Universal laws and economic phenomena

Abstract

Despite the idiosyncratic behavior of individuals, empirical regularities exist in social and economic systems. These regularities often arise from simple underlying mechanisms which, analogous to the natural sciences, can be expressed as universal principles or laws. In this essay, I discuss the similarities between economic and natural phenomena and argue that it is advantageous for economists to adopt methods from the natural sciences to discover “universal laws” in economic systems. © 2011 Wiley Periodicals, Inc. Complexity, 2011

Commodity and resource ETF trading patterns during the financial crisis

“…Therefore, the possibility that the investigated data might be lognormal distributed cannot be ruled out, but conjecturing in line with the literature that the power law hypothesis holds, the consequences are not favorable for the prospect of predicting, regulating, or controlling trade and speculation in the goods (and respective ETFs) under investigation: As outlined by various scholars ,], not only do infinite moments present a problem, but also is the likelihood of tail events systematically underestimated. Although Taleb's argument holds that predictability with Gaussian statistics as commonly used in economics and finance is already severely impeded with infinite kurtosis (i.e., for α < 5), even lower exponents would tend to make it even more difficult.…”

“…The general case is most illustratively argued in models of spacial organization or implicit optimization of information packaging in languages widely discussed in 1960s; for a recent overview see Mitzenmacher . Gerig , in turn, gives the non‐normal price change distribution (although he does not explicitly assume a power law) for financial assets as one of his examples for universal laws in economics. Self‐organization mechanisms are also the basis for the different proposed models that explain the emergence of financial power laws.…”

“…Following the finding that log returns of financial assets may be power law distributed, the ramifications of this for the use of Gaussian statistics (as is common in economics and finance) have been vividly debated . As higher moments do not exist for power law distributions (depending on the exponent), models relying on such moments (potentially including mean and variance) may lead to flawed results, particularly the underestimation of tail events (colorfully termed “black swans” or “dragon kings”) such as financial crashes.…”

“… Log returns are commonly used in finance as they are an approximation of raw returns but have certain desirable qualities, among others, they are additive in time and lognormally distributed raw returns would show up normally distributed in log returns form (which is a well founded and convenient prior but turns out not to be supported by the data because log returns are still heavy tailed ).…”