Seminario "Pricing Mistakes and the Non-Neutrality of Money"
Martes 7/7, 11h
Seminario presentado por Tomás Caravello
Abstract
Using microdata from the UK and Argentina, I document that sizes of consecutive price changes are negatively correlated, and that this predictability increases with adjustment frequency. Standard sticky-price models (including random menu costs, multiproduct firms, imperfect or infrequent information, and mean-reverting desired prices) cannot jointly match these facts, while models with imperfect reset, where choosing prices precisely is costly and there are ex-post mistakes, can. In a model with endogenous inaction, I show analytically that imperfect reset increases monetary non-neutrality, and it can even exceed the standard theoretical upper bound. If reset can be imperfect, standard pricing moments (frequency and kurtosis of price changes) alone cannot identify non-neutrality. I derive a generalized sufficient statistics result that restores identification by augmenting those moments with predictability coefficients. Accounting for imperfect reset increases inferred non-neutrality in the UK data by at least 60%.
Tomás Caravello
Ph.D. Candidate in Economics at MIT. His main research interests include macroeconomics, monetary economics, time series econometrics and asset pricing.
Using microdata from the UK and Argentina, I document that sizes of consecutive price changes are negatively correlated, and that this predictability increases with adjustment frequency. Standard sticky-price models (including random menu costs, multiproduct firms, imperfect or infrequent information, and mean-reverting desired prices) cannot jointly match these facts, while models with imperfect reset, where choosing prices precisely is costly and there are ex-post mistakes, can. In a model with endogenous inaction, I show analytically that imperfect reset increases monetary non-neutrality, and it can even exceed the standard theoretical upper bound. If reset can be imperfect, standard pricing moments (frequency and kurtosis of price changes) alone cannot identify non-neutrality. I derive a generalized sufficient statistics result that restores identification by augmenting those moments with predictability coefficients. Accounting for imperfect reset increases inferred non-neutrality in the UK data by at least 60%.
Tomás Caravello
Ph.D. Candidate in Economics at MIT. His main research interests include macroeconomics, monetary economics, time series econometrics and asset pricing.