Autokorelasi

Introduction to Econometrics Lecture 8 Autocorrelation INEMET [U13783] Guy Judge March 2007 Econometric problems INEMET [U13783] Guy Judge March 2007 Topics to be covered Overview of autocorrelation First-order autocorrelation and the Durbin-Watson test Higher-order autocorrelation and the Breusch-Godfrey test Dealing with autocorrelation Examples and practical illustrations INEMET [U13783] Guy Judge March 2007 Autocorrelated series and autocorrelated disturbances INEMET [U13783] Guy Judge March 2007 Overview of autocorrelation What is meant by autocorrelation The error terms are not independent from observation to observation – ut depends on one or more past values of u What are its consequences? The least squares estimators are no longer “efficient” (i.e. they don’t have the lowest variance). More seriously autocorrelation may be a symptom of model misspecification How can you detect the problem? Plot the residuals against time or their own lagged values, calculate the Durbin-Watson statistic or use some other tests of autocorrelation such as the Breusch-Godfrey test How can you remedy the problem? Consider possible model re-specification of the model: a different functional form, missing variables, lags etc. If all else fails you could correct for autocorrelation by using the Cochrane-Orcutt procedure or Autoregressive Least Squares INEMET [U13783] Guy Judge March 2007 First-order autocorrelation INEMET [U13783] Guy Judge March 2007 The sources of autocorrelation INEMET [U13783] Guy Judge March 2007 The consequences of autocorrelation INEMET [U13783] Guy Judge March 2007 Detecting autocorrelation INEMET [U13783] Guy Judge March 2007 The Durbin-Watson test INEMET [U13783] Guy Judge March 2007 More on the Durbin-Watson statistic INEMET [U13783] Guy Judge March 2007 Using the Durbin-Watson statistic INEMET [U13783] Guy Judge March 2007 Durbin-Watson critical values INEMET [U13783] Guy Judge March 2007 The Breusch-Godfrey (LM) test INEMET [U13783] Guy Judge March 2007 The Breusch-Godfrey test continued INEMET [U13783] Guy Judge March 2007 Dealing with autocorrelation How should you deal with a problem of autocorrelation?  Consider possible re-specification of the model:  a different functional form,  the inclusion of additional explanatory variables,  the inclusion of lagged variables (independent and dependent)  If all else fails you can correct for autocorrelation by using the Cochrane-Orcutt procedure or Autoregressive Least Squares INEMET [U13783] Guy Judge March 2007