Empirical Research of Business Development Potentiality of Cross Border E commerce between China and the Belt & Road Countries

Since the financial crisis, the global economy has fallen into a trough and recovered slowly. Under the double impact of the “new normal” of the internal economy and shrinking external demand, the growth rate of China’s traditional foreign trade has shrunk dramatically. In contrast, cross-border e-commerce has become a key support point for China’s foreign trade. In addition, the “Belt and Road” strategy is proposed to provide good development conditions for cross-border e-commerce. This paper takes the “Belt and Road” cross-border e-commerce as the background, builds a systematic cross-border e-commerce impact mechanism evaluation system, and analyzes the potential of cross-border e-commerce export from China to the countries along the “Belt and Road” based on the stochastic frontier gravity model. The conclusion is that improving infrastructure and logistics construction, improving the development of e-commerce environment, strengthening financial services, and depreciating RMB can promote cross-border e-commerce export efficiency, and thereby tap the export potential of cross-border e-commerce. Finally, this essay puts forward specific policy recommendations to help the development of cross-border e-commerce, combining the actual conditions of China and the countries along the “Belt and Road”.


Introduction of Stochastic Frontier Gravitation Model
In order to solve the problem of production efficiency, Meeusen and Broeck (1997) and Aigner et al. (1997) first propose the stochastic frontier method, which defined production efficiency as the ratio of actual output to theoretical maximum output. The formula for applying the stochastic frontier method to the gravity model is as follows: Where T ijt represents the actual trade volume between country i and country j in period t; x ijt is the core variable affecting the trade volume in the gravity model, such as economic size, population, distance, etc.; β is the parameter to be estimated. Formula (2) is the logarithmic form of Formula (1).
In stochastic frontier gravitation model, v ijt represents measurement and specification error. u ijt stands for trade inefficiency, including factors that promote or restrict trade. In this paper, the inefficiency effects are modeled in terms of other variables, as suggested by Battese and Coelli (1995) and expressed as Formula (3), where z ijt is a vector of explanatory variables associated with the technical inefficiency effects, α' is a vector of unknown parameters to be estimated, and ε ijt represents random disturbance term. u ijt and v ijt are independent of each other, and u ijt follows a truncated normal distribution.
In Formula (4) and (5), * is the trade potential, which stands for the maximum trade volume that country i can trade with country j in period t; TE ijt is trade efficiency, which is the ratio of actual trade volume to trade potential.
The trade development potential can be judged by trade efficiency: when u ijt = 0, TE ijt = 1, and there is no trade inefficiency between two sample countries, the trade volume reaches the maximum, so the actual trade volume is equal to the trade potential; when u ijt > 0, TE ijt ∈(0,1), and there is trade inefficiency between two sample countries, so the actual trade volume is less than the trade potential.

Specific form of the Stochastic Frontier Gravity Model
The stochastic frontier gravity model and its variables are specifically shown in formula (6) and Table   2: = 0 + 1 + 2 + 3 + 4 + 5 + 6 + 7 + 8 + − Where T ijt represents the volume of China's cross-border e-commerce exports to country j in year t.
However, cross-border e-commerce data is difficult to obtain directly, so this paper refers to the method of Chinese scholars and adopts the cross-border e-commerce data processing method released by iResearch, as shown in formula (7):

The Specific Form of the Inefficiency Model
This paper applies the method of constructing the trade facilitation indicator system (Wilson et al., 2003) currently used by most scholars in the field of cross-border e-commerce, based on the authority, operability and quantification of data. Also, this paper combines the background of "Internet +" and the logistics, customs environment, regulatory environment, e-commerce, financial services, and further refines these five first-level indicators divided into 17 secondary evaluation indicators. Since the value range of 17 secondary indicators is different, in order to make the data more intuitively comparable, this paper unitizes the data of these 17 secondary indicators, and then assigns the same weight to each unitized secondary indicator to construct the primary indicators. Quality of railroad infrastructure X 2 1-7 Quality of port infrastructure X 3 1-7 Quality of air transport infrastructure X 4 1-7 Regulatory Environment (REG) Public trust in politicians X 5 1-7 Judicial independence X 6 1-7 Burden of government regulation X 7 1-7 Efficiency of legal framework in settling disputes X 8 1-7 Transparency of government policymaking X 9 1-7 Burden of customs procedures X 17 1-7 Note. In the range of values, 1 (0) = worst and 7 (100) = best.
The inefficiency model and its variables are specifically shown in formula (8) and

Model Checking
In this paper, the maximum likelihood estimates tool FRONTIER4.1 is used to build the stochastic frontier model. It is necessary to use the likelihood ratio (LR) test to validate the model because the stochastic frontier model has very high requirements on the function form.
According to Table 5, the LR statistic of "there is no cross-border e-commerce export barrier" is much greater than the critical value of 1%. The null hypothesis is rejected at a 1% significance level, which means that it is necessary to use the stochastic frontier gravity model. Table 6 shows that CTG, CUS, TAR, and FTA fail to pass the critical value of 5%, indicating that these are not important variables affecting cross-border e-commerce exports. Given that the stochastic frontier analysis is highly dependent on the functional form of the model, these variables should be eliminated.

Analysis of Influence Factors
In this paper, one-step estimation is used in the stochastic frontier gravity model, and the empirical results are shown in Table 7. Model a contains all explanatory variables; Model b to e are the regression models with the insignificant variables CTG, CUS, TAR, and FTA deleted respectively; Model f is the regression model after removing all the insignificant variables, that is, the final function of the stochastic frontier gravity model.
According to model f, the GDP of China and the import country, population size (POP), per capita income gap (DPGDP), and common language (CLG) have positive effects on cross-border e-commerce exports, while geographic distance(DIS) has negative effects on cross-border e-commerce exports Impact. Infrastructure and logistics (FRA), e-commerce (ICT), financial services (FIS), foreign currency exchange rate against the RMB (EXC), SCO member, and WTO member are significantly negatively related to export inefficiency term (u). Therefore, the empirical results are in line with expectations.

Analysis of the Export Potential of Cross-border E-commerce
For cross-border e-commerce exports, the higher the value, the higher the cross-border e-commerce export efficiency, while the lower the value, the greater the cross-border e-commerce export potential to explore. As can be seen from Figure

Figure 2. The Efficiency Changes of China's Cross-border E-commerce Exports to Countries along the Belt and Road
It can be seen from  This paper further divides the countries along the "Belt and Road" into four quadrants by taking the cross-border e-commerce export efficiency as the abscissa axis and the cross-border e-commerce export growth rate as the ordinate axis from 2009 to 2017. The four quadrants are export development zone, export core zone, export remodeling zone and export key zone, which are shown in Figure 3 and Table 9.
The export core zone is characterized by "high export efficiency and high export growth rate". The export development zone is characterized by "low export efficiency and high growth rate".

Qatar (QAT), India (IND), Pakistan (PAK), Bangladesh (BGD), Nepal (NPL), Georgia (GEO), Bosnia
and Herzegovina (BIH), Montenegro (YUG), and Albania (ALB). The high growth rate and low export efficiency indicate that China's cross-border e-commerce exports to these countries are growing rapidly, but the export obstacles are relatively large, mainly due to the backwardness of infrastructure and logistics and e-commerce level (FRA average coefficient is 0.521, ICT average coefficient is 0.599).
China should strengthen cooperation with countries in the zone in infrastructure and logistics, digital finance, and e-commerce.
The export remodeling zone is characterized by "low export efficiency and high growth rate".

Conclusion
This The results show that: (1) From the perspective of macroeconomic and natural factors, the larger the GDP of China and import country, the population and the GDP per capita gap between China and import country, the more China's cross-border e-commerce exports to the country along the Belt and Road. The use of a common language has a positive effect on China's cross-border e-commerce exports to the country along the Belt and Road, while geographical distance has a negative effect on exports. Additionally, whether China and the import country are adjacent does not have a significant impact.
(2) From the unique factors of cross-border e-commerce, the improvement of the quality of infrastructure and logistics, e-commerce environment and financial services can increase China's cross-border e-commerce exports to the country along the Belt and Road. Moreover, the depreciation of the RMB and the importing country being the SCO or WTO members are beneficial to exports.
(3) From 2009 to 2017, China's average export efficiency to 52 countries along the Belt and Road was 0.603, of which 30 countries were above the average level and 22 countries were below the average level. China's export efficiency to Central Asian countries is much higher than other regions. enabled China and relevant countries to better improve logistics, customs, finance, infrastructure and other related systems that are conducive to the development of cross-border e-commerce. The joint development of cooperation laws and regulations has made the operation of cross-border e-commerce business more standardized. Actively participating in free trade zone negotiations can also provide a supportive policy environment for the development of cross-border e-commerce.

For Enterprises
First, strengthen the establishment of overseas warehouses. The construction of the Belt and Road overseas warehouse can enable consumers to receive goods quickly after placing an order, effectively improve transportation efficiency, reduce transportation links, and also facilitate product after-sales processing, thereby effectively reducing logistics and operating costs. Also, the short-term logistics lag and the long-term difficulty in global stocking can be effectively alleviated through overseas warehouses during the COVID-19 epidemic. Therefore, Chinese cross-border e-commerce companies should actively cooperate with local companies and related logistics companies to jointly promote the establishment of overseas warehouses in countries along the Belt and Road.
Second, strengthen information construction and improve service levels. In order to increase the number of orders, cross-border e-commerce enterprises and platforms should further improve the rapid and effective supply and demand matching and docking technology to accurately locate customer groups, and at the same time, enterprises should strengthen product quality follow-up. Especially during the COVID-19 epidemic, for urgently needed products such as Masks and hand sanitizers, a quick and convenient purchase channel should be established to meet consumer demand. Meanwhile, enterprises and platforms should strengthen cooperation with financial institutions on the basis of high-quality information matching and high-quality service level to improve the efficiency of payment and settlement and the efficiency of cross-border e-commerce transactions, thereby better promoting cross-border e-commerce development.