Ask Colin

How is the use of the Fibonacci ratio justified in technical analysis?

The question in full:

As a relative newcomer to the field of TA I am amazed at the credence given by many experts in the field to some concepts which seem to have no plausible justification. One example is the use of ratios derived from the Fibonacci series, as applied to index and/or stock prices to forecast advances and retracements. In the "rabbit population" example (which I believe came from Fibonacci himself) it is easy to see how the series arises from the initial assumptions. Other concepts in TA, such as "support" and "resistance", lend themselves to plausible explanation/justification. However, in the several "Fibonacci" articles I have come across in the ATAA Journal and elsewhere I have not found any attempt at a justification. It seems to be simply taken for granted that "it works" though most of the examples are far from convincing to the sceptical reader. Do we have to attempt a justification by assuming that the Fibonacci theory is so widely accepted that it has approached a "self fulfilling prophecy" - ie so that a significant proportion of the investing community is avidly plotting these ratios from all significant chart features, and making buy/sell decisions accordingly? Or is there a simpler justification?

My original response in 2001 was this:

I am not personally a user of Fibonacci ratios. However, Tony Plummer, who will be teaching at the Australian Trading and Investment Camp in October 2001, is an expert in this field. He has kindly penned an explanation, which I have also requested be published in the ATAA Journal. It is in a pdf file.

This is an update at August 2012:

The Fibonacci is widely used in technical analysis as you have observed. I have never seen a justification for it in terms of the psychology of interection between buyers and sellers. Mostly its adherents try to justify it by showing lots of examples, which are at best approximate.

Also, in September 2006 a research paper was published by academics Roy Batchelor and Richard Ramyar called "Magic Numbers in the Dow" in which they found that Fibonacci numbers only seemed to work as often as would happen by pure chance. This research paper is available on the internet and is worth reading.