Is it stationary

Identifying whether a time series is stationary or non-stationary is very important. If it is stationary you can use ARMA models to predict the next values of the time series. If it is non-stationary then you cannot use ARMA models, however, as you will see in the next lesson, you can often transform non-stationary time series to stationary ones.

In this exercise you will examine some stock and earthquake data sets in order to identify which are ready for ARMA modeling, and which will need further work to make them stationary.

Question

The top plot shown is a time series of Amazon stock close price.

Is the stock close price stationary?

Possible Answers

No, because in the top plot, the dynamics change with time.
No, because in the top plot, the variance changes with time.
No, because the top plot has a trend.
Yes, the top plot appears to be stationary.

Exercise

Is it stationary

Instructions 1/3

Question

Possible Answers