BookmarkSubscribeRSS Feed
Fluorite | Level 6
Hi all,

I am new to the field of econometrics and appreciate any assistance to make the theories to be more easily understandable.

If I tested a time series and it is stationary. However, does it mean that I do not have to test for heteroscedasticity or autocorrelation? Since a stationary process means constant mean/variance/autocorrelation?

Thank you in advance.
SAS Employee

Hi @itslarajean , stationarity refers to the series moving around a fixed mean, this is important in forecasting as you're trying to predict future values based on the self-relationship of the series. Typically this is done by modelling the autoregression (relationship between x_t, x_t-1...) and the moving average (which models the impact of exogenous shocks to the series). 

Where data is already stationary, one might use an ARMA model. If the data needs differencing to make it stationary that's when the more generalised ARIMA is used. With regards to heteroskedasticity this is where there is some stochastic volatility in the series (such as financial time series) you may use a model like ARCH which models the variance of the error term. It would be worth going through some online resources to familiarise yourself with some of the different modelling techniques and their intuition. The Wiki page for Time Series is a good starting point 


Available on demand!

Missed SAS Innovate Las Vegas? Watch all the action for free! View the keynotes, general sessions and 22 breakouts on demand.


Register now!

Multiple Linear Regression in SAS

Learn how to run multiple linear regression models with and without interactions, presented by SAS user Alex Chaplin.

Find more tutorials on the SAS Users YouTube channel.

Discussion stats
  • 1 reply
  • 1 like
  • 2 in conversation