There are many studies on the influencing factors of foreign direct investment in China, and the research focus is not the same. Some scholars have studied the impact of single or similar factors on China's absorption of foreign direct investment. Such as Liu Min (2013) used the quarterly data from 1999 to 2011 combined with the cointegration method to carry out empirical research, from the level, fluctuations and expectations of the three aspects of the exchange rate on China's foreign direct investment and found that the value of RMB and appreciation Are expected to be conducive to the absorption of foreign direct investment, while the impact of fluctuations in foreign direct investment is not significant.

In the research object, some scholars choose provincial panel data. Zhang Chaolei, Jin Qiuming (2013) used 8 years of data from 10 provinces and cities in Shandong and Sichuan to explore the human resource stock, economic development prospect and labor cost, marketization degree and market scale by using the effective variable intercept model. The Influence of Factors on China 's Foreign Direct Investment.

Some scholars have studied the influencing factors of foreign direct investment in a given industry. Zhong Xiaojun (2015) first analyzes the influencing factors and the mechanism of foreign direct investment in service industry, and then establishes the econometric model to test the size and direction of influencing factors of foreign direct investment in service industry. The results show that population density, consumer demand and manufacturing FDI have the greatest impact, ranking the top three, and thus the service multinational companies are mainly demand-oriented.

Jiang Dajian and Yue Gongzheng (2013) divided China's major foreign direct investment into Hong Kong, Macao and Taiwan foreign direct investment and foreign direct investment in Europe, America and Japan according to different sources of FDI, and discussed their influencing factors respectively. The results show that: per capita GDP, tertiary industry, infrastructure construction, foreign investment policy both have an impact on the former determinants of wages, the latter also by the quality of labor, trade policy factors.

Based on China's specific national conditions and the characteristics of absorbing foreign capital, Liu Ju-lin (2006) chooses the per capita GDP and accumulates foreign direct investment from three aspects: the determinants of GDP, the cumulative effect of FDI and the policy and environment of investment. The amount of investment in capital construction, the expenditure on scientific research in the state financial expenditure, the total retail sales of social consumer goods, the expenditure of social education and education in the state financial expenditure and the policy variables to explore the problem.

To sum up, in the study of FDI factors, many scholars only selected one part or from one aspect to start the study. And for such a huge and complex economy in China, if only from a certain aspect of the analysis, it is inevitable not comprehensive. Therefore, this paper hopes to integrate the existing research, select more economic variables index, and then use the econometric method to test each variable, and finally come to a more comprehensive model of the impact of foreign direct investment in China model.

2 model establishment and empirical analysis

2.1 Variable selection and data sources

Based on the economic variables data of 1983 ~ 2013, this paper chooses 11 indexes including explanatory variables, and uses the method of multiple regression to measure the influencing factors of foreign direct investment. Data are from the National Bureau of Statistics and the network, select the variables are as follows.

Explained variable:

FDI: foreign direct investment. Measured with the actual use of foreign direct investment (in thousands of dollars). Explanatory variables:

GDP: gross domestic product (RMB billion), reflecting the size of the economy, is the market capacity of alternative variables.

SALES: Total retail sales of social consumer goods (100 million yuan) to measure market demand.

CPI: consumer price index (billion), to measure the degree of economic stability.

PORT: total import and export (million), reflecting the international economic ties, is an important measure of economic openness indicators.

TIP: The proportion of the added value of the tertiary industry to the proportion of gross domestic product (%), mainly reflecting the financial, information, transportation and other industries development level.

DEBT: The financial institution's financial resources to the international financial institutions (billion), measure the solvency of a country's economy, reflecting a country's investment environment.

TAF: total amount of fixed assets (100 million yuan). To a certain extent, reflects the return on investment capital.

UR: Urbanization rate, measure the level of a country's infrastructure. Good infrastructure can reduce the cost of foreign direct investment.

WAGE: the average wage of workers in the post, measuring labor costs. Low labor costs are good for companies to profit.

ER: RMB exchange rate against the US dollar (1 US dollars = 100 yuan), used to measure the purchasing power of foreign currency on foreign currency. Foreign direct investment is suppressed to a certain extent when the appreciation of a country's currency and the relative depreciation of foreign currencies.

2.2 Model setting and inspection

In order to eliminate the heteroscedasticity, other variables other than TIP and UR are logarithmic, and then the model is initialized. From the initial regression results can be found, the coefficient is high R2, F value is also significantly significant. But the coefficient t test of multiple explanatory variables is not significant, indicating that the model may have severe multiple collinearity.

In this paper, stepwise regression is used to eliminate multiple collinearity. By observing the results of the preliminary regression, it can be seen that the LN of the LNCPI variable is the largest, 0.970748, the most significant effect on the total amount of FDI, and the LNCPI is retained. On this basis, adding other variables, respectively, to observe the initial regression results that LNTAF variable R2 = 0.974663, the most significant impact on the total amount of FDI, so the LNTAF variables remain. And so on, until the addition of any of the remaining variables, the regression effect will be reduced, then the model no longer add other variables, step by step regression process. The results of the stepwise regression are as follows:

Since the selected economic variables are time series data, most economic time series are nonstationary. If the non-stationary time series is directly used as a stationary time series regression analysis, pseudo-regression phenomenon may occur, resulting in reduced model reliability. It is also necessary to carry out the smoothness and cointegration test of the economic variables in the model. First of all, with the ADF test to determine the stability of the variables. The results show that the other variables are first-order monotonic sequences except that TIP is a stationary sequence. Then we use the EG two-step method to test the cointegration relationship between variables, and conclude that the t value is less than the critical value when the significance level is 0.05, so the covariance is between the variables.

Finally, the autocorrelation test and correction. According to the DW value of the previous regression results, we can see that the model has autocorrelation problem. Therefore, the first order difference is obtained for the five variables other than TIP, where the ρ value can be obtained by the negative first order regression of e and e. On the basis of the above generalized differential regression method, we can see that the DW value is 1.944 and the autocorrelation has been eliminated. At this point the final model is available as follows:

3 Economic interpretation of the model

According to the final model established in this paper, the gross domestic product (GDP), the consumer price index (CPI), the added value of the tertiary industry as the proportion of gross domestic product (TIP), total fixed assets (TAF) The average wage (WAGE) of the workers explained the changes in foreign direct investment (FDI) of 98.99%.

In the case of other factors unchanged, GDP, CPI, TIP, TAF four economic variables have a role in promoting FDI, TAF which the strongest impact on China's market size, economic environment, the tertiary industry development, The return on capital investment has a positive impact on the attractiveness of foreign direct investment in China. In the case of other factors, LNFAGE will decline by 1.54% for every 1% increase in LNWAGE, indicating that with the labor force of our country The gradual increase in the cost of nearly 30 years to a certain extent, inhibit the growth of FDI.

In general, foreign investors in the investment, the main value is the level of investment and economic development of the level of stability, the tertiary industry development level, the real return on investment and labor costs and other factors. Some of the variables presented in this paper, but not shown in the model, have a strong alternative to the existing variables in the model, such as the total retail sales of social consumer goods and the consumer price index, per capita gross national product And gross domestic product, etc .; there are some variables, because the data does not represent the economic factors that it actually reflects, so in the data analysis results, did not show a significant impact.

Policy and advice

Based on the above analysis, we know that the factors that affect the level of FDI in China are the market size, the degree of economic stability, the level of financial information industry development, capital return rate and labor costs. The scale of the market, the degree of economic stability, the level of development of the financial information industry and the rate of return on capital have a positive impact on foreign direct investment, and the labor cost has a negative impact. Combined with the model results, this paper gives the following policy recommendations.

4.1 to maintain economic stability

This paper chooses the CPI as an indicator of the domestic price level. To maintain price stability, the need to reduce the market supply and demand fluctuations. This requires the relevant government departments in accordance with different markets to improve the corresponding laws and regulations, especially in the transparency of commodity information to be improved to reduce the information asymmetry brought about by the vicious transactions, thereby enhancing the degree of economic stability.

4.2 to expand the market size

This paper chooses GDP as the measure of China's economic scale, which has a positive impact on FDI. China should use the current stable external environment, and actively develop their own economies, stimulate domestic demand, in order to strengthen the market scale to attract foreign investment advantage. Specific from the expansion of GDP and increase the actual disposable income of residents to stimulate domestic consumption in two areas. The current domestic social security system is not perfect, while the traditional consumption habits, residents will be more money for the savings rather than consumption. So the government can improve the social security system to improve the current consumption growth of the situation.

4.3 regulate the development of finance and information industry

The current era of rapid development of information, Internet banking gradually occupy the market. In the innovation at the same time, also need effective institutional constraints. The essence of finance is the use of trust, but the financial risk is also largely from information asymmetry and trust collapse. Therefore, the government should encourage the Internet financial innovation at the same time, do a good job of corporate trust level of the investigation, take appropriate measures to reduce the risk, thereby enhancing the financial and information industry development level.

4.4 Increase the rate of return on capital

Although many of our enterprises are developing rapidly, but in achieving high return on capital

The area is still at a disadvantage. Therefore, business management should focus on improving

Business level and asset profitability, more effective use of capital to achieve greater business

Of the proceeds.

4.5 Adjust labor costs

Relatively low labor force is very helpful to attract foreign investment in China, but because of the relatively low labor cost base in China, it will reduce the living standard and happiness of our citizens. Therefore, this paper argues that labor costs should be adjusted according to the level of regional economic development, set a minimum level, and can not blindly reduce.