Time-varying evidence of predictability of financial stress in the United States over a century : the role of inequality

Abstract

In this paper, we analyze time-varying predictability of financial stress due to growth in income inequality of the United States (US) over the annual period of 1913 to 2016. In order to ensure that we remove the asset price effects on income inequality, and provide incorrect inferences regarding the impact on financial stress, we work with capital-gains excluded six alternative measures of top shares of pretax income and wages. We find that the top 10%, the top 10% to 5%, and the top 5% to 1% inequality growth rates tend to predict financial stress relatively better than the corresponding inequality growth rates associated with the top 1%, top 0.1%, and the top 0.01% of the income distribution. Moreover, all the six metrics of inequality growth is capable of predicting the heightened financial stress observed during the onset of the Great Depression and the same associated with the recent global financial crisis. Finally, our in-sample evidence of predictability tends to carry over to an out-of-sample forecasting exercise under four out of the six measures of inequality considered, and in particular for the broader measures of inequality – a result consistent with our in-sample analysis.