Cycles review, part VIII: Fibonacci numbers can apply not just to price but also time

Many traders are familiar with the use of Fibonacci numbers in analyzing setups and price objectives. These are used in Fibonacci-based setups (such as those described in the book authored by Hobbs called "Fibonacci for the Active Trader" which I've found quite helpful), in Elliott Wave, and in other methods. Apparently there's also one by Larry Pesavento, Fibonacci Ratios with Pattern Recognition. Not as many people realize that Fibonacci numbers are also used for time analysis; and for that matter, some put time and price together with the use of Fibonacci arcs (which can look like spirals or circles). Fibonacci numbers are very useful in sizing up markets. For example, this evening there's news that Birinyi is projecting that the equities markets are about to launch higher into a bull market that will take the S&P up 88%, which I calculate to be approximately 1690. Hmm, I seem to remember that there is a number in that neighborhood that actually is a Fibonacci number! So I can agree that, if and when the market goes into bull market mode, that may be a price area to focus on. (We'd also want to look at 1.382 times the amount of the drop from 1576 to 667, which would point to 1923 SPX as one projection for an expanded flat "B" wave, in Elliott Wave parlance - also a Fibonacci calculation.) Meanwhile, do we think this is likely to happen? One of the methods we're using to size up that possibility is observing how the market's reacting to the near-term Fibonacci numbers we've been working with. For example, the Nasdaq, the Dow Industrials, and the Dow Transports have - so far - exhibited probable reversal patterns at a 1.382 extension (NDX) and a .786 retrace (Dow Industrials and Dow Transports, see Not to add to your bearishness, but ... Dow Industrials' reversal from Fibonacci level may have larger significance (5/7/09)). If the markets move up higher above those levels, then we'll have to adjust our views and work with higher price targets.

But let's focus back to Fibonacci applied to time. I believe that the end of May time frame can be important because it's 1.618 years after the October 2007 highs. I also believe that this September 2009, and June 2010, can be important time points because they represent .382 of the range October 2002 to October 2007, and .382 of the range January 2000 to October 2007, respectively. As for the May-end time (which may also correspond neatly with Andy Askey's Gann analysis I'll post about next, being 90 days after the March 6-9 lows, not to mention the Bradley model, and some dates identified by Merriman) - I posted about this idea of 1.618 year from the October 2007 high in Fibonacci time as well as price projections can help identify probable market bottom (3/7/09) and you can use the "Fibonacci" label to find posts here that include Fibonacci analysis. But - I should mention that the idea of 1.618 year isn't really the typical way of calculating Fibonacci for time in the markets ... although let's face it, the March 6-9 time window at 1.382 year from October 2007 worked out pretty well! (Sure, there were LOTS of other important signals for that time window too - but it's neat at least, that it occurred at that 1.382 year point. And that the 1.618 year point coming up soon also correlates to other signals for this being another important time window.) The more standard ways to use Fibonacci calculations on the big picture would be the others I describe above, applying Fibonacci extension from a range measured from a significant swing low-to-high (or perhaps high-to-high in some cases), that point toward September 2009 and June 2010.

In my March 7 post, I quoted from a paper that presents these concepts succinctly, called "Fibonacci Studies with mGlider" and it's on the internet at http://www.mglider.com/fibonacci_studies.doc. (Please remember, I don't know that group and this is not intended to be any endorsement, just recognition that they put forward a good summary about this topic). Here's a brief recap from that paper, focused just on the Fibonacci numbers applied to time, and also the Fibonacci arcs or spirals that are intended to describe both time and price:

Fibonacci time projection days are days on which a price event is supposed to occur. Time projection analysis is not lagging but is of forecasting value. Trades can be entered or exited at the price change rather then after the fact. The concept is dynamic. The distance between two turning points is seldom the same, and time projection days vary, depending on larger or smaller swing sizes of the market price pattern. This base for drawing this shape is 2 critical points: two highs, two lows or a low and a high. Fibonacci levels are projected into the future based on those points and at this time it is impossible to say whether those levels mark peaks or valleys. If price is declining or rising approaching a given Time Projection level, it is likely this level will mark an end or a pause of a particular trend. It is always recommended to combine Time Projection with other Fibonacci tools for more dependable signals.


Fibonacci spirals provide the optimal link between price and time analysis and are the answer to a long search for a solution to forecasting both time and price. Each point on a spiral manifests an optimal combination of price and time. Corrections and trend changes occur at all those prominent points where the Fibonacci spiral is touched on its growth path through price and time.

Here is some information on one Fibonacci setup pattern, called the "Gartley" (I do believe it's named after someone called Gartley). As you look at it, you can see that it carries some implications about probable time, but it's really just based on measuring price only. This information is from http://www.investopedia.com/articles/trading/05/AdvFibonacci.asp:

The Gartley pattern is a lesser-known pattern combining the "M" and "W" tops and bottoms with various Fibonacci levels. The result is a reliable indicator of future price movements. Figure 4 shows what the Gartley formation looks like.

Figure 4: An example of what bullish and bearish Gartley Patterns look like.

Source: http://www.harmonictrader.com/

Gartley patterns are formed using several rules regarding the
distances between points:

X to D - Must be 78.6% of the segment range XA.
X to B - Must be near 61.8% of the XA segment.
B to D - Must be between 127% and 161.8% of the range BC.
A to C - Must be 38.2% of segment XA or 88.6% of segment AB.

How can you measure these distances? Well, one way is to use Fibonacci retracements and extensions to estimate the points. You can also download a free Excel-based spreadsheet from ChartSetups.com to calculate the numbers.

You can locate more background information on Fibonacci numbers at http://en.wikipedia.org/wiki/Fibonacci (which also features the portrait, shown here at upper right, by Leonardo of Pisa entitled simple "Fibonacci").