1. ET life formed and vanished so quickly that they didn't even get to the stage where they generate electromagnetic signals.

2. ET life formed and got to generate EM waves, but:

- they're either too far from us that we haven't detected those waves yet;

- they only recently got to this stage and the signals they produce have not reached us yet;

- or both.

3. ET life formed way sooner than we did; figured out how to generate EM waves, and then vanished entirely. So, any signals of their civilization they must have sent out there have already reached earth, but before we began listening for such signals.

4. Then again, since there are practically infinite number of planets with habitable conditions for ET life, statistically speaking, we must have already received a signal from at least one of them.

Which leads me to this:

5. ET life doesn't exist, because the universe as we know it is most probably a simulation inside a computer and the computational resources can only render so many different objects at a time. (This is serious stuff, researchers have already proposed methods for verifying this: https://www.washington.edu/news/2012/12/10/do-we-live-in-a-c...).

Looking at our own universe, building a simulator for it inside itself looks like a non starter. We’re even detecting gravity waves from black hole collisions billions of light years away, that's how detailed the simulation would have to be. There’s no way to build a simulator with that level of detailed fidelity of a universe in that same universe. Imagine designing a computer chip to run a simulation of itself. Yes we run low fidelity weather simulations, but to simulate outcomes in our universe at a general level like that you wouldn’t need to simulate consciousness. In the same way that we don't think superintelligent AI would necessarily need to be conscious, the actors in a history simulation wouldn't need to be either. Our universe is too detailed to be a statistical simulation and too complex to be an accurate simulation. The parameters don’t make sense.

That only really leaves the possibility that the simulator is running in a dramatically more complex universe, so simulating ours would be relatively easy. But then, there would be almost infinitely many less complex universes they could simulate, why choose ours? The more complex the universe running the simulator(s), the more lower complexity universes there are available to simulate. As the number and complexity of universes available to run simulators goes up, the number of lower complexity universes (and the complexity of them) that it's possible for them to simulate also goes up, so the probability they will bother with ours probably stays steady at a very low prior.