The Economic Powerhouse: Corporate Profitability and Growth from the 19th to the 21st Century

Abstract

Three research projects are at the core of this thesis. Their common point of departure is the seminalpaper by Alfarano et al. (2012), in which the authors analyze the statistical and distributional patternsthat characterize profitability of corporations in the U.S. during the years 1980 to 2011. They find theprofitability of competitive corporations, measured by their return on assets (ROA), to be bestdescribed by a stationary diffusion process comprising a deterministic mean-reverting tendency andstochastic noise. The parameters of the process then are the median (location) and mean deviationfrom the median (dispersion) of the profit rates of the overall population of competitive firms and afirm-specific diffusion constant. In this thesis, we find this process to describe the statistical behaviorof profit rates in very different countries and very different time periods. Time-series and cross-sections of profit rates fall onto Laplace distributions for the U.S. between 1863 and 2013 (chapter 3),for Germany between 1919 and 1944 (chapter 4) as well as 45 countries around the world between1983 and 2013 (chapter 5). So profit rates display qualitatively the same statistical behavior,culminating in the Laplace distribution of profit rates, across time and space. Significant departuresfrom the distribution are transitory and only occur under exceptional circumstances, such as severeeconomic depressions or wartime. Quantitative differences in the parameter values, however, appearpersistently across countries, potentially being the result of country-specific characteristics like thelegal system. Thus, departures from the distribution mark exceptional economic conditions and severedistress while different parameter values seem to be the result of country-specificities. Overall,profitability qualifies as reliable vital signs of the economy with robust statistical patterns across spaceand time. Departures from these statistical patterns then carry substantial information about thepotential length and severity of crises.