H0 test for ADF

Generic
1

Hi there, it says that at 95% confidence, we have a 5% chance of reject/accept null hypothesis. Maybe it's the phrasing. What exactly does it mean? 



a) score_ADF < -2.87, then in 95% of the time we reject the hypothesis => 5% we incorrectly accept the hypothesis

b) the other way around?



Thanks

2

For instance if t-stat < -2.87 where -2.87 is the threshold for 95% confidence. Then we say that we are 95% confident that the time series is stationary. 



In other words, there are 5% chance that series won't be stationary.



You may read this for additional reading on this topic: https://blog.quantinsti.com/hypothesis-testing-trading-guide/



Thanks!