Ok guys I spent quite some time researching and developing. Think this might really interest you.
It started when I saw BG taking this trade.
Screen Shot 2016-06-09 at 12.04.08.png
It exactly did as it was supposed to; going short since the Diff exceeded our threshold of 2.4. As you can witness, price RRT'd, reversed on us and closed out at loss 15-20 minutes later dealing a fair blow to our account. When I saw this an image of Riemann sums and fundamental principles of velocity and acceleration came spontaneously to mind. Suddenly a lot of pieces came together and I think I acquired a most important insight. A sort of 'once-seen-cannot-be-unseen'-thingy if u will. For those not familiar with Riemann sums I won't go into too much detail and the technicals of it as this would digress us unnecessarily. In short it is a mathematical approximation method - pertaining to the discipline of integral calculus - to calculate the 'area' under a function. This is done by dividing the area in N partitions and subsequently summing them up. Pretty straightforward. The higher N, the more precise the value of the 'area'. Needless to say, in its limit (N to positive infinity) it returns the very exact value of the area - otherwise known as an 'integral'. A visualisation makes it much easier to grasp, doesn't it?  (You'll have to click on it.)
Riemann_integral_regular.gif
"What in the world has this got to do with anything?", you may ask. Let me show you.
Price has wave-like features and properties. We all know this. Turn your everyday candlestick / bar chart into a line chart to see it better. A lot of trading methods are based on these wave-ishness of price action (e.g. Elliott waves). Now let's assume a rather "stable" scenario within which price is trading between well defined S/R levels and is behaving rather well. To do this, let's plot a sine-function and smack on a S/R.
Screen 1.png
There we are. Now let us throw Wilder's DMI into the mix on the M5. To be clear, these are DMI histogram bars.
Screen 2.png
The very schematics in the last picture I manually, amateurishly drew contain the patterns that will occur most often. On first sight it should be immediately clear where the opportunities lie; when price is accelerating and gaining velocity. As you can see, the pattern I already mentioned in previous posts occurred over the first 3 histogram bars. We would've entered on the close of bar 3. and surely hit our TP in the next 5 minutes. The pattern is broken over bar [3,4,5] as Diff_[4,5] is not nearly enough to testify for acceleration. No trade. Diff_[4,5] suggests deceleration and a possibility of a reversal. Think of it as a yellow wavy occurring above that 5th candle. Then price indeed proceeds to fall (as if a blue wavy kicked it in motion) and forms the stairs 5 times with subsequent profit-takings - apart from the last set at point B. This is exactly why we're incorporating a safety-mechanism of no more than 2 - 3 trades in the same direction, for no longer than 10-15-20 minutes a time (we'll have to see what's the better configuration). Immensely important. So when 2 - 3 shorts occur from point A. onwards, we'll forcibly suspend trading and only resume once the stairs-pattern has been broken by a non-stairs pattern. This happens at the top, at point B. Absolutely necessary in order to make this work. Now we're allowing BG to resume its operations - when stairs in either direction get confirmed. I reiterate the technique once more:
Boulder » Mon Jun 06, 2016 11:45 pm wrote:
The process for these secondary trades would go something like this. I define Diff_1 and Diff_2 as the Diffs between respectively the first pair and second pair pertaining to a set of 3 bars.
Go long if:
1. 3 rising histograms
2. Both Diff_1 and Diff_2 between 0.4 and 1.8.
3. Open trade if maximum 1 trade in the same direction has been triggered in the last 10-15 minutes.
4. TP = e.g. 5 pips
5. SL = 35 to avoid disaster
6. If 4. and 5. didn't happen after 15 minutes; exit the trade.
That's it. Analogously for a short.
In essence, we don't want to trade at tops and bottoms. We want to catch and ride the transitions. These transitions contain rather robust DNA. BG cold-bloodedly and unemotionally confirms that DNA and acts accordingly with minimum market exposure, aiming for feasible short & sweet profit targets.
Visually we would be doing this:
giphy.gif
Of course if it were that easy, it wouldn't be any fun now would it? Luckily we have our dearly-beloved market makers around lurking for our stop losses and snatch them whenever they see fit. If you are not aware of these practices, I STRONGLY advise you to read up on CJ's material here. Rather volume-heavy, but I guarantee you won't regret it.
How would we bypass those nasty stop hunts then? Have a look at point C. and D. I noticed these patterns occurring at the top and bottom all the time. Taking care of these will almost surely raise the probability of us avoiding being prey of stop hunts. It appears that the probability of a stop hunt and a subsequent reversal is highest when a 'deformed' stairs-pattern is printed. Diff_[6,7] is much larger relative to the preceding Diff. This increases the chances of, and can be thought of as, market makers reaching out high/low and pulling back. We want to avoid these. The system would go like this:
- 1. If the Diff_2-to-Diff_1 ratio is bigger than X (e.g. 2), then no trade.
2. I feel there's an absolute value under the momentum threshold of 2.4 here too to be avoided, though I'm not sure of this. We would not be taking a trade if the Diff falls between, say, 2 and 2.4. Again, not sure about this one.
Point D. also shows a place where trades should be avoided. Operations resume after the non-stairs pattern of [12,13,14] occurs - leading us to the final part of this system. This integrates the primary/momentum trades which I originally came up with and suggested in the first place. These trades follow after price explodes and leaves a large Diff as represented by point E and Diff_[16,17]. It explodes in a certain direction (in casu upwards) and trades again in a certain range. Here's the place where the entire picture clicked and all sorts of scenario's passed in front of my eyes. Range, explosion, range. Range up/down with a gentle slope with plummets/upsurges. Etc, etc, ... All of them could be covered with a rather high probability. Again, I stipulate, nothing is guaranteed. However it is most definitely worth an attempt. In fact I think these stairs should be the primary method of trading of the EA, taking precedence over momentum trades. Considering that momentum trades typically will follow the stairs, now have a look at the trade I posted up top .
Screen Shot 2016-06-09 at 15.52.00.png
Yes those would be 2 winning trades immediately after each other.
Cheers,
Aram
PS: for those planning on writing a lengthy post - I advise you to save your draft intermittently. I just had to write this entire post from scratch after my internet just bailed on me. 
PSS: the Riemann sums by their visual resemblance to the DMI histograms merely triggered the connection between waves, maths, velocity and acceleration. We won't be calculating integrals or anything of the sort of course  |
