Cover Image
Read Now

Autocorrelation-Corrected Standard Errors in Panel Probits An Application to Currency Crisis Prediction

Many estimates of early-warning-system (EWS) models of currency crisis have reported incorrect standard errors because of serial correlation in the context of panel probit regressions. This paper documents the magnitude of the problem, proposes and tests a solution, and applies it to previously publ...

Full description

Bibliographic Details
Main Author: Berg, Andrew
Other Authors: Coke, Rebecca
Format: eBook
Language:English
Published: Washington, D.C. International Monetary Fund 2004
Series:IMF Working Papers
Subjects:
Dynamic Treatment Effect Models
Financial Crises
Currency; Foreign Exchange
Exports And Imports
Simulation Methods
Time-series Models
Estimation
Foreign Exchange
Currency Crises
International Trade
Capital And Ownership Structure
Goodwill
Macroeconomics
Estimation Techniques
Early Warning Systems
Econometrics
Financial Risk And Risk Management
Econometrics & Economic Statistics
Financing Policy
Financial Risk Management
Single Equation Models
Economic & Financial Crises & Disasters
Macroeconomic Aspects Of International Trade And Finance: Forecasting And Simulation
Econometric Analysis
Single Variables: Discrete Regression And Qualitative Choice Models
Value Of Firms
Diffusion Processes
Trade: General
Crisis Management
International Economics
Export Performance
Exports
Econometric Models
Real Exchange Rates
Dynamic Quantile Regressions
'panel Data Models
Spatio-temporal Models'
Argentina
Online Access:
Collection: International Monetary Fund - Collection details see MPG.ReNa
Description
Summary:Many estimates of early-warning-system (EWS) models of currency crisis have reported incorrect standard errors because of serial correlation in the context of panel probit regressions. This paper documents the magnitude of the problem, proposes and tests a solution, and applies it to previously published EWS estimates. We find that (1) the uncorrected probit estimates substantially underestimate the true standard errors, by up to a factor of four; (2) a heteroskedasicity- and autocorrelation-corrected (HAC) procedure produces accurate estimates; and (3) most variables from the original models remain significant, though substantially less so than had been previously thought
Physical Description:21 pages
ISBN:9781451845860

Similar Items