특허정보 기반 기술혁신 지표의 주가예측력 비교 연구

We estimate seven representative technology innovation indicators proposed by previous studies based on patent data registered by non-financial listed companies in Korea from 1985 to 2019, and verify whether these indicators possess meaningful predictive power for stock prices and generate excess profits. Through cross-sectional regression analysis of future stock returns of sample companies, we find that only R&D intensity (RDI), defined as R&D expenditure divided by market capitalization, shows significant explanatory power during the sample period after 2000. This result is consistent with the profitability analysis of portfolios constructed based on the seven indicators, where only portfolios constructed based on RDI exhibit a monotonically increasing return as RDI increases. Based on these findings, a hedge portfolio composed of buying the highest RDI group and selling the lowest RDI group achieves a significant market-adjusted excess return of 0.75% per month on average, while a value-weighted portfolio achieves an even higher market-adjusted excess return of 1.16% per month on average. The predictive power of RDI is more pronounced in large-cap stocks than in small-cap stocks, remains unchanged after risk adjustment using a four-factor model, and remains significant up to three years after portfolio construction. This suggests that information related to technology innovation indicators such as RDI is reflected with a significant delay in stock prices due to investors' limited attention.