Statistics based system; Technical Documentation | Forex Factory

Options

Permalinks
Print Thread

Similar Threads

Grid system based on optimization and statistics 1 reply

Documentation prior to each trade and Time management 1 reply

FXpro new Amendments to the Legal Documentation 5 replies

Trader Statistics and Broker Statistics 0 replies

Help! MetaEditor Documentation font size shrunk! 2 replies

Attachments: Statistics based system; Technical Documentation
Exit Attachments

Statistics based system; Technical Documentation

  • #1
  • First Post: Apr 1, 2022 7:48am Apr 1, 2022 7:48am
Just starting a journal here to document my system.
There are too many pieces of it, but I would do my best to break them down here.
It is also to help me see if there are better ways to do the same thing.

(reserved this space to create menu with bullet points here)
  • #2
  • Edited 9:38am Apr 1, 2022 8:04am | Edited 9:38am
Intraday 123 pattern retracement zone

Comparing with this post from simnz 's thread
https://www.forexfactory.com/thread/...8#post13944218

I personally think there are better ratios we can use for each pullback/retracement for intraday swings.
I had found that this differs in the following;
1-per pair
2-timing of the day (broken down into timings based on 96 x M15)

In the past, I had tried to use certain fixed ratios, and found that while it helps, eventually I need something much more specific and tuned for each pair.
Since every instrument / pair has different volatility depending on time of the day, hence there has to be a way to have the optimal stop loss (2 standard deviation time and price based) for the given time of the day.

The scatterplots are based on....
1-Past 2 years days worth of intraday swings, which are only in the same direction with the daily / higher timeframe swing.
2-Only 123 pattern intraday swings qualify, ignore all other patterns (there is another pattern which will be explained later)
3-Take only the 2nd swing retracement candle length and price length
4-Normalize price by using rolling 5 days of volatility.
5-The end time of 1st swing determines the indexing of timings based on 96 x M15. e.g. 0 = 0000 hrs , 1 = 0015 hrs , 2 = 0030 hrs
6-Store this data into a CSV and load it for live use

How to use or apply;
1-Apply the filters based on the current 1st swing's timing index
2-Using the scatterplot, try and catch the best possible bottom of 2nd swing for entry
3-Set price stop loss just outside the 2nd standard deviation price, and consider early exit timing wise if 2 standard deviation time hit.
4-Ride the 3rd swing for breakeven and profit
5-Repeat until higher timeframe exhaust or hit stop loss
Attached Image(s) (click to enlarge)
Click to Enlarge Name: screenshot.png Size: 73 KB
Click to Enlarge Name: screenshot.png Size: 60 KB
Click to Enlarge Name: screenshot.png Size: 77 KB
Click to Enlarge Name: screenshot.png Size: 65 KB
Click to Enlarge Name: screenshot.png Size: 389 KB
Click to Enlarge Name: screenshot.png Size: 162 KB
  • #3
  • Apr 1, 2022 8:44am Apr 1, 2022 8:44am
Quoting Relativity
Comparing with this post from simnz 's thread https://www.forexfactory.com/thread/...8#post13944218 I personally think there are better ratios we can use for each pullback/retracement for intraday swings. I had found that this differs in the following; 1-per pair 2-timing of the day (broken down into timings based on 96 x M15) In the past, I had tried to use certain fixed ratios, and found that while it helps, eventually I need something much more specific and tuned for each pair. Since every instrument / pair has different volatility...
Thank you for your initiative to educate me scatter plots. I have never used them. Let me go through your posts and I will respond.
Practice makes a person perfect
  • #4
  • Edited 11:56pm Apr 1, 2022 11:30pm | Edited 11:56pm
Intraday parabolic pattern, Figure 2.31 retracement zone

Taking inspiration from Ray Barros.
What if 123 pattern does not happen? Assuming one is still holding the trade because....
1-The second swing triggered entry
2-The third swing did not materialize into a fresh break to complete the 123 pattern
3-Should we hold the trade...
a-until it fresh break later
b-with same stop loss price wise
4-Given new developments, should we...
a-adjust price stop loss tighten
b-adjust price stop loss loosen
c-adjust time stop loss widen

The process of gather statistics;
1-With the earlier Intraday 123 pattern retracement zone, find those that held price stop loss, but did not hit 3rd swing fresh break
2-Add filter that had fresh break much later, regardless of swing count
3-Add filter that price stop loss from Intraday 123 pattern retracement zone still holds, regardless of swing count
4-Gather all the price and time data points for each occurrence, ensure to use normalized price
5-Note that each occurrence has different swing counts, don't force a specific number. Just gather the data.
6-The end time of 1st swing determines the indexing of timings based on 96 x M15. e.g. 0 = 0000 hrs , 1 = 0015 hrs , 2 = 0030 hrs
7-Store this data into a CSV and load it for live use

How to use or apply;
1-Apply the filters based on the current 1st swing's timing index
2-If Intraday 123 pattern retracement zone holds price stop loss, ok to display F231 retracement zone
3-F231 retracement zone has its own 2 standard deviation price stop loss, and 2 standard deviation time stop loss
4-Keep price stop loss from Intraday 123 pattern retracement zone, but be prepared to adjust if F231 zone shows tighter price stop loss
5-Ride the Nth swing for breakeven and profit
6-Repeat (using both types retracement zones as and when) of until higher timeframe exhaust or hit stop loss
Attached Image(s) (click to enlarge)
Click to Enlarge Name: screenshot.png Size: 282 KB
Click to Enlarge Name: screenshot.png Size: 111 KB
  • #5
  • Apr 1, 2022 11:55pm Apr 1, 2022 11:55pm
Quoting Relativity
Intraday parabolic pattern, Figure 2.31 retracement zone Taking inspiration from Ray Barros. What if 123 pattern does not happen? Assuming one is still holding the trade because.... 1-The second swing triggered entry 2-The third swing did not materialize into a fresh break to complete the 123 pattern 3-Should we hold the trade... a-until it fresh break later b-with same stop loss price wise 4-Given new developments, should we... a-adjust price stop loss tighten b-adjust price stop loss loosen c-adjust time stop loss widen The process of gather statistics;...
Great. This is what I was looking for. Thank you.
Practice makes a person perfect
  • #6
  • Apr 3, 2022 6:06am Apr 3, 2022 6:06am
https://www.forexfactory.com/thread/...5#post13955235
Could this study be of any use to your system?
Practice makes a person perfect
  • #7
  • Apr 3, 2022 10:37am Apr 3, 2022 10:37am
Quoting simnz
https://www.forexfactory.com/thread/...5#post13955235 Could this study be of any use to your system?
Nope as my swings are really customized per pair as well, and I am aware of this book. Thanks for the offer thou.
  • #8
  • Apr 6, 2022 5:46am Apr 6, 2022 5:46am
If you can find a pattern of hugging or hippo-jumping Consolidation measured by time, major news event, Thursday and Fridays (weekly options expiration days), quarterly triple/quadruple wrenching dates, we can derive useful insights.

Presently, as a thumb rule, I use Murrey Math lines based on 64 factor as a daily guide to find the lines where price keeps hugging or revolving. It is similar to what you call going easy on the mind. The indicator on top left shows the H-L range for previous 4 candles plus the current candle moving dynamically.
Attached Image (click to enlarge)
Click to Enlarge Name: 2022-04-06_18h38_35.png Size: 72 KB
Practice makes a person perfect
  • #9
  • Apr 6, 2022 3:57pm Apr 6, 2022 3:57pm
Pre-FOMC Minutes release, the consolidation lasted for 6.5 hours prior to the release time.
The Range was 50 HL pips for the consolidation during which buys and sells were recorded .
The consolidation period prior to major data release lasts from 16 hours to 6 hours.
Attached Image(s) (click to enlarge)
Click to Enlarge Name: pre fomc minutes release bought positions.jpg Size: 1.3 MB
Click to Enlarge Name: Pre Fomc minutes sold positions.jpg Size: 1.7 MB
Practice makes a person perfect
  • #10
  • Apr 10, 2022 8:45am Apr 10, 2022 8:45am
Quoting simnz
If you can find a pattern of hugging or hippo-jumping Consolidation measured by time, major news event, Thursday and Fridays (weekly options expiration days), quarterly triple/quadruple wrenching dates, we can derive useful insights. Presently, as a thumb rule, I use Murrey Math lines based on 64 factor as a daily guide to find the lines where price keeps hugging or revolving. It is similar to what you call going easy on the mind. The indicator on top left shows the H-L range for previous 4 candles plus the current candle moving dynamically. {image}...
Quoting simnz
Pre-FOMC Minutes release, the consolidation lasted for 6.5 hours prior to the release time. The Range was 50 HL pips for the consolidation during which buys and sells were recorded . The consolidation period prior to major data release lasts from 16 hours to 6 hours. {image} {image}
I understand where you are coming from.
However I do not rely on Murray Math lines, as I believe there should be a direct or better way.
I am currently experimenting my current swing work to see how I can extract the data, without breaking MT4
  • #11
  • Apr 11, 2022 12:12am Apr 11, 2022 12:12am
Quoting Relativity
{quote} {quote} I understand where you are coming from. However I do not rely on Murray Math lines, as I believe there should be a direct or better way. I am currently experimenting my current swing work to see how I can extract the data, without breaking MT4
I just shared my experience of working with retirees for your reference. Now waiting for a tool from you which can generate data that can be made use of.
Practice makes a person perfect
  • #12
  • Apr 11, 2022 5:39am Apr 11, 2022 5:39am
Quoting simnz
{quote} I just shared my experience of working with retirees for your reference. Now waiting for a tool from you which can generate data that can be made use of.
Apologies but I dont create tools for anyone, nor share it out. Just to make it clear.

What I typically share are steps one may use to recreate what I had done. There are too many programming languages / trading platforms out there.
  • #13
  • Apr 11, 2022 10:23am Apr 11, 2022 10:23am
Quoting Relativity
{quote} Apologies but I dont create tools for anyone, nor share it out. Just to make it clear. What I typically share are steps one may use to recreate what I had done. There are too many programming languages / trading platforms out there.
That is the impression I got from your earlier responses.
That is my error.
Anyway, whatever steps you want to share,please do .
Practice makes a person perfect
  • #14
  • May 10, 2022 2:16pm May 10, 2022 2:16pm
Following
It's not over, until I win.
  • #15
  • Edited 7:09am Sep 18, 2022 6:22am | Edited 7:09am
Major breakthrough using combination of MQL4, Powershell scripts and Forexfactory calendar data. I had completed a news engine that measures and provides impact (as a number), based on news event, and historical pre and post Consolidation Price Action for that news event, and for per pair.

RIP Queen Elizabeth
Attached Image(s) (click to enlarge)
Click to Enlarge Name: IMG-20220918-WA0002.jpg Size: 570 KB
Click to Enlarge Name: IMG-20220918-WA0007.jpg Size: 299 KB
Click to Enlarge Name: IMG-20220918-WA0006.jpg Size: 314 KB
Click to Enlarge Name: IMG-20220918-WA0005.jpg Size: 367 KB
  • #16
  • Edited 6:15pm Jan 26, 2023 5:07pm | Edited 6:15pm
Code Freeze for my work other than minor bug fixes.

1a-News review, customized impact scoring based on historical data as per news event.
Attached Image (click to enlarge)
Click to Enlarge Name: screenshot.png Size: 49 KB

1b-News last rolling 24 hours relative to price action, and what's left for the week.
Attached Image (click to enlarge)
Click to Enlarge Name: screenshot.png Size: 40 KB


2-How often price stays or moves away or comes back to the open price of the day. Historical data based on time (96 x M15) of the day volatility and day of the week (mon/tues/wed/thur/fri). Projected forward for each day.
Attached Image (click to enlarge)
Click to Enlarge Name: screenshot.png Size: 40 KB

2b-Keeping only the bounds for both price and time. It almost serves like a statistical backed Asian session breakout range
Attached Image (click to enlarge)
Click to Enlarge Name: screenshot.png Size: 35 KB


3a-Projected Close based on time of current high and current low of the day.
Attached Image (click to enlarge)
Click to Enlarge Name: screenshot.png Size: 20 KB

3b-Based on historical data of High to Close distance + Low to Close distance for final major move of the day if down, and Low to Close distance + High to Close distance if up.
Attached Image (click to enlarge)
Click to Enlarge Name: screenshot.png Size: 21 KB
Attached Image (click to enlarge)
Click to Enlarge Name: screenshot.png Size: 19 KB

3c-Highs and Lows are also timing sensitive. The earlier the high and lows are, the more chance it has to reverse to the other side, to account for days when price just range expand progressively through the day.
Attached Image (click to enlarge)
Click to Enlarge Name: screenshot.png Size: 30 KB

Attached Image (click to enlarge)
Click to Enlarge Name: screenshot.png Size: 13 KB

Attached Image (click to enlarge)
Click to Enlarge Name: screenshot.png Size: 12 KB
  • #17
  • Edited 6:51pm Jan 26, 2023 6:25pm | Edited 6:51pm
4a-Major moves that are smaller than 2 standard deviation are grayed. Point of control line based on previous major move, where price tends to return to for a ranging day
Attached Image (click to enlarge)
Click to Enlarge Name: screenshot.png Size: 11 KB

4b-Historical distance from point of control when price is returns to point of control and makes a new opposite high/low, and projected forward per day.
Attached Image (click to enlarge)
Click to Enlarge Name: screenshot.png Size: 28 KB

4c-Historical distance from point of control when price returns or does not to point of control, plus does not makes a new opposite high/low, and projected forward per day.
Attached Image (click to enlarge)
Click to Enlarge Name: screenshot.png Size: 20 KB

4d-Combined plots, and keeping only the average and 2 standard deviation levels
Attached Image (click to enlarge)
Click to Enlarge Name: screenshot.png Size: 78 KB
Attached Image (click to enlarge)
Click to Enlarge Name: screenshot.png Size: 56 KB

4e-On trending days, price trends to stick close to the 2 standard deviation levels, and bounce off / retracement at specific times of the day.

Attached Image (click to enlarge)
Click to Enlarge Name: screenshot.png Size: 49 KB

5a-Trending days have intraday​ 123 patterns, which the price and time stop loss can be placed behind 2 standard deviation of historical retracements. These retracements are time of the day sensitive, and are plotted with price and time bell curves.
Some timings are not worth it to use 123 pattern e.g. much later time of the day. This justifies not keeping a position from current day to next day most of the time, and close off before end of the day.
Attached Image (click to enlarge)
Click to Enlarge Name: screenshot.png Size: 22 KB

5b-Intraday 123 patterns, projecting forward when and how often fresh breaks trend to occur.
With target project time and price points, boxed into 4 sections; above or below average time and price wise, until 2 standard deviation time and price wise.
Historical data of 123 patterns used for these projections.
Attached Image (click to enlarge)
Click to Enlarge Name: screenshot.png Size: 27 KB
Attached Image (click to enlarge)
Click to Enlarge Name: screenshot.png Size: 11 KB
  • #18
  • Jan 26, 2023 6:55pm Jan 26, 2023 6:55pm
6-Putting it all together, where everything comes together with confluence.
Historical price behavior combined together to create a picture of how the day may turn out, like a intraday template showing the important movements to look for.
These data is self adapting and as per symbol; each pair has its specific behaviors so that trader can trade based off it.

Attached Image (click to enlarge)
Click to Enlarge Name: screenshot.png Size: 83 KB
Attached Image (click to enlarge)
Click to Enlarge Name: screenshot.png Size: 77 KB
Attached Image (click to enlarge)
Click to Enlarge Name: screenshot.png Size: 93 KB
Attached Image (click to enlarge)
Click to Enlarge Name: screenshot.png Size: 75 KB

Attached Image (click to enlarge)
Click to Enlarge Name: screenshot.png Size: 73 KB
  • #19
  • Last Post: Feb 14, 2023 5:56pm Feb 14, 2023 5:56pm
7-Adding on a minor enhancement. M15 bars High-Low average size and Volume average sizes, showing the odds of when these sizes are above average.
Filtered by weekday and M15 time of day.

Attached Image (click to enlarge)
Click to Enlarge Name: screenshot.png Size: 46 KB
Attached Image (click to enlarge)
Click to Enlarge Name: screenshot.png Size: 23 KB
Attached Image (click to enlarge)
Click to Enlarge Name: screenshot.png Size: 21 KB
0 traders viewing now