My doctoral thesis consists of four single papers (chapters) which are intended for publication in academic journals. Although all these papers are related and cover different aspects of a common theme, each of them can be read on its own and does not require any prior knowledge from the other chapters. It naturally follows that this style of presentation results in some overlap between the chapters, e.g. with respect to the core model, data description, and related literature. The readers of my entire dissertation will hopefully excuse the presence of potential redundancies that cannot be avoided at this point. Up to the date of submission of this dissertation in February 2017, two out of four papers have been published in international peer-reviewed journals. The article “Gibrat’s law redux: think profitability instead of growth”, co-authored with Mishael Milakovi´c and Simone Alfarano, was published in Industrial and Corporate Change, 2016, Vol. 25, No. 4, 549-571. The second article ”The real versus the financial economy: a global tale of stability versus volatility”, co-authored with Niels Förster, Simone Alfarano, and Mishael Milakovi´c, was published in Economics: The Open-Access, Open-Assessment E-Journal, 2014, Vol. 8, 2014-17. The two remaining papers “The ultimate corporate objective is survival” and ”A statistical equilibrium approach to forecasting corporate profitability” are unpublished to date. In one way or another, all four essays revolve around statistical regularities in the dynamics of business firms. My interest in the behavior of large, publicly traded corporations rests on the granular view of aggregate fluctuations that has received growing attention in recent years (e.g. Acemoglu et al., 2012; Carvalho and Gabaix, 2013; Gabaix, 2011). The fundamental idea of this literature is that the dynamics of large corporates may be of huge quantitative importance for macroeconomic quantities, e.g. GDP volatility, implying that a better understanding of these dynamics may contribute to a kind of new “micro-foundation” of macroeconomics that is empirically sound and well-founded. Although macroeconomic insights are beyond the scope of this dissertation, I would hope that the identification of empirical regularities in firm dynamics is a first step towards such a theory. To explain how these four papers are related and why they hopefully contribute to a larger whole, it is instructive to add some remarks about the underlying vision and research methodology. At the most fundamental level, the common theme underlying all papers is the problem of aggregation that economic theory often circumvents with the representative agent paradigm. If agents were homogeneous, it would be sufficient to consider the behavior of a single agent and the aggregate properties of the system could be derived from the behavior of this individual. However, research originating in the field of statistical physics suggests that this reasoning is misleading in complex systems that are characterized by a large number of heterogeneous interacting agents, or as Aoki and Yoshikawa (2007) put it (p. 26), “Micro behaviors of the representative agent do not mimic the behavior of the macroeconomy. Macroeconomic phenomena are the outcomes of interactions of a large number of economic agents such as households and firms.“ Statistical physicists developed an alternative methodology to describe these systems. Their approach builds on the perception that complex dynamics and interactions among the system’s constituent units may produce robust statistical regularities at a higher level of aggregation that do not directly reflect the behavior of the subsystems. Following this idea, we need to shift the focus from a deterministic to a probabilistic modeling approach that takes into account the presence of fluctuations (even in equilibrium), or to quote Aoki and Yoshikawa (2007) again (p. 26), “Equilibrium in the macroeconomy is better described by a probability distribution than by a ‘point’ in some space or set.” Thus, the present collection of articles approaches the dynamics of business firms from a probabilistic perspective. To detect these macro-regularities, one needs to select suitable quantities to characterize the states the system may take at every point in time. Browsing the literature that has been published in the research field of industrial dynamics so far (part of which is quoted in the first chapter), one is left with the impression that most papers in this vein focus on the growth rate of firm size. To this end, the first chapter of this dissertation explores the cross-sectional and time series properties of profit and growth rates in closer detail to show that profit rates are economically more fundamental and statistically more convenient than growth rates, and that Robert Gibrat’s seminal idea of a common law governing the dynamics of all firms applies to profitability but not firm growth. The paper presented in the second chapter replicates some of the results pertaining to the statistical properties of profit and growth rates on a much broader basis considering data of about 30,000 firms from more than 40 countries. It shows that the stability of the average profit rate and its volatility is not a peculiarity of the US data but a rather universal feature that can be observed across countries. Defining firm size in terms of market value, this paper also contributes to the literature stressing “excess volatility” in financial returns. The third chapter investigates the impact of firm idiosyncrasies on profitability. Given that profit rates are in statistical equilibrium, one should expect that profit rates are governed by a common probabilistic law of motion for all firms. Such a view is diametrically opposed to previous findings in the industrial organization, strategic management, and accounting and finance literature which stress the importance of sectoral or firm-level effects on performance. It turns out that survival time may be a potential explanation for the emergence of such divergent views because a common law for profitability only prevails conditional on survival. This leads to the fundamental question of what the ultimate business objective should be. Finally, the fourth chapter presents an application of the statistical equilibrium methodology to forecasting profitability. Overall, I think that the results presented in this dissertation suggest that statistical equilibrium is a very reasonable first order approximation to the profitability of surviving corporations, and that methods for the modeling of complex systems do have explanatory power for problems in economics. One of the major questions for future research should be if these insights on the dynamics of large, long-lived corporations can be exploited to draw inferences on the macroeconomy.weiterlesen