analysis with enterprise characteristics shows that only when facing the downside foreign exchange tail risk, OFDI can play a significant hedging role. When the exchange rate rises, the effect of OFDI's overseas asset conversion risk will offset the effect of OFDI as an operational hedging tool; When the competitive effect of OFDI is insufficient and the hedging effect is low, OFDI may even become a significant source of risk. The above conclusions provide a new perspective and inspiration for foreign exchange risk prevention.

Starting from the motivation of outward foreign direct investment (OFDI), this paper sorts out the main influence paths of OFDI on employment, puts forward a measurement model based on time-varying state space model and input-output model, and calculates the home country employment effect and opportunity cost of China's OFDI to the U.S., using China's manufacturing OFDI and export data to the U.S. from 2007 to 2014 and the world input-output table. The results show that:1) China's OFDI to the U.S. generates an overall creation effect on employment, with an average of 1.47 million jobs offered to China each year; 2) From the perspective of investment motivation, market-seeking OFDI has a creation effect on home country employment, while resource-seeking OFDI has a substitution effect on it; 3) The opportunity cost of China's OFDI to the U.S. for employment was relatively large before 2012, while after 2012, OFDI created more employment in home country compared with domestic investment.

China's external crude oil dependency rate is high, supply and demand shocks in the international oil market have significant effects on China's economic development. This paper innovatively constructs a research framework that incorporates supply shock-oil price volatility-economic impact (SS-PV-EI) in order to measure the impact of supply shocks on China's economy. First, a PPM-KM model with historical event analysis is developed to analyze historical supply shock. Second, the time-varing elasticity and impulse response of supply shocks is estimated based on TVP-SV-VAR. Third, scenario analysis combined with historical event analysis is applied to analyze four scenarios that may casued by the sanctions imposed on Iran. Four, MIDAS-AR model and NARDL model are used to measure the asymmetric effects of supply shock on China's economy. The results show there are 42 historical events, of which 19 were caused by supply shocks. The oil price will increase by 0.25%~0.45% when the world oil supply decreases by 1%. Whereas supply shocks caused by Iranian sanctions will likely resulted in oil prices rising between 2%~100%. Oil price has impacted China's economy asymmetrically, as the impact of oil price rises is greater than decline. China's GDP will decline by 0.04%~4.17%, and CPI will rise by 0.14%~6.92% under various conditions.

We use the machine learning methods such as penalized linear regression and PCA to fully mine the big data of financial fundamentals of Chinese A-share listed firms, and try to construct Chinese parsimonious multi-factor pricing model. We find that in Chinese A-share market, PC factors as pricing factor is better than the characteristics-based factors. In terms of shrinking the pricing factors, the elastic net method performs slightly better than LASSO. In addition, momentum, trading friction, and value and growth factors have more contribution to pricing, which suggests that our capital market is not perfect and the trading habits of market participants should be improved, and reflects the continuous improvement and effectiveness of Chinese stock market. The multi-factor model we construct is superior to main pricing models. We introduce the machine learning methods to asset pricing, shrink the high-dimensional data, select the optimal variables, build a multi-factor asset pricing model for Chinese A-share market and give a clear economic explanation. We provide new ideas for asset pricing in Chinese A-share market from the aspect of big data and machine learning, which clarifies the contribution of each pricing factor in the multi-factor model, helps understand the characteristics of our capital market and provides a method to solve the problem of high-dimensional financial fundamental data for Chinese capital market.

In the era of big data, the frequent occurrence of information security issues has given rise to consumers' requirement of privacy protection. Privacy protection is essential to enterprise competition. Based on the assumption that consumers have both brand preferences and privacy protection needs, this article uses game theory to analyze the effect of the privacy protection strategies on enterprise competition and social welfare. The result shows that compared with only brand competition, competing through both brand and privacy protection can improve social welfare. The impact of privacy protection on enterprises depends on the gap in the degree of privacy protection. When consumers pay more attention on product brands, companies with high level of privacy protection can set higher prices for higher profit as the gap increases, while companies with low privacy protection are affected in contrast. When the gap is large enough, companies with low privacy protection can obtain higher profits. When consumers pay more attention on privacy security, both firms can get more profit under certain conditions.

Modelling the US West Texas Intermediate (WTI) crude oil spot price structure has received an increasing attention in recent years. However, most of the existing literatures focus on classical econometric methods and point-valued counterfactual analysis, which may suffer from informational loss. This paper proposes a novel interval-based counterfactual analysis to explore the impact of crude oil fluctuation events that appeared on WTI crude oil spot prices in 2011. The interval-based models can be used to simultaneously estimate hypothetical values of level and volatility of WTI price that are not affected by the event. These counterfactual estimations are based on the cross-sectional correlations between WTI and other 10 international crude oil spot markets. Our findings reveal that both the level and volatility of WTI price were lowered by the oversupply of crude oil production caused by shale oil boom, and this effect did not disappear until the end of 2014. The results can provide an important reference for the government, industry practitioners and individual investors.

In order to thoroughly investigate the effects and internal mechanism of government innovation support policies, this paper takes the research and development subsidy policies and tax preferential policies proposed in the "National Medium and Long-term Science and Technology Development Plan Outline (2006——2020)" as the research object, adopts the double difference method based on propensity score matching identifies the double effects of government innovation support policies. The empirical results show that the government R&D subsidy policy has both the effects of "pre-incentive" and "post-reward", which significantly improves the innovation ability of corporate, while preferential tax policies cannot promote corporate innovation. Further analysis shows that continuous and stable R&D subsidies, geographical dispersion of business operations, and external innovation environment have significantly strengthened the positive effect of innovation support policies on corporate innovation. This article not only makes up for the existing research on the ignorance of the internal motivation and potential mechanism of government innovation support policies, but also extends the research perspective from "government hands" to "market hands", makes a necessary supplement to the logic chain of policy influence on corporate innovation capability. This paper reveals the internal law of the implementation of government innovation support policies, provides theoretical support and policy suggestions for government support to drive corporate innovation for our country to cope with increasingly fierce international competition and realize the innovation-driven development strategy.

Since December 2019, COVID-19 epidemic is continuing to spread globally. It not only jeopardizing the lives and health of people around the world seriously and putting a severe test on the public medical and health system, but also causes a huge impact on economic and trade activities and has a deep influence on the international community. In order to help researchers and policy makers understand the mechanism of virus transmission and adopt reasonable anti-epidemic policies to inhibit the further spread of the virus, some studies have adopted mathematical prediction models to simulate the spread of the virus and the development of the epidemic. However, the existing research has certain limitations, such as single method selection, excessive reliance on model parameters selection, and virus transmission and policy adjustments caused time variability of data. To solve the above problems, this paper proposes a comprehensive ensemble forecasting framework, which bases on six single prediction models, including time-varying Jackknife model averaging (TVJMA), time-varying parameters (TVP), time-varying parameter SIR (vSIR), logistic regression (LR), polynomial regression (PNR), autoregressive moving average (ARMA). The proposed method is used to predict the cumulative number of confirmed cases in the 6 most severely affected countries in different regions. Empirical results show that for a single prediction method, the TVJMA method outperforms the other five methods; the comprehensive ensemble forecasting method is significantly better than any single method in most cases, especially, the multi-model combined forecasting method based on error correction weights improves the prediction accuracy significantly. For different prediction steps, the comprehensive ensemble forecasting method is robust.

The accurate forecasting of container throughput is an important basis for reasonably planning port construction, making port operation plan and adjusting port development direction. Aiming at the complex nonlinear characteristics of port container throughput, a hybrid model based on selective deep-ensemble for container throughput forecasting (HMSD) is presented in this paper. First, this model decomposes the original container throughput time series into several intrinsic mode functions and a residual by empirical mode decomposition. Considering highly nonlinear characteristics of each intrinsic mode function, the proposed model trains three deep neural networks, namely, long short term memory, gated recurrent unit and convolutional neural network, as base learners to predict intrinsic mode functions. Then, this model establishes selective deep-ensemble forecasting model by improved group method of data handling on intrinsic mode functions and obtains their ensemble forecasting results. Furthermore, this model uses an autoregressive integrated moving average model to predict the linear residual. In order to verify the performance of the proposed model in container throughput forecasting, six ports with significant differences in throughput in China are selected for empirical testing, and the results show that the model has the best forecasting effect on all six ports. Finally, the monthly out-of-sample forecasts of container throughput of six ports from January 2021 to December 2022 by HMSD model is given.

The seasonal fluctuation data has complex characteristics such as long-term trend, seasonal fluctuation and local random oscillation, which makes it difficult to construct a reasonable prediction model. In this paper, first of all, the ranking function f(x_ik) was used to mine the ordinal relationship and evolution rules of related elements in the seasonal fluctuation data, thus realizing the feature extraction of the seasonal fluctuation data and the construction of the driving term. Then, by constructing the fractional order multi-variable grey prediction model FMGM(1,N), the expansion and optimization of its accumulation order from positive integer to all real numbers were realized. Finally, FMGM(1,N) was applied to the fitting and prediction of monthly GDP data with seasonal fluctuation characteristics in China. The modeling results showed that the simulation and prediction accuracy of FMGM(1,N) was superior to current mainstream univariate and multi-variable grey forecasting models, the nonlinear regression model, Arima model and intelligent modeling methods (support vector machine, SVM; long short-term memory, LSTM). The research results provide a new prediction modeling method for studying seasonal fluctuation data and have positive significance for enriching the prediction model method system.

In cross-border e-commerce, free-shipping service and product price jointly influence consumers' purchase decisions. When brand-owners offer free-shipping services, they will bear both logistics and import tariff fees. Otherwise, these fees will be paid by the consumers. In this paper, we study the free-shipping decisions of two brand-owners A and B when they sell through a common cross-border e-commerce platform with the consideration of the two-dimensional competition of product and logistics service. We develop four strategy combinations (N,N), (N,F), (F,N) and (F,F) and focus on the strategies of brand-owners A and B in terms of free-shipping. We find that the relative size of logistics service costs paid by brand owners and consumers and product differentiation have a decisive influence on brand owners' decisions on free-shipping. We find that there exist three equilibrium strategies (N,N), (N,F), and (F,F) when the two brand-owners' products are highly differentiated, depending on the logistics fee directly borne by consumers. In contrast, when product competition is fierce, (N,N) and (F,F) are the equilibrium strategies. The tariff rate may affect the possibility of various equilibriums significantly, and the increase of tariff rate may lead to the convergence of brand owners' free-shipping strategies, that is, (N,N) and (F,F) are more likely to occur. This shows the unique role of tariffs in the decision of corporate in cross-border settings.

This paper considers estimating individual treatment effects in nonlinear difference in difference models with covariates. Under the general framework of difference in difference (DID) models, we impose some new identification conditions, which assume that changes of the potential outcome distributions given covariates in two periods are invariant for the treated group and for the control group, and the conditional rank of the potential outcomes receiving treatment or not are invariant. Under these assumptions, we propose an estimator for the individual treatment effects. Simulation results illustrate effectiveness of the proposed estimator. Finally, the model is used to analyze the effect of implementing the Minimum Wage Act in the United States on counties' unemployment rates.

Considering China interest rates liberalization reform,the treasury market is an important part of the monetary policy transmission mechanism,in which the treasury yield and fluctuations will be affected by central bank communication information shocks.Compared with formal written communication,oral communication takes place more frequently and more timely,therefore,this paper uses China's central bank oral communication practices from 2003 to 2018 to construct quantitative indicators from multiple dimensions including tone and text complexity for each communication,as well as the textual similarity between current and last communication.Based on the EGARCH model,we comprehensively examine the impact of central bank communication on the yield of the treasury market.We find that communication has an impact on the treasury yield since 2008 financial crisis,and it is more significant for long-term treasury,indicating the information channel of central bank communication is improved gradually.Besides,significant shift in tones,stable and intelligible phraseology can effectively reduce market fluctuation,indicating the coordination channel of central bank communication performs well.That is,China's central bank can effectively maintain financial market stability by expectation management.

TONF, that is trade-in for a new oil-fuelled automobile (OFA), has been taking as a common trade-in strategy in the vehicle industry. With the rapid development of new energy vehicles (NEVs), some vehicle manufacturers have also launched TONE, that is trade-in for a new energy vehicle. However, the co-existence of TONF an TONE possibly aggravates competition between new OFAs and NEVs. Therefore, the vehicle manufacturers and retailers need to determine how to select trade-in strategy and make the corresponding optimal prices. To solve this problem, in the case of consumer mileage anxiety to NEVs, we construct a two-echelon supply chain model composed of a vehicle manufacturer and a retailer in which they cooperate in trade-in, and discuss the threshold conditions under which the retailer only provide TONF, TONE and both of them, as well as the corresponding optimal pricing decisions. We also analyze the influence of the manufacturer's pricing decisions, consumer mileage anxiety to NEVs on the retailer's choice of trade-in strategy. The results show that the retailer's trade-in strategy is not only related to the production costs of OFA and NEVs, but also related to the manufacturer's recycling prices. Consumer mileage anxiety to NEVs will also have an important impact on the retailer' choice of trade-in strategy. In particular, when the production costs of OFA and NEVs are moderate, consumer mileage anxiety to NEVs play a dominate role in retailer' choice of trade-in strategy.

In this paper, we studied reservation control strategy of shared parking spaces with uncertain supply and stochastic demand in the framework of stochastic dynamic programming model. A product-based decomposition model and a period-based decomposition model have been proposed to analyze the optimal solution of the stochastic dynamic programming model. The objective function of the product-based decomposition model is supermodular and the objective function of the period-based decomposition model is concave. We proposed three reservation control algorithms to solve the stochastic dynamic programming model. Finally, numerical simulations have verified the effectiveness of the proposed models and algorithms. The results will provide a support for reservation control in the shared parking system.

As a new marketing method,live broadcast with goods is developing rapidly.How to optimize the dynamic price decision to maximize the broadcast platform and the anchor's profits is an essential task.Based on differential game theory and method,considering the reference effect and anchor influence on the pricing strategy,we study three pricing strategies:Basic pricing,static pricing and dynamic pricing strategy considering the reference effect.The results show that:Under static or dynamic pricing strategies,the optimal pricing of the platform is positively correlated with the initial reference price.When the initial reference price is higher (lower),the optimal price strategy is similar to skimming pricing strategy (penetration pricing strategy);With the increase of the anchor's effort level,the stronger the anchor's influence on consumers'purchasing decision,the platform tends to set higher prices;The dynamic pricing strategy can optimize the profit of the live broadcast platform,but compared with the static pricing strategy,the profit of the anchor is lower;The evolution rule of the live broadcast platform optimal steady-state price and profit with different system parameters.The platform should design the optimal price strategy based on the consumer price reference effect and anchor influence level to improve profitability.

In the cross-national supply chain network system, upstream and downstream enterprises inevitably face tariff policies in the transaction process. At the same time, the substitution of local products will also have an impact on the tariff implementation and the decision-making of all parties. Considering product substitutability, this paper constructs a profit maximization model of multinational supply chain members affected by tariff policy under random demand, and using the complementary system, the equilibrium decision of transnational supply chain network is obtained. The results show that: 1) As the country increases the tariffs on imported products, the profits of the enterprises producing and trading substitute products will increase; 2) When the demand fluctuation of a country's imported products increases, the trading volume of the country's local products will reduce, and both manufacturers' profit and dealers' expected profit will decrease; 3) If the manufacturer improves the substitution degree of its products for imported products, the trading volume of products will increase, which will also increase the profit of the manufacturer and the expected profit of dealers dealing with it, the implementation of national tariff policy will amplify the positive impact of local products on the improvement of the substitution degree of imported products.

Identifying and effectively supervising the important risks of the insurance industry are of great significance for maintaining the stability of the insurance industry and the entire financial industry. Most studies estimate the level of major risks based on quantitative indicators, but there are some indirectness and hysteresis. Regulators usually require insurance companies to disclose current or future potential risks in their textual financial reports. Comprehensively extracting such risk information can gather the experiences of all insurance companies' managers and identify important risks of insurance industry more directly and proactively. Therefore, the text mining method is adopted to identify the important risks of insurance industry from massive financial reports and analyze their evolutions. Based on 1682 textual risk disclosures of 214 listed insurance companies in the United States in 2006-2018, the empirical study identifies 29 important risk points. By analyzing the evolution trend of each risk point, the importance of risk points related to operational risks has shown a significant upward trend, especially the largest increase in "information system security". It is recommended that insurance companies and regulators should now attach great importance to the operational risk brought by new technologies and new business models.