Can a Factor Model with Macroeconomic Underpinnings Determine the Persistence of R&D Investments? The Case of Central and Eastern European Countries

In this paper I investigate whether a factor model that includes macroeconomic variables can determine the persistence of firms’ innovative activity. I model innovations as a continuous process with transition probability models that are a part of Markov chain analysis and I estimate factors with common principal component analysis. Focusing on Central and Eastern European public companies from 2010 to 2018, I find no evidence that factor models determine the persistence better than models without factors. Nonetheless, inclusion of factors increases the significance of the size effect. Specifically, smaller firms are less likely to start innovating than larger firms. However, if smaller companies can overcome the hurdle of starting the innovative activity, they are more likely to switch to higher states of innovativeness than large-scale enterprises. This effect is more significant on input innovations than on output. I also show that apart from firm-specific characteristics like size or ownership, the country’s market reforms (the switch from planned to market economy in the late 20th century) still affect the innovativeness of firms. Companies that operate in countries of fast market reforms are more innovative than firms in "slow-reformed" countries.