The Relationship Between Vertical Integration and Risk. A Statistical Analysis of Changes in Vertical Integration within Industries Affected by the Global Semiconductor Shortage

Abstract: Succeeding a brief discussion of what vertical integration (VI) is, its interaction with risk and uncertainty, and the context of this study, namely the COVID-19 pandemic and its simultaneous semiconductor shortages, the purpose of this paper is outlined to fill a research gap regarding an empirically supported relationship between vertical integration and systemic risk. The aim and contribution here is subsequently twofold: to examine whether one can predict strategic trends and behavior among multinational companies when it comes to value chain ownership and control decisions in response to abnormal levels of systemic risk, and, to more generally statistically prove or disprove the seeming theoretical consensus that vertical integration can be a key risk-reduction strategy for various reasons, something that will be detailed and explored more closely in an extensive literature review. The literature review is guided principally by Transaction Cost Theory, otherwise referred to as Transaction Cost Economics and commonly abbreviated to TCE, in order to provide a theoretical background, connection, and support to both the purpose and expected and actual results of this study. To answer the research question of whether or not there is a statistically significant and verifiable relationship between vertical integration and systemic risk, this paper uses a quantitative approach to measure the change in vertical integration level and systemic risk between two specified years of interest, one representative of the semiconductor shortage effects and the other a base year, for 101 multinational companies, or MNCs, operating in semiconductor-reliant industries. Simple linear regression is used to test for a relationship between the independent variable, the change in systemic risk, and the dependent variable, the change in vertical integration level. Finally, the resulting finding of a positive and statistically significant relationship between the two is then critically analyzed with a connection to relevant theoretical assertions, whereafter the limitations of this study and their potential to segue into future studies are discussed.