Scribd
Upload
5 views4 pages

Autocorrelation in Econometrics

Autocorrelation in econometrics refers to the correlation of a variable with its own past values, often seen in time-series data where regression residuals are not independent. It can lead to inefficient estimates, misleading inference, and model misfit, necessitating detection through graphical analysis and statistical tests like the Durbin-Watson test. Remedies include modifying the model, using Generalized Least Squares, adjusting standard errors, and checking for specification errors to ensure valid model outcomes.

Uploaded by

bronishan241
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
5 views4 pages

Autocorrelation in Econometrics

Autocorrelation in econometrics refers to the correlation of a variable with its own past values, often seen in time-series data where regression residuals are not independent. It can lead to inefficient estimates, misleading inference, and model misfit, necessitating detection through graphical analysis and statistical tests like the Durbin-Watson test. Remedies include modifying the model, using Generalized Least Squares, adjusting standard errors, and checking for specification errors to ensure valid model outcomes.

Uploaded by

bronishan241
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
You are on page 1/ 4

Autocorrelation in Econometrics

Autocorrelation refers to the correlation of a variable


with its own past or lagged values. In econometrics,
it typically arises in time-series data where the
residuals (errors) of a regression model are not
independent but exhibit a systematic pattern over
time. This violates one of the classical assumptions
of Ordinary Least Squares (OLS) regression, namely
that the error terms are uncorrelated (E(utus)=0E(u_t
u_s) = 0 for t≠st \neq s).

Key Characteristics
1.Presence of Patterns: Autocorrelation indicates
that current values are influenced by past values.
For example, in economic data, GDP growth in
one quarter may depend on growth in previous
quarters.
2.Positive or Negative:
o Positive autocorrelation: Successive

residuals are positively correlated; errors


tend to persist in the same direction.
o Negative autocorrelation: Successive

residuals tend to alternate in sign.


Consequences of Autocorrelation
1.Inefficient Estimates: OLS estimates remain
unbiased but are no longer efficient (i.e., they
don’t have the minimum variance among linear
unbiased estimators).
2.Misleading Inference:
o Standard errors of the estimates are

underestimated or overestimated.
o tt-statistics and FF-statistics become

unreliable, leading to incorrect hypothesis


testing.
3.Model Misfit: The presence of autocorrelation
signals that the model may be missing important
explanatory variables or lags.

Detection of Autocorrelation
1.Graphical Analysis: Plot residuals against time
or lagged residuals to visually inspect patterns.
2.Statistical Tests:
o Durbin-Watson Test: Common for first-

order autocorrelation. Values near 2 indicate


no autocorrelation.
o Breusch-Godfrey Test: Suitable for higher-
order autocorrelation.
o Ljung-Box Test: Often used for more
general autocorrelation detection.

Remedies for Autocorrelation


1.Modify the Model:
o Include lagged variables or differencing to

account for the time-series nature.


o Use autoregressive distributed lag (ARDL)

models.
2.Generalized Least Squares (GLS):
o Transform the model to eliminate

autocorrelation.
3.Newey-West Standard Errors:
o Adjust standard errors to account for

autocorrelation and heteroskedasticity.


4.Check for Specification Errors:
o Ensure all relevant variables and dynamics

are included.
Autocorrelation is especially important in time-
series econometrics and requires careful handling to
ensure valid and reliable model outcomes.

You might also like

pFad - Phonifier reborn

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.

Note: This service is not intended for secure transactions such as banking, social media, email, or purchasing. Use at your own risk. We assume no liability whatsoever for broken pages.


Alternative Proxies:

Alternative Proxy

pFad Proxy

pFad v3 Proxy

pFad v4 Proxy