We don't, in some cases. It is how you measure your universe and whether you place any limits on the physical constraints. Do we say when you take the integral of the wave function for instance, that you allow it to move from $$-\infty$$ to $$\infty$$ in time? Or do we say that the wave function of the universe is finite and there is some unseen location in the future of our universe which will define a symmetry in time?
Not only that, but time itself can be scrutinized under two different disciplines. One which allows a boundary for the universe, the other solution does not. So long as there is a big bang, this is an event measured in real time. If you move 90 degrees off this space angle, you move into the imaginary time axis. On this potential, your system will act as though each event is unique, but without a cause or reference to any other event.
Imaginary time was first viewed this way by professor Hawking in a search to see if he could find any way to remove the singularity. His solution is a strange one, because there is always a description of the big bang, only out of ignorance would you beleive this, or maybe out of being biased over one solution or another.
An end of the universe is just as significant as finding a beginning to the universe. Imaginary time wipes that description away. Real time events works differently. Perhaps the universe will undergo a rapid freezing era once it has expanded sufficiently. That is the most likely scenario.