R&D expense stickiness in Chinese listed firms

Abstract

In 2021, the Chinese government announced China’s objective to increase spending on research and development (hereafter R&D) by more than 7% per annum between 2021 and 2025, making China the world’s second largest spender on R&D. Out of China’s total R&D spending of RMB 2.39 trillion in 2022companies accounted for 77.6 percent (globaltimes, 2023). While companies make significant contributions to R&D they are faced with declining sales revenue in China and increasing competition internationally, thus, remaining profitable is a challenge for businesses. On the one hand R&D is a source of competitiveness and long-term profitability, on the other hand the combination of sales decline and more intense competition makes one question the prudence of increasing R&D investment. Therefore, the empirical, and timely question remains: How do firms adjust their R&D spending in response to changes in sales or to changes of other internal and external firm factors.
Drawing on research into cost stickiness, this thesis addresses three research questions, on how firms’ spending on R&D changes in response in fluctuations in sales, and how this relationship is affected by: (1) management’s motivation for earnings management and management’s ability; (2) concentrated ownership in conjunction with management ability, and (3) intensity of product market competition in conjunction with management ability.
To address these questions, data is drawn from firms listed at the Shanghai and Shenzhen Stock Exchanges. A panel data is constructed and we explored the change of R&D expenses in response to the changes in sales revenue. We allow the changes to be different in sales increases and decreases using interaction variables – we interacted the growth of sales revenue with a dummy variable that indicates whether the sales had increased (growth positive) or decreased (growth negative). This modelling structure follows the customary setup used by the empirical literature on cost-stickiness. Instead of studying the overall costs, we focus on R&D spending. Although there are subtle differences in motives for controlling the expenses related to sales, general and administrations compared to those for controlling R&D, the incentive structure for the managers is similar, as is the principal-agent relationship between the managers and the shareholders.
Managers, as agents, act on behalf of shareholders. Yet, managers’ and shareholders’ interests might diverge, as shareholders are interested in the long-term performance of their shares while managers seek to maximise their (short-term) utility. At the same time, managers are not homogenous, some possess higher others lower abilities to understand the business environment and to make best use of firm’s capabilities. Such management ability can be used to in times for sales decline to retain R&D investments to reap future benefits while low ability managers might engage in earnings management activities either to protect their short term interest or to compensate for their relatively low ability compared to firms in the same industry.
In the first paper we find that when sales revenues fall, managers who are motivated to avoid loss or profit decreases will aggressively cut R&D costs. When managers need to manipulate earnings, more capable managers will increase R&D more slowly when sales increase and reduce R&D more quickly when sales decline. China’s investor landscape is increasingly dominated by institutional investors and particularly state ownership. Research found that large blockholders are often more long-term oriented, give that they cannot quickly sell off their shares without causing a decline in the firm’s share price. This increases the motivation for blockholders to supervise management to secure their interest, which is to overcome the free rider issue faced by divergent shareholding (who can diversify their shareholding and to whom it is relatively costly while less beneficial to supervise management -providing benefits to other shareholders). At the same time blockholders have access to private information through direct interaction with management and the board. Blockholders could also have representatives on the board to exercise more supervision over management and to secure their long-term interests, which might include to remain R&D activities when firm’s sales decline.
The findings of the second paper suggest that highly capable management, in the presence of high ownership concentration, strengthens R&D expense stickiness. This underpins long-term focused blockholders encourage R&D and are more ‘forgiving’ when sales and profits decline. Concentrated ownership also means that the shareholders monitor the management better, making it harder for the management to extract wealth for themselves through earnings management.
Another important aspect to shape firm’s R&D investment is market competition. Firms do not operate in a vacuum but compete at least with firms in the same industry. Consequently, management cannot afford to ‘fall behind’ and might have the motivation to retain R&D despite sales declines.
Empirical results for the third paper indicate that managerial ability is positively associated with R&D stickiness. Furthermore, this association is driven by market competition for the product. At the same time, management ability cannot be neglected, which shaded additional light on the effect of competition on R&D stickiness.
The above findings were subject to further scrutiny through a battery of tests, including mechanism tests commonly used in research.
To summarize, these findings have extended our knowledge on how principal-agent conflicts between management and shareholders may play out with regards to R&D expenditure under different scenarios. The findings of the thesis provide insights into the role of concentrated ownership and principal-principal conflicts suggested by agency theory. The research also extends the resources-based view on the interaction between resources: managerial ability and R&D under product market competition. Implications of these findings are discussed.

Date of Award	Jul 2024
Original language	English
Awarding Institution	University of Nottingham
Supervisor	Shuai Yuan (Supervisor) & Ada Hoi Yan Ma (Supervisor)