Rolando Gonzales (Bayes Foundation): Modeling Hyperinflation Phenomena: A Bayesian Approach

Hyperinflations are short-lived episodes of economic instability in prices, characteristically lasting twenty months or less. Classical statistical techniques applied to these small samples could lead to incorrect inference problems. This paper describes a Bayesian approach for modeling hyperinflations which improves the modeling accuracy using small-sample inference based on specific parametric assumptions. A theory-congruent model for the Bolivian hyperinflation was estimated as a case study.