Testing The Limits: A Robustness Analysis Of Logistic Growth Models For Life Cycle Estimation During The COVID-19 Pandemic

Abstract

The semiconductor industry operates in a dynamic environment characterized by rapid technological advancements, extensive research and development investments, long planning horizons, and cyclical market behavior. Consequently, staying vigilant to technological disruptions and shifting trends is crucial. This is especially challenging when external shocks seriously affect supply chain processes and demand patterns. Particularly, recent events, such as the COVID-19 pandemic, the ongoing Russian invasion of Ukraine, and high consumer price inflation impacting the semiconductor cycle emphasize the need to account for these influences. In this context, we analyze growth patterns and life cycles of various technologies within the semiconductor industry by estimating logistic growth models. The logistic growth model was originally formulated to describe population dynamics. However, many processes outside the discipline of ecology share the fundamental characteristics of natural growth: self-proportionality and a self-regulating mechanism. Out of the different applications, two are of particular interest in the context of strategic business decisions: (1) modeling innovation diffusion and technological change to predict the mid- to long-term growth of a market, and (2) modeling of product life cycles. To obtain market growth and life cycle predictions, we apply the logistic growth model to forecast cumulative revenues by technology over time. This model treats the analyzed technology as a closed system. However, in practice, external shocks are the norm. To analyze the robustness to such external shocks, we compare technology life cycle estimates derived from logistic growth models before and after the effects of COVID-19 became evident for a wide array of semiconductor technologies. We find that the impact of COVID-19 on these life cycle estimates is mixed, but the median change is low. Our findings have implications for the application of logistic growth models in strategic decision-making, helping stakeholders navigate the complexities of technological innovation, diffusion, and market growth.