fibonacci retrace and extension, how do you use them practically
hi there guys,
reading the lesson on this, i would like to know when do you use fib retrace and ext. you use both of them, or do you have to choose 1 over the other. they say 1 is used for resistance and support and the other is for taking profits. if you are using them, please let me know how you have been using them.
If you were to search around the forum for "fibonacci" or "fibonacci retracement", you're likely to find some great threads that explain the use of this fibonacci study and are filled with examples of fibonacci retracements in action. Discussion and use of fibonacci extensions is not as common.
Fibonacci extensions can be a little confusing, and I usually think of them as more of an intermediate level tool, whereas working with fibonacci retracements and projections are better to cut your teeth on. To answer your initial question: fibonacci retracements can be determined without use of extension, but not really the other way around. I'll explain why in a moment.
First: projections and extensions are very similar, but are constructed in different ways and are more or less useful in different scenarios. A couple notes before getting to that:
Fibonacci retracements are draw from a low to a high (or high to low) and are then used to map forward over time to determine those levels price is likely to pullback (throwback) to.
Building on that, fibonacci projections outline further levels to which price will gravitate if the .786% and then 100% retracement levels are breached. Once price has retraced a low to high/high to low 100%, it is no longer "re-tracing". Once price breaks through that 100% low or high and keeps going, the old retracement levels are good to have around for future support/resistance levels, but do nothing for determining where price is headed next. That's where projections fit in: 127.2%, 161.8%, 261.8%, 423.2% of the retracement are common projection points. Most fibonacci retracement studies plot projection points by default. I find projections highly useful and would recommend learning to apply them.
Building a bit further, we come to fibonacci extensions. The BP school does a great job of explaining them. Extensions follow a 1-2-3 pattern (in some cases, what's called a "measured move down" or "measured move up" chart pattern). Points 1 and 2 are established by the fibonacci retracement. Point 3 is the deepest point to which price then retraced. The "extension" is mapping the range of points 1 and 2 over the chart starting at point 3. In that way you come up with a 1-2-3-4 zig-zag pattern, where point 4 is the 100% extension. The range of points 1 and 2 is then equaled by the range between points 3 and 4; but as with projections, price can go beyond the 100% extension point.
The difference is that projections measure where price will go if it exceeds a 100% retracement. Extensions measure forward movement of the 3-4 wave. So, you can think of projections as a V where the right side continues above the highest level of the left side, while extensions are lightning bolt shape where the bottom right tail can potentially keep going.
By definition, projections build off of retracements. Extensions also need a retracement to be constructed, but that retracement (point 3) might not coincide exactly with a fibonacci level. Regardless, once point 3 is in, fibonacci ratios are used to determine the extension levels from point 3 to 4 and beyond. You can see then, why retracements are useful (to determine pullbacks/throwbacks) without extensions, whereas extensions cannot be created without a previous retracement.
Practically, use all three of these studies, and a couple more besides. Where a low is followed by a high that seems to produce sturdy resistance, plot a fibonacci retracement. That will give you good idea of what levels between the high and low price is likely to respect and pivot off of. Some retracement levels produce reversals more than others (.618 and .786 come to mind). If price cuts through all of those levels and then violates the low, you know you're probably in for a significant down move; but to where? Well, fibonacci projections will help with that: will it be 127.2%, 161.8%, 261.8% or even 423.2% of that move from the high back to the low? On the other hand, if price goes from low to high, and then retraces 78.6% back to the low and then pivots, take the range from the original low to the high and then map that over the future, beginning with the new low at the .786% retracement level. That will create your extensions.
Long story short: Retracements don't need extension or projections; but extensions and projections need retracements. Over the arc of a retracement/projection, you have a V shape. Projections are a continuation of price retracing. Over the arc of a retracement/extension, you have a zig-zag/lightning bolt. Extensions are the measure of the high-low/low-high mapped over the future beginning with a retracement of that high-low/low-high. Both extensions and projections develop price targets.
I realize that's a lot of words with no charts for reference. If you like, I'm happy to pull some charts and post them here. Otherwise, I hope that gives a bit of clarification.
Last edited by Andrewunknown; 08-01-2009 at 02:25 PM.
Here's a basic example of a fibonacci extension in action.
The green fib study from swing low on 07/13 (point 1) to a swing high on 07/16 (point 2), followed by a 50% retracement on 07/17 (point 3). This retracement could be construed as a 38.2% retracement on a closing basis, or 50% if we look at the intra-candle low.
Next, look at context: note the local importance of the 1.63 level. Price topped here on 07/06 before moving to 1.60 (actually, that makes up points 3 and 4 of measured move down/extension!). Then, note the congestion on 07/13 between 1.6258 and 1.6311. Those points are the 50% and 38.2% retracement of the green fib range. The 0300 ET candle on 07/17 leaves a long lower wick before pulling back up to the 38.2% level. The 0600 ET candle (the ultimate low of the retracement) goes to the 50% but also pulls back up to the 38.2% level. The long lower wicks produced on both candles at those levels (the second candle qualifying as a hammer) is not coincidental.
I plotted my extension from the ultimate low, some 7 pips above the 50% retracement level. As you can see, the black fib study extends to exactly where Friday's high, where price plateaued. Also notice how price topped at the 61.8% extension (the upper dashed line of the black fib study) in the 1.6543 area and bottomed out around the 23.8% extension area at 163.70, producing a loose box channel - even at one point creating a spinning top with a wick that extended back to the 38.2% retracement on the green study.
Long above the green line at 1.6481 with SL at 1.6265 would be the cookie-cutter set-up here, but you would've endured a very significant drawdown before your trade went into profit unless you took profit at the 61.8% extension at 165.42 on 07/20. But that wouldn't be sound money management. Another set-up that I like is going long on a cross above the red dashed line (break of the corrective retracement wave) on the 2100 ET candle on 07/19. That would've brought an entry at 163.95, giving a target at 165.42 and a stop loss at 162.65 a slightly greater than 1:1 ratio. Not great, but a better proposition.
The real win here, obviously, if you could stick out the box channel oscillation (or better yet, trade it back and forth) is long at 1.6395, stop loss at 1.6265 targeting 1.67ish for something like a 2.25:1 ratio. That trade turned into a swing timeframe, but would've worked out very well in the end. Or, you may still be holding into next week for a try at 127.2%, 161.8% etc.
Last edited by Andrewunknown; 08-01-2009 at 05:04 PM.
Or, you may still be holding into next week for a try at 127.2%, 161.8% etc.
Actually took a very similar trade in the scope of the one I outlined above for +131 last night; thought I'd share the results in case anyone is interested in a few of the finer points of fibonacci.
This was a basic 1-2-3 pattern carried over from last week. Picking the low on 07/29 at 1.6338 as Point 1, the swing high early the next morning at 1.6526 and then the low shortly thereafter at 164.56, the original trigger for this trade would've been either
a) a close above the corrective trendline at 1000 ET on 07/30 at 1.6492, which subsequently spun its wheels until, the second and more conservative entry point at
b) on the close above the range of the first fibonacci study in the 0000 ET 07/31 candle at 1.6542.
With neither of those applying by Sunday night, I entered a couple pips above the 161.8% extension of the green study from the previous chart (the top dashed green line @ 1.6735) at 1.6737. Breaking above the congestion at 1.6760 (the 161.8% extension of the second green study on this chart) confirmed the entry. I then time stopped (part of my overall trading methodology) the trade to take profit this morning at 1.6868. If one were to draw an retracement over the congestion from 1200 ET 07/31 to 0000 ET 08/03, they'd find price retraced 78.6%, from which an extension could be drawn that pegs the 161.8% level for each in the 168.50 area. The next area of interest is the 261.8% of these and the latter green study on the above chart, all falling at 1.6937-1.6954.
The trade itself is noteworthy because of the hesitancy I would otherwise have over taking an entry after a parabolic rise like the one that occurred on 07/31.