Conjoint effects of R&D on the financial performance of semiconductor companies: Rule-based granular computing

Research and development (R&D) is inherently important to high-tech industry, such as semiconductor companies. However, the complex and ambiguous relationship between R&D and the subsequent financial performance varies and is still underexplored. Previous studies based on statistical analyses are not able to provide individual guidance considering the specific circumstance/status of a company. Therefore, the aim of this study is to explore the inherent patterns or rules - by a two-stage soft computing model - to guide semiconductor companies to improve its financial performance, considering R&D and the other factors. This research uses a group of semiconductor companies in Taiwan as an empirical case, and the research period ranges from 2006 to 2013. Compare with the conventional discriminant analysis approach, the obtained decision rules from the proposed rule-based analytic model shows superior results that may classify the financial performance of semiconductor companies by a set of financial and R&D indicators with 75% accuracy in average. In addition, we also show how to use the retrieved knowledge (i.e., rules) to guide a real company to improve its performance in the future. The findings contribute to the understanding of R&D influences to the financial performance of semiconductor companies with managerial implications in practice.