Following is a quick review of a Fibonacci-based methodology I used to employ in trading. Please, consider following the actual trade with its original predictive/forecasting analysis at this link: .
1 - FIRST and foremost, I make it a rule that my zero and 100 Fibonacci levels are confirmed by the existence of a point of repose, which has to be a minimum of 38.2%, as illustrated in the chart:
If a temporary retracement occurs at the 0.114 or the 0.214 levels (i.e.: mathematical complements of the 0.886 and 0.786 minor Fibonacci numbers), then the 100-Fib anchor cannot be defined at the level where these internal retracement, or contraction levels, occur.
2 - Next, I look for significant extensions. Keep in mind that there are several of these levels (i.e.: 1.131, 1.272, 1.313, 1.414 and 1.618).
HOWEVER, in the most common price actions, the 1.618 is the level at which a stop-loss can be sensibly defined.
PEARL: Now, remember that a lot of black-box or even institutional traders might already know this, so watch for an AGGRESSIVE market reversing at the 1.414 instead. Here, we are dealing with a standard level of retracement, namely at the 1.618 level, as illustrated in the following chart:
3 - Next, we are to define a probable level of contraction, or from the current price swing. So, define the standard 0.382, 0.618 and 0.786 as the levels corresponding to vanishing probability levels of retracement, as illustrated in the chart:
Note how price in the LARGER Fibonacci matrix allows its 0.382 contraction level to align with the SMALLER (former) 0.618 level ... one being the mathematical complement of the other (i.e.: 0.382 + 0.618 = 1.000).
Is this making sense to you? Is this perhaps changing the way you might define your trading, stop-losses, or targets? If so, good. If not, feel free to share your thoughts in the thread.
Predictive Analysis & Forecasting
Durango, Colorado - USA
Thank you in advance.
Price is driven by fundamental agents acting upon knowledge of price moving data, such as interest rates, for instance. In a pair, either side have a relative strength. Both can be moving under their own bearish data, and the most bearish will cause the other to rise, for instance. The greatest movement is if the pair moves on correspondingly opposite fundamental data, which is what is typically best to trade. For instance, a BOE decision to increase interest rates against a RBNZ decision to decrease interest rates at the same time would generate an increadible rallying in the GBP vs. NZD Forex pair.
When it comes to trade market geometry pattern, I would remain dubious about the feasibility of a pattern to trump newly released data that speak of the contrary, especially if that data is truly new and had not been released or expected before. In the case of a prescient, or expected data (for instance, the Fed's decision to keep interest rates where they are is of little surprise to most, and that is what probably stomped further rallying of the USD pairs.
Perhaps, your question could be restated not in terms of the Wolfe Wave or the Geo (which I only use as the best illustration that has tended to remain on pace with my independent, but leading method, which is the Predictive/Forecasting Model.
The Predictive/Forecasting Model offers a direction, a strength, as well as identify high-probability future pivot (via quantitative targets, named TG-1, TG-2, ... TG-n, which are forecast retracement levels whereby price would counter-trend in the order that should not exceed 0.618 Fibonacci from prior swing) levels, as well as defines moderate-probability tip-top or bottom-tip reversal levels (via moderate probability qualitative targets named TG-Hi/TG-Lo and low-probability targets named TG-Hix/TG-Lox, for highs/lows and extreme-highs/extreme-lows, respectively).
In the case of the GBPNZD analysis, the basis was of the forecast is purely based on Fibonacci levels. I thought that price was nearing a significant level at 1.618. However, this alone is not sufficient as a technical data to rely upon - and we has seen this, since price continued to move on up.
My decision to enter at the level shown is based on the Predictive/Forecasting Model - Hence, this past June 18th, based on the folowing chart, I decided to take a short - See chart, which is a cut/paste of the price action as of 18 JUN 2015:
$GBPNZD - Weekly chart on 18 JUN 2015:
The difference between the Fibonacci levels, the advanced market geometries such as the Wolfe Wave, and a quantitative-based model such as the Predictive/Forecasting Model (which does not use price, unlike the other two methods mentioned), is that each carry their own probability values.
In my view, a Fibonacci-based trading represents the lowest affordable gain in terms of probability. For instance, shorting a Forex pair once price reaches a specific extension level is simply not good enough a method, especially since many of the software, black boxes and institutionally-trained, Pavlovian traders already know and rest orders at these levels, producing a less efficient and too predictable pattern, which exposes their positions to vying institutional traders.
In contrast, pattern-based trading offers a higher level of probability. I have written and discussed probability strength on various occasions. Although I do not have good data, some sites have a range of success rate data on Scott Carney's patterns which appears consistent with my trading (been trading unsatisfactorily with most of these standard patterns, and since moved on, discovering my own patterns - Great White, Janus, Euclid, Alternate Shark - and developing my own quantitative model since then. I can say that theShark/5-0, followed by The Crab, and then the Bat are the best, most successful pattern in this very order of vanishing probability.
Last, but not least, in terms of highest probability is whatever you have been able to come up with. In my case, I have relied on the "Predictive/Forecasting Model", which is simply a combined set of quantitative and qualitative data that is derived from a non-price field - I won't tell you what it is, but this is what I use as a "foreground" analysis. In contrast, the geometries are simply a "background" visual support, and for this, I have now tended to use the Wolfe Wave, and elaborated on it by developing rules for the "Geo". But these are simply moderate probability tools relative to what I consider a high-probability Predictive/Forecasting Model.
OVERALL, we are only dealing with probabilities. I am not the one forecasting, and I certainly do not have any directional "feelings" about a market. I simply look into any price moving asset (stock, Forex pair, index, commodity) through the "lenses" of the Predictive/Forecasting Model, and I write down and share what "it says". Since we are dealing with probabilities, I simply layer all that is probable in decreasing order, and share the chart based on:
1 - Predictive/Forecasting Model (high-probability)
2 - Advanced market geometries: Wolfe Wave or Geo (moderate probability)
3 - Fibonacci levels (low probability)
All these layered probability expressions are what is defined in the chart - I often remove too many of these indicators, so as to keep the chart clean and intelligible to the mind's eyes, but this is how I arrive at the levels defined. At the end, it either goes there, or it does not. It either reverses there or carries on like a steam roller.
Once a "steam roller" scenario occurs, then I simply look at the 4-fold timeframe above present timeframe (e.g.: 15-min x 4 = 1-hour; 1-hr x 4 = 4-hour timeframe, 4-hr x 4 ~ daily timeframe, and daily x 4 approximates a weekly timeframe, and weekly x 4 = monthly).
I hope this elucidates a bit my approach, and explains why a forecast is only a matter of probability that is based only on a layered set of technical data, and not necessarily trying to combined the fundamental data, since by nature, technical analysis exists on its own presumption of discounting that data into its price field.
If and once price errs further out, then I would look for a larger timeframe in which to re-apply the Model. At 2.30275, the Model suggests a short entry. However, the Model is not aware nor does it depend on the underlying development of a geometry, be it a Wolfe Wave, a Geo, or any of the Scott Carney's patterns.
Hence, at that level of entry (which would have appeared as a moderate-probability target (i.e.: as a TG-Hi level), price would be expected to decline significantly.
What this also signal to me is that whatever pre-existing geometry that I may have considered, should now be superseded by a larger geometry: Looking at the chart, it would be correct to use the Model's short entry as a price level corresponding to Point-3, and expect that a Geo might work in this instance, simply by looking back at the probable 1-2 Leg and use its inherent reciprocal symmetry expressed as a simple ab =cd pattern.
Since the 1-3 Line is required to be rising when Point-3 is atop the geometry, then a reasonable construction would start with he following premise, where the ab =cd construction of the 1-2 Leg starts the larger Geo (here, the assumption is to call it a Geo and not as Wolfe Wave, since it is making use of internal geometric conditions which do not exist in the WW):
Then, you would simply build the Geo from here, such that the 1-3 Line and the 1-2 Line remain apparent, but the 1-4 Line remains undefined, since we have not yet seen the final level of Point-4, although one can approximate it using a fainter line projecting from Point-2 forward, and follow the structural lows ahead, such that you would end up with this speculative geometry:
At this point, it is assumed already that the smaller geometry has lost its relevance, not simply because of the extreme level attained from the 5-second (here, each trader would have to define their own level of tolerance, either as a fixed amount of acceptable loss, or better yet, as a percentage value of a pre-defined risk management plan.
OVERALL, there are several levels shown in the chart: One where the Predictive/Model defines probable levels of attainment, retracement and reversals (which is shown on a separate engine and is never shown on the charts I share, and the other where I ascribe a probable geometry, for which I have decided that the Wolfe Wave or Geo are the best representative agents, as they tend to correlate well once their geometry is forming.
To your question: "a Predictive / Forecasting model works or not, always?", the answer is simply that it does not always work, but it represents the highest probability indicator in terms of market direction, strength, future pivot definition levels and reversal levels.
I also believe that you are assuming that both the Model and the geometric points have something in common. They simply don't. Therefore, as you ask: "what tolerability shows the model after point 5" ?", I want to emphasize that the Model is not cognizant of the geometries and is therefore not designed to work with the geometries. It's simply that the geometries used (WW and Geo) have been the most responsive to the Model's forecast levels of probable price retracement and reversals - I hope this is clear.
The question perhaps might be: What tolerability shows the model after TG-1, if the target were a quantitative one, or TG-Hi in the case of a qualitative one. The answer would simply rely upon whatever the trader defined as her risk tolerance, which should be defined by a pre-determine percentage of her asset (say, trade only 10% of the total assets, and only risk 2% of that exposure per trade, for instance).
I am expanding on topics that cross-reference your question on purpose, and go a bit off tangent, but I want to be sure that I shed light on that important question, as well as preempt the question of other related questions.
I appreciate you reading thus far.
Have a fantastic weekend.