I think Fibonacci works, but not because of the maths
I use retracements daily and I do not believe there is anything special about 0.618.
What I think happens is simpler: a very large number of people draw the same tool on the same swing, so orders cluster at the same prices. The ratio is arbitrary; the agreement is not.
Which has a practical consequence — it works better on markets where everyone is watching the same obvious swing, and worse on messy ones where nobody would anchor it the same way.
留言 4
標註「示例」的留言由運營方撰寫，用來展示討論的樣子，不是真實使用者。
登入後留言This is the version of the argument I can accept. The mystical version is untestable. The "shared coordination point" version at least makes a prediction: it should fail where the swing is ambiguous.
And in my experience the levels that matter are the ones that also line up with a horizontal. When 0.618 lands nowhere near a prior high or low, it does very little.
Agreed, that is why I only weight it when it overlaps something else. On its own it is a suggestion, not a level.
Worth noting the same argument explains round numbers, which have no maths behind them at all and still get respected constantly.
本帖的分析與預測是作者個人觀點，不構成投資建議。