Please use this identifier to cite or link to this item: https://hdl.handle.net/10419/220156 
Year of Publication: 
2015
Series/Report no.: 
Discussion Paper No. 67
Publisher: 
Institute for Applied Economic Research (ipea), Brasília
Abstract: 
When analyzing the demand for money in high inflation processes it has been suggested [Tourinho (1995)] that we should consider not only the effects of changes in the expected inflation rate but also changes in the expected variability of inflation. The model in Lima & Ehlers (1993) is extended here to deal more accurately with the uncertainty produced by the variability of inflation: a term proportional to the expected quadratic error in forecasting inflation is included in the demand for money equation. The problem of what estimate to use for the expected variance of inflation, is addressed by a Bayesian estimation procedure. Model parameters are allowed to vary slowly over time and Bayesian monitoring and intervention procedures are then used to cater for structural changes. We estimate the model with data ranging from first quarter of 1973 to fourth quarter of 1995, thus taking into account many stabilization plans for the Brazilian economy. We find that the presence of variance of inflation in our money demand equation is important in two ways: a) it prevents the monitor from signaling again in 1990 after an intervention period in 1986 and b) its effect turns out to be significant after 1986 when many stabilization plans contributed to increase uncertainty.
Document Type: 
Working Paper

Files in This Item:
File
Size
357.4 kB





Items in EconStor are protected by copyright, with all rights reserved, unless otherwise indicated.