Short excerpt of a YouTube talk by Kosta Ouzas explaining an indicator used to identify "turbulent" regions associated with market bubbles. The speaker describes that the indicator relies on "markup chains" to detect different probability states (e.g., 2 or 3 states) and suggests adding an extra transitional state to better capture regime changes.
This indicator has been very useful in the past because you could tell when we were in these turbulent region that are associated with bubbles. We spend one year going up when you're going down. But usually once we cross this line right where it's more sophisticated how the indicator finds these regions. It uses something that is called markup chains. Basically identify different state probability state states. And he says, oh, I see. You can say how many identify in the data two states, three states. In fact, we could add one more state that is a state of transition.
