This study examines the role of business groups in creating innovative firms using micro data from Japanese firms. The main findings are as follows. First, the partially owned subsidiaries (POs) which are created or operated with the involvement of a business group are important for innovation. Of newly established firms, including independent firms, 19% of the new firms that exceed a certain threshold (50 employees and 30 million yen of capital) are PO firms, but they account for 47% of R&D and 90% of patent holdings.
Second, PO governance provides an important means of developing new companies. About 23% of the PO companies established during the observation period converted from independent companies, and 8% of the PO companies became independent. Independent companies that converted to POs tend to be large and have high R&D investments on average. Compared with the control group matched by industry and firm age, both R&D spending and patent holdings decreased significantly after the introduction of PO governance, but its sales increased, and profit margins improved. These results suggest that the duplication of R&D and other spending was reduced through the process of becoming a group firm. At the same time, the greater the increase of the equity capital due to the governance change, the greater the expansion of R&D and the larger the profit margin increase. In particular, when the governance change of an independent firm is associated with a change of industry affiliation from the same industry with the parent firm to a different industry, its R&D expenditure and equity capital increase significantly. These points suggest that internal capital markets can alleviate equity capital shortages for undertaking such transformation and R&D.
On the other hand, when a PO governance company becomes independent, R&D investment and profitability decrease compared with the control sample with matched industry, firm age, and growth rate before independence. This tendency is stronger for firms with the same industry affiliation as the parent firm. However, for PO companies with low parent company stakes before independence, the decrease in R&D investment is small and the number of patents held increases, suggesting that the capability of responding to stronger incentives significantly explains the performance from independence.
Third, as the more diversified the parent firms are, the higher their level of technological assets are, and the higher their Tobin's Q are, the greater number of R&D-intensive PO subsidiaries are created. Even if the fixed effects of the parent firms are controlled, their R&D scale and diversification are positive and significant for the creation of such firms. Additionally, firm age was found to be positive and significant, which also clarified that conventional cross-section analysis on the effect of firm age on new firm creation may result in a wrong conclusion. These suggest that the accumulation of technological assets and the age of the parent firm are important for new firm creation, and that PO firms are important channels for realizing the technological potential of existing firms.