SSRN Author: Ole PetersOle Peters SSRN Content
http://www.ssrn.com/author=2561166
http://www.ssrn.com/rss/en-usThu, 27 Jul 2017 03:00:49 GMTeditor@ssrn.com (Editor)Thu, 27 Jul 2017 03:00:49 GMTwebmaster@ssrn.com (WebMaster)SSRN RSS Generator 1.0REVISION: An Empirical Test of the Ergodic Hypothesis: Wealth Distributions in the United StatesClassical studies of wealth inequality make the "ergodic hypothesis" that rescaled wealth converges rapidly to a stationary distribution. This allows powerful analytical techniques but constrains models. We test the hypothesis and find it unsupported by data. In a simple model of a growing economy with wealth reallocation, we fit the reallocation parameter to historical United States wealth data. We find negative reallocation, i.e. from poorer to richer, for which no stationary distribution exists. When we find positive reallocation, convergence to the stationary distribution is slow. Both cases invalidate the ergodic hypothesis. Studies which make it should be treated with caution.
http://www.ssrn.com/abstract=2794830
http://www.ssrn.com/1612226.htmlWed, 26 Jul 2017 18:31:03 GMTREVISION: An Empirical Test of the Ergodic Hypothesis: Wealth Distributions in the United StatesClassical studies of wealth inequality make the "ergodic hypothesis" that rescaled wealth converges rapidly to a stationary distribution. This allows the use of powerful analytical techniques but it also places a strong constraint on admissible models. Here we test this hypothesis and find that it is not supported by data. In a simple model of a growing economy with wealth reallocation, we fit the reallocation parameter to reproduce historical wealth inequality in the United States. To our surprise, we find negative reallocation, i.e. from poorer to richer, over the last four decades. Under such conditions, wealths diverge positively and negatively and no stationary distribution exists. When we find positive reallocation, it is so slow that convergence to the stationary distribution occurs over decades or centuries. In all cases the ergodic hypothesis is invalid. Studies which make it are, by design, unable to detect the dramatic conditions found here.
http://www.ssrn.com/abstract=2794830
http://www.ssrn.com/1562291.htmlTue, 31 Jan 2017 03:15:38 GMTREVISION: An Empirical Test of the Ergodic Hypothesis: Wealth Distributions in the United StatesClassical studies of wealth inequality make the "ergodic hypothesis" that rescaled wealth converges rapidly to a stationary distribution. This allows the use of powerful analytical techniques but it also places a strong constraint on admissible models. Here we test this hypothesis and find that it is not supported by data. In a simple model of a growing economy with wealth reallocation, we fit the reallocation parameter to reproduce historical wealth inequality in the United States. To our surprise, we find negative reallocation, i.e. from poorer to richer, over the last four decades. Under such conditions, wealths diverge positively and negatively and no stationary distribution exists. When we find positive reallocation, it is so slow that convergence to the stationary distribution occurs over decades or centuries. In all cases the ergodic hypothesis is invalid. Studies which make it are, by design, unable to detect the dramatic conditions found here.
http://www.ssrn.com/abstract=2794830
http://www.ssrn.com/1560032.htmlSun, 22 Jan 2017 15:53:43 GMT