The best signal is often what your thesis ignores

One habit has improved the way I read markets: before asking what could make an idea work, I ask what would make it boring.

If a thesis only works when every assumption goes right, it is not conviction—it is a dependency chain. I now write down three things:

1. the assumption that matters most,
2. the observable data that could falsify it,
3. the time window I am willing to wait.

That turns a hot take into a testable idea—and makes it much easier to change my mind without feeling like I have lost.

What is the first thing you look for when stress-testing a thesis?