Autonomy of profit rate distribution and its dynamics from firm size measures : a statistical equilibrium approach

This paper presents an empirical analysis of the distributional and dynamic properties of firm profit rates, measured by returns on assets, using panel data on 1095 long-lived Japanese (non-financial) listed firms over the 1971-2012 period. In particular, this paper tests the validity of statistical equilibrium approach of Alfarano et al. (2012), by investigating whether the two representative firm size measures of total assets and total sales are the significant determinants of key parameters ruling over the distributional outcome and stochastic motion of firm profit rates: a system-wide aver- age rate of profit, a system-wide dispersion measure of profit rates, and an idiosyncratic noise factor reecting individual firm characteristics. Employing information-theoretic model selection approach and standard panel data econometric techniques which control for both unobserved individual firm heterogeneity and time effects, this paper finds: (i) under the various levels of aggregation using the two size measures as firm classification instruments, the empirical density of profit rates is well described by the Laplace distribution; (ii) the key parameters characterizing the profit rate distribution and its dynamics are independent of the movements in firm size measures. These findings confirm the fundamental predictions from statistical equilibrium approach and the finding (ii) implies that firm competition is an autonomous system, immune to the size of individual firms.