Identifying bubbles is not all that hard; very few bubbles are not widely recognized before they burst, and very few not-bubbles are widely identified as bubbles.

Predicting the timing of a bubble bursting, OTOH, is hard, as is predicting the natural support level they will revert to.after bursting.

There are means of prediction/identification that will not identify exactly when it collapses but identifies the potential.

Has the cost of the thing separated from it's usefulness? (Bitcoin is too slow and expensive for transactions. It's too volatile to be a storage of worth.)

Does slapping unrelated products with it's name boost that products purely from branding?

New financial instrument backing the bubble? In this case we have the holy grail where the backing instrument is the thing itself! https://www.forbes.com/sites/investor/2017/05/31/cryptocurre...

Bubble Talk Cycle: Not talking about a bubble, discourse of a bubble, denying a bubble, seeing a bubble, late investors rushing into to get gains from the bubble.

Shoeshine boy: Everyone is excited about bitcoin, seeing the large numbers you start getting random people proffering as the next sure thing.

> but identifies the potential

That's not what the parent said. They said "detect one".

"detect" != "identifies the potential"

no?

Identifying bubbles with absolute certainty may be impossible, but identifying them with a reasonable amount of confidence is possible. They follow similar paths.

Of course, there's the old saying that the financial media has correctly predicted 20 of the last 3 bubbles

I think the original quote is "The stock market has forecast nine of the last five recessions." 1966 https://en.wikiquote.org/wiki/Paul_Samuelson But the idea probably applies to most predictions :)

Do you have a data model you can share that would show "a reasonable amount of confidence"?