Investigating the semiconductor industry cycles
Résumé
Abstract The aim of this article is to investigate the links between semiconductor sales and various macroeconomic, financial, industrial variables including inventories, equipment orders or semiconductor sector stock index. Statistical properties of these variables are studied. Both short-run and long-run interactions are analysed. On the short-run, our results indicate that relationships often imply feedbacks. Through the implementation of cointegration analysis, we separately identify both sales value and investments in the semiconductor market. An impulse-response analysis confirms the relevance of our choice of data and stability tests demonstrate that the parameters remain constant during the entire sample. The Vector Error Correction Models (VECMs) offer a representation respecting cycle theories and market actor analyses.