Does the Cumulative Effect of R&D Investment Exist in High-Tech Enterprises? —Empirical Evidence from China A-Share Listed Companies

In the international society, a new-round revolution is being launched with the continually rise of new technologies. Technical innovation in global perspective has become corporate and national strategy. This article takes the high-tech enterprises from China A-share listed companies as research sample, empirical results demonstrate: The input of R&D investment has a lag period of 2 years and cumulative effect on output; the regression coefficient of initial investment of R&D funding to the current performance is 1.17 that is the type of increasing returns. This article suggests that high-tech enterprises should constantly improve corporate governance mechanism, accelerate transformation of R&D achievements, shorten R&D process to improve R&D efficiency, and establish a value evaluation system for independent innovation projects. The research enriches literature of independent industry, different with the existing research models, with cumulative effect using Cobb-Douglas production function model based on the random effect of panel data.


Introduction
Scientific and technological innovation of high-tech enterprises is mainly reflected in investment of knowledge innovation, technology innovation, management innovation and Research & Development (referred to as R&D) activities. R&D expenditure and human capital investment occupy more and more shares in enterprises' capital expenditure. In recent years, the increase propor-More and more countries recognize the importance of science and technology to the national economy in the context of global economic integration. Science and technology promote the growth of national economy and maintain the international competitive position effectively. China has carried out the construction of innovation strategy. Innovation activities are the basis of scientific and technological progress and play an increasingly important role in economic and social construction. However, current research of R&D activities in China shows few literatures about the empirical relationship between R&D investment and enterprise performance based on high-tech enterprises. Existing two main limitations: First, the disclosure of R&D information is inconsistent in various companies, which brings a large difference in the statistical data caliber, making further research difficult. Second, the incompleteness of R&D investment information disclosure makes it difficult to obtain a large amount of basic R&D activity data. This article hopes that the research on the relationship between R&D investment and corporate performance can attract the attention of high-tech enterprises, formulate long-term and effective future development strategies, and provide certain theoretical support for corporate strategic decisions.

Literature Review
The correlation between R&D investment and corporate performance is concentrated in three aspects. Initially based on the endogenous growth theory, the R&D investment and enterprise productivity research was formed, using the improved Cobb Douglas production function model and beyond logarithmic production function model in economic theory to discuss the contribution of R&D investment to productivity. The corporate performance of high-tech enterprises introduces profitability into the research framework to discuss the correlation between R&D investment and corporate performance. With the deepening of research, more scholars have found that the intensity of corpo-R. X. Qi DOI: 10.4236/ojbm.2020.83072 1124 Open Journal of Business and Management rate R&D investment and the company's future performance are more significant.

Correlation between R&D Investment and Corporate Performance
R&D investment has a positive impact on the company's profitability (Branch, 1974). Positive correlation between R&D investment intensity and operating income market share (McLean & Round, 1978). Different R&D investments have different contribution rates, companies with large R&D investments are superior to small companies (Amir & Sougiannis, 1999), but moderate R&D investment contributes the most to performance of US companies and gives enterprises a comparative advantage (Bosworth & Rogers, 2001). A sample of 244 enterprises from developed countries and regions in the world were selected to study the high-tech enterprises that R&D investment expenditure exceeded 100 million US dollars, found a positive correlation between R&D investment intensity and annual sales revenue growth rate and positive impact is more significant (Roberts, 2001 (Garner, Nam, & Ottoo, 2002). Empirical research on US chemical and pharmaceutical industry from 1975-1996 shows that investment in R&D funds can bring profits to enterprises more than investment in fixed assets (Hsieh, Mishra, & Gobeli, 2003). The 2005 R&D Scoreboard by the UK Department of Trade and Industry showed that: enterprises that use R&D as main influencing factor for competition, such as in medicine and information technology fields, the strength of R&D investment directly affects the rate of return on business profits is positive. In French and German enterprises, R&D expenditure and corporate value show a significant positive correlation, similar to the situation in the US, but for Italian enterprises is not obvious, the control of major shareholders have a certain degree of influence on the correlation between R&D investment and corporate performance (Hall & Oriani, 2006). By analyzing 770 listed companies in the UK, concluded that a positive correlation between R&D investment intensity and the enterprises total revenue (Seraina & Levis, 2008). R&D investment is the most important factor affecting patents and has a positive impact on sales revenue growth (Sohn, Hur, & Kim, 2010). Under the international economic situation, the growth of enterprise R&D investment has a positive correlation with enterprise performance (Ren, Eisingerich, & Tsai, 2015).
Scholars in China have formed two academic views. One view hold that a linear correlation between R&D investment and corporate performance in high-tech enterprises, and the research results are mainly based on a positive correlation (Chen, Zhang, & Chang, 2006;Liang & Yan, 2006;Zhang & Li, 2009;Feng, 2010). In terms of negative correlation: R&D investment and innovation per-R. X. Qi Open Journal of Business and Management formance have a weak negative correlation (Zhang, Su, & Li, 2011). R&D investment significantly positively correlated with corporate profitability in China's high-tech listed companies, but not significantly positively correlated with development capabilities and cannot efficiently promote the growth of high-tech enterprises (Du, Yan, & Chen, 2014). The fluctuation and intensity of R&D investment affect the corporate performance together, the greater R&D investment fluctuation jump, the higher corporate performance. In addition, the dual innovation focus of decision makers plays an intermediary role (Wu & Xiao, 2015).
While another view holds that a non-linear correlation between R&D investment and corporate performance, a few scholars mainly rely on the "inverted U" empirical results (Zhou & Xu, 2007;Cheng & Dai, 2012).

The Lag of R&D Investment to Corporate Performance
In terms of the lag of corporate R&D investment on performance, length of the lag period varies from industry to industry. Due to the particularity, the pharmaceutical industry has a lagging effect of R&D investment on corporate performance for 5 years, the electronic information industry is lagging by 2 years and other ordinary industries are lagging by 3 years (Goto & Suzuki, 1989). The empirical results show that R&D expenditure increment is 1, the company's surplus increment in the next 7 years is 2, the market value increment is 5 and the company's R&D investment significantly lags behind the improvement of performance (Sougiannis, 1994). Studied the impact of R&D expenditure current year on the operating income the next year by enterprises with high R&D density in different industries, the conclusion drawn that the lag period is different with industries (Lev & Sougiannis, 1996). Focused on the relationship between R&D and future growth, empirical results show a significant positive relationship (Deng, Lev, & Narin, 1999). Follow-up research indicates that R&D investment increases the value of future growth (David & Stephen, 2000). In high-tech listed companies not exist correlation between R&D investment funds and ROA, R&D investment has a two-year lag period for corporate performance (Liang & Zhang, 2005). An empirical study of China's electronic information industry from the micro level shows that the reason for the insignificant positive impact of R&D investment on corporate performance is the lagging effect (Wang & Wang, 2008). R&D investment has a non-positive impact on corporate performance, but R&D human resource input is significantly positively correlated with corporate performance in SME board listed companies and not exist lagging effect (Wang & You, 2009). Using inter-period valuation techniques, investors can readjust the low valuation in the future year if they do not fully confirm the value of R&D in the current year. Based on this, R&D investment is not correlated with the change of stock price in the same period, but is significantly positively in the next year (Luo, Zhu, & Li, 2009). Empirical analysis of the impact of R&D investment on the enterprise operating performance in information technology and manufacturing industry, research results show that it is negative correlation R. X. Qi Open Journal of Business and Management (Lu & Wang, 2011). Research on listed companies in pharmaceutical manufacturing sector found that the R&D efficiency and R&D investment intensity is low. R&D expenditure is negatively correlated with corporate performance, the lagging effect in 2011 is significant (Wan & Chen, 2014).

Empirical Analysis
Hysteresis is that R&D investment does not generate revenue in the year when it occurs, but in several fiscal years after the investment. The theoretical root of the lag in R&D investment from the birth of new technology to the final harvest, it needs to go through many links such as research and development, testing and inspection, application and promotion. This process is not overnight and it takes a period of time for the expected effect to appear. The number of lag periods varies depending on the industry development cycle and technical proficiency.
Cumulative effects originated from environmental impact assessment, which emphasizes that the environmental impact of a single impact activity in a short time may not be obvious and ignored, but if it is repeated multiple times, then the impact will be significant in the future. The cumulative effect arises from the cross-effect and interaction between the multiplier effect and the acceleration effect. The cumulative effect of R&D investment on corporate performance means that the enterprises R&D activities in the short term will not have a significant impact on the corporate performance, but if continuous R&D investment and the resulting promotional force will significantly improve.

Model
Considering the characteristics of R&D activities of high-tech enterprises that need to go through project approval, research, development, application and other links, it will take a long time to bring economic benefits to the enterprises and should be continuously accumulated, only when the R&D products be mass-produced can the company's performance be improved.
Hypothesis 1: The impact of high-tech enterprise R&D input on output has a lag period and establishes a multiple linear regression model.  Table 1.
Investment_int i-j represents R&D investment intensity in year i − j, ln_size i represents natural logarithm of total assets in year i, lev i represents debt asset ratio in year i, om i is a dummy variable that consider Chairman-CEO Duality in R. X. Qi Open Journal of Business and Management year i, ε is random disturbance term, i represents the year (i = 3, 4), j represents the lag period (j = 1, 2, 3).
Hypothesis 2: The impact of R&D investment of high-tech enterprises on output has a cumulative effect. The corporate performance in year i is positively correlated with the previous cumulative R&D investment in year i − j.
To establish the linear regression equation, take the natural logarithm on the formula: In the model, Profit i represents the explained variable of output that uses main operation profit indicator, i represents the year (i = 3, 4) and j represents the lag period (j = 1, 2, 3), ε is random disturbance term, A is a constant term. Investment i − j represents the enterprise R&D investment in the i − j year, Labor i represents the employees' number in high-tech enterprises, Asset i represents the total assets in year i. a, b, and c represent the output elasticity coefficients of R&D investment, employees' number and total assets.

Data and Variable Selection
This paper selects three main industries: information transmission, software and The variables that this paper involves are explained below, showing in Table   2.
1) Core variables R&D investment intensity: High-tech enterprises are a typical technology capital-intensive industry and R&D activities are concentrated on capital investment. R&D investment intensity is an indicator internationally used that can objectively and credibly reflect the investment intensity of R&D funds.
Employees' number: High-tech enterprises are mainly small and medium-sized enterprises that with small scale but good growth. The number of employees used as a regulating variable to effectively reduce the asymmetry of corporate performance.
Assets: External financing ability constrained by the scale of an enterprise.
Due to the economies of scale, large enterprises' bargaining power has a comparative advantage. At present, China's high-tech enterprises are with short development process and no strong economies of scale. R&D investment and human resources are limited to a certain extent, so this article examines the impact of firm size on empirical research.

2) Control variables
Debt asset ratio: Due to the high level of corporate motivation for using liabilities, financial pressure is greater and R&D capital investment will be reduced to ensure the normal operation. However, debt capital can reasonably avoid taxes, which is conducive to the rapid growth and enterprise value growth.
Chairman-CEO duality: With the separation of ownership and management, the managers who get equity incentive will attach great importance to the investment of R&D activities and share ownership as a contract makes the interests of shareholders and management consistent. At the same time, the existing literature has proved that the proportion of management shareholding has a positive correlation with R&D investment. Therefore, corporate governance becomes an indispensable factor.

Factor Analysis
The explained variable's factor analysis process as follow:

1) KMO and Bartlett's test
In Table 3, the KMO value is 0.693 > 0.5, which is suitable for using the factor analysis method. When KMO is close to 1, the analysis effect is better. The significance level of Bartlett's sphericity test is 0.000 and the P value is extremely significant, which is suitable for factor analysis.   Table 5. When the factor loading matrix rotated, the maximum variance method in SPSS is used to highlight the practical significance of extraction of the four common factors, and the score of them showing below and total score of the corporate performance can be obtained. Only cited common factor 1 as an example:

2) Extraction of common factor
For the comprehensive evaluation of corporate performance of high-tech listed companies, 4 common factor scores were substituted into the model with the contribution rate of common factor variances as the weight, the new explained variable is Performance i :

Empirical Result
The regression results showing in Table 6 show that R&D investment intensity of high-tech enterprises has a significant positive correlation with corporate performance whether the lag period is 1 or 2, the coefficients are 2.153 and 4.102 which demonstrates that the role of R&D investment has been gradually brought out with an obvious increasing trend in three years and passed the significance test of 10% and 1% respectively.
From 2010 to 2013, accumulated R&D investment in previous period i (i = 1, 2, 3, 4) to the current corporate performance showing in Table 7 that the sum of regression elasticity coefficient a, b, and c is 1.17 (>1) which determined as the type of increasing return. High-tech enterprises applying new technologies to expand production scale can effectively increase the scale of output.

Robustness Test
This article used Bootstrap 300 times test showing in Table 8. The R&D investment intensity lags1 period passed the 10% significance test and lags2 period passed the 1% significance test, the significance level is consistent with the original. The significance level of the coefficients of variables size, lev and om also same as the original, which further proves the accuracy of model setting.

Conclusion and Suggestions
Through the above research model and test of original hypothesis, the following conclusions are obtained: The input of R&D investment has a lag period of 2 years and cumulative effect on output, when lagged by 2 periods, the regression coefficient is 4.1, which is significant at the 1% level.

3) Establish a value evaluation system for independent innovation projects
High-tech enterprises have the characteristics of large capital demand, long R&D investment cycle, high capital utilization rate and high project risk. Therefore, the value assessment and feasibility analysis of the initial investment of R&D funds are particularly important for high-tech enterprises. The risk factors in the environment control the project risks and losses within a certain range and limit. Establishing a whole set of feasible and perfect value evaluation system will be the key to whether R&D investment can effectively bring enterprise performance and this is also the key to the high-tech enterprises' sustainable growth.

The Limitation and Outlook
For the undisclosed data of some years, the research period is limited, the follow-up research in this paper will try to make up by in-depth collection of first-hand data.
Meanwhile, the research object is high-tech enterprises, mainly selected 3 typical can expand the research conclusion from the micro level to the meso level, and can make subsequent attempts. Besides, based on the limited available data, only A-share listed companies are included in considered, whether the research conclusions equally applicable to non-listed companies, it remains to be further proved that the entire capital market should be included in the scope of the research as far as possible.

Conflicts of Interest
The author declares no conflicts of interest regarding the publication of this paper.