Is autocorrelation an indication of Non Stationary Series

[Khurram]

I have time series data and it has following Autocorrelation plot for each lag , x is lag number and y is correlation

I read https://support.minitab.com/en-us/minitab/18/help-and-how-to/modeling-statistics/time-series/how-to/... 

at bottom of this link

"A stationary time series has a mean, variance, and autocorrelation function that are essentially constant through time. The data is non-stationary when there is a large spike at lag 1 that slowly decreases over several lags. If you see this pattern, you should difference the data before you attempt to identify a model. To difference the data, use differences. Once you difference the data, obtain another autocorrelation plot."

do we really need to have constant autocorrelation for each lag for data to be stationary?

Thanks

[HarrySnart]

Stationarity in a time series effectively means the series fluctuates around a given mean for the series, this is a useful property for creating a reliable forecast because you are modelling the fluctuation of the series around the mean, not the series trend. When you have non-stationarity such as an upward trend in a stock price, for example, typically you difference it to make the series stationary. This is where models such as ARIMA are useful as they encapsulate three important aspects of time series forecasting:

Autocorrelation (AR): how does x(t) depend on x(t-1), x(t-2), x(t-n)...

Integration (I): how does the series need to be differenced to make it stationary

Moving Average (MA): for how long do exogenous shocks impact the model not captured in the series autocorrelation

[Ksharp]

It would be better if you post it at Forecasting Forum.

https://communities.sas.com/t5/SAS-Forecasting-and-Econometrics/bd-p/forecasting_econometrics