Rel's Lab | Page 2 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Rel's Philosophy of Trading Discussion 24 replies

Rel's Forge 53 replies

Rel's Evaluation 13 replies

Nikola's Trading Lab 247 replies

Spud's Lab Accident 22 replies

Attachments: Rel's Lab
Exit Attachments

Rel's Lab

  • #21
  • Aug 15, 2014 7:13am Aug 15, 2014 7:13am
Quoting LiquidGenius
Some thoughts before I nap
The more I think about it the more I feel like correctly developing states is just as hard as correctly developing a good tool to map price.
For example your wave tool is created in such a way that each bar or wave, in order to register, has to be "significant"; a term which you ultimate had to define yourself.
Yes it is just as hard, but at least knowing the tool has mapped out price movement correctly, you can settle your mind better at the next step of developing states.
To clarify;
1-The bars are created using a "significant" value, to determine the best range bar value to use.
2-The waves tool is generic at 1 period and is applied to the range bar chart(treat this as 1st level of price frame)
3-Wave lengths are then statistically extracted to determine what is the significant period to use
4-The wave tool is reapplied again using the significant period (treat this as 2nd level of price frame)
5-Repeat 3 and 4 to determine the next price frame level. Personally I stop at 3rd level.

Quoting LiquidGenius
I think a "correct" wave tool does the following:
-correctly maps price (not too much distortion of missing highs/lows)
-follows mechanical rules (RB/modded BS swings)
-offers the user a way to actually trade from them (ratios)
-offers a pattern (freq. dist./awareness of: number of continuation moves, trouble areas, make or break areas, areas of no return, turning points, etc).
Yes for all, at least based off my experiences, especially the last point is most important.

Quoting LiquidGenius
At a glance states run into the same issue of "tailored to fit". I think it's impossible to say that one persons classification of states is better than another's, although their states should match up in some scenarios (strong trends).
Yes, strong trends should not be denied. Other states IMO I think it can differ, but my understanding and persistance of pushing the limits of accuracy come from this fabled post:
http://www.forexfactory.com/showthre...73#post5911673
Quoting iDouble
Market Type 1: Vertical (Bullish, or Bearish)
Market Type 2: Horizontal (Channeling)
Market Type 3: Transitional (Unstable)
Everyone is familiar with a Type I and Type II market, and most Traders believe they have a strategy, tactic or a system for dealing with both types. However, the dominant market type is Type III. It is in effect a more higher percentage of the time than either Type I, or Type II. Thus, is behooves the Trader to understand the market that dominates every chart.
A Type III market condition is one where price itself is going through a shift in phase and momentum (either From or To) a Type I, or a Type II market. It is the market behavior just before the Horizontal channel is formed, and just before either the Bullish Vertical or Bearish Vertical is solidified.
Without a doubt, Type III is the most illusive market type to nail down with a degree of specificity, and it happens to be the market type where most Traders lose the vast majority of their capital. Only the Bermuda Triangle holds more mystery and more lost vessels.
Everyone is searching for the so-called "holy grail" of trading, yet precious few even know where the problem is located. The problem for the discovery of the "holy grail" is located in the Traders Bermuda Triangle, otherwise known as: The Type III Market.
Find a compass that enables you to circumnavigate the Traders Bermuda Triangle, and you will have found your "holy grail." More precisely - if you can know the Start and End points of the Type III Market, then you can know the Beginning and Ending of every "trend," and the Beginning and Ending of every "channel."
Impossible to know, you say? No. Not impossible. But, it is not simply sitting there on the superficial surface, just waiting for you to discovery its existence either.
The Transition Phase is the most important phase of Price Behavior in existence - bar none. Develop an indicator, or a set of indicators that helps you map that very specific price behavior, and you will discover a whole new realm of trading has just opened up to you, that you never knew existed before.
Hogwash, you say?
When you start developing a sequence of indicators that help you identify and measure the Transition Phase (Type III), then you will also discover a world of trading where the Hogs actually wash themselves and the horses run forever.
Thats also why I insisted on market states in terms of swings to overlap, so that its much clearer to see how the transition occurs, which I will explain more later. Classical Elliot does not have such a feature, and I have yet to see any WA out there with this, with the minor exception of GWT (trend -> stronger trend; small 1-2-3 -> big 1-2-3).

Quoting LiquidGenius
Question: What qualities are in a "correct" state? Likely something that answers "yes" to your two questions haha
I'm thinking this is something that one learns from actually creating states and working through them (aka the need from going from 3 states to 6? lol) but I feel like there are a couple of different sets one could use that would in theory work:
1. trend/expand/squeeze (with or without it's extensions)
2. moving/flat (exhale/inhale)
3. strong bullish/bullish->flat/ flat/bearish->flat/strong bear
etc and of course even if we were...
Thats where the Dow Theory and Ray's work come in. I would explain the determination of states soon. But yes, these 3 you listed seem to show up most of the time. The variations are another story.

Quoting LiquidGenius
Lastly perhaps BB isn't the way but I feel like you have to use averages in some fashion. After all, since price is in a state, it must be in a state compared to something else right? How else would you be able to tell the difference between an expand or flat, if it weren't for knowledge of previous sizes of said states? {image}
Yes there has to be a 'reference' to compare, a qualifier. I was never comfortable with averages, but I have not given up as I still have ideas to test e.g. GMMA. Even MasterForex-V uses a set of averages and measures how far or close the averages are, as a means to determine volatility. After all volatility is related to market states as you can see.
  • #22
  • Aug 15, 2014 7:28am Aug 15, 2014 7:28am
Quoting PipMeUp
This thread is very interesting. But I'd like to drop this little post to warn you about the fact that the Bollinger Bands is a defective tool. It leads you to wrong interpretations. The first issue with the BB is that it uses as many samples to estimate the variance as it uses to compute its moving average. For instance a BB with a SMA(30) will use the last 30 samples to estimate the variance (the square of the width of the bands). This is absolutely wrong! The value you get this way for the variance comes with an error of up to 300% (yes really!)....
Thank you for your comments. It is really an eye opener to see someone else to be able see that classical tools we have for trading are flawed! And take action to determine what is right. Welcome to the club =)

Your explanation is very clear. This made me more doubtful about using bands. I think instead of using Std Devs bands, there should be other ways to capture PA using averages. GMMA is the closest I can think of, but I would also want to keep certain things simple.

My problem on the coding level with averages is that one has to track every candle against the current MA line, in order to gather relevant statistical data. It is a huge task; I see that it can be done, but I see that I would rather avoid this problem instead by using my wave swing points instead. e.g. what swing points are in confluence with which MAs?

But this means my swing points must be top notch to replace candle data. So far I think I am on the right track, but this would be outside the current discussion we are having now.

http://i59.tinypic.com/2wnmefa.png
  • #23
  • Aug 15, 2014 8:56am Aug 15, 2014 8:56am
-Market Conditions (part 2)

Dow Theory says an uptrend moves with higher highs and higher lows, which leads to;
-2nd swing low does not exceed 1st swing low
-3rd swing high exceed 1st swing high
which produces Trend state

Therefore logically, a failure to uptrend means either;
-2nd swing low does not exceed 1st swing low
-3rd swing high does not exceed 1st swing high
which produces Squeeze state

OR
-2nd swing low exceed 1st swing low
-3rd swing high exceed 1st swing high
which produces Expand state

Opps so there is a 4th state? Where is it? Here we are :
-2nd swing low exceed 1st swing low
-3rd swing high does not exceed 1st swing high
which produces Resume/Reverse state

http://i58.tinypic.com/6th46f.png

There are no other states logically speaking other than these 4 (at least according to my WA). This simplifies trading down so much.

The 4th state looks very familiar isn't it? The reason why I stated the 1st three often but do not mention the 4th... is because this 4th state, I found that it is the most consistent and profitable pattern, but also the hardest to pinpoint down. I find that this pattern not much studied when I look at all the WA work available. This itself deserves it own topic.
  • #24
  • Aug 15, 2014 9:04am Aug 15, 2014 9:04am
state #1 is made of higher highs and higher lows. The opposite (down trend) is then made of lower highs and lower lows. But the state #4 is also made of lower highs and lower lows. How do you distinguish the state #4 and the bearish version of the #1? The amplitude?
No greed. No fear. Just maths.
  • #25
  • Aug 15, 2014 9:11am Aug 15, 2014 9:11am
Quoting PipMeUp
state #1 is made of higher highs and higher lows. The opposite (down trend) is then made of lower highs and lower lows. But the state #4 is also made of lower highs and lower lows. How do you distinguish the state #4 and the bearish version of the #1? The amplitude?
I would just flip the 4 states the opposite way for downside. I do understand your point that this would seemingly lead to 4 x 2 states for both up and down, but I treat these states as directionless (yes, even though they are somewhat indicated by direction)

It would be better explained by my next post, to show how swings overlap.
  • #26
  • Aug 15, 2014 9:14am Aug 15, 2014 9:14am
Quoting Relativity
-Market Conditions (part 2) Dow Theory says an uptrend moves with higher highs and higher lows, which leads to; -2nd swing low does not exceed 1st swing low -3rd swing high exceed 1st swing high which produces Trend state Therefore logically, a failure to uptrend means either; -2nd swing low does not exceed 1st swing low -3rd swing high does not exceed 1st swing high which produces Squeeze state OR -2nd swing low exceed 1st swing low -3rd swing high exceed 1st swing high which produces Expand state Opps so there is a 4th state? Where is it? Here...
Your patterns account for A + B + C swings have you considered the D + E swing in relation to your data? Or 5 wave pattern?
  • #27
  • Aug 15, 2014 9:25am Aug 15, 2014 9:25am
Quoting Baillie
{quote} Your patterns account for A + B + C swings have you considered the D + E swing in relation to your data? Or 5 wave pattern?
Good question. I've decided to analyze 3 swing types 1st as these make up majority when I go by swing count; 60 to 70% of time. I may attempt at 5 swing types as you can see its around 20% to 25% in future. You had just added something to my to do list =)
http://i59.tinypic.com/1hbion.png

Just to be sure on both sides of the market, I separated downtrend and uptrend. Not much difference as I see it
  • #28
  • Aug 15, 2014 9:27am Aug 15, 2014 9:27am
Do you distinguish between these 2 sub-cases of the state #4 --depending on the ranges of the swings 1 and 3 overlap or not-- ?
Attached Image
No greed. No fear. Just maths.
  • #29
  • Aug 15, 2014 9:28am Aug 15, 2014 9:28am
Quoting Relativity
{quote} Good question. I've decided to analyze 3 swing types 1st as these make up majority when I go by swing count; 60 to 70% of time. I may attempt at 5 swing types as you can see its around 20% to 25% in future. You had just added something to my to do list =) {image} Just to be sure on both sides of the market, I separated downtrend and uptrend. Not much difference as I see it
HAHA, that was my next question Thanks
  • #30
  • Aug 15, 2014 9:34am Aug 15, 2014 9:34am
Quoting PipMeUp
Do you distinguish between these 2 sub-cases of the state #4 --depending on the ranges of the swings 1 and 3 overlap or not-- ? {image}
This is a sub state which I will chose not to disclose because it answers this burning question and is a huge money maker :

Does support become resistance when broken and vice versa?

Sorry! ;d

I love your questions... haha...
  • #31
  • Aug 15, 2014 10:02am Aug 15, 2014 10:02am
Quoting Relativity
{quote} This is a sub state which I will chose not to disclose because it answers this burning question and is a huge money maker : Does support become resistance when broken and vice versa? Sorry! ;d I love your questions... haha...
In a way, not answering is an answer in itself eh...? hehehe
  • #32
  • Aug 15, 2014 10:07am Aug 15, 2014 10:07am
Quoting LITEchild
{quote} In a way, not answering is an answer in itself eh...? hehehe
I don't want to provide the statistics to prove the success rate and optimal entry levels, thats all
  • #33
  • Aug 15, 2014 10:20am Aug 15, 2014 10:20am
Quoting LITEchild
{quote} In a way, not answering is an answer in itself eh...? hehehe
I asked because I distinguish 3 cases: inner swing, outer swing and equal swing. There are then 9 possible sequences of 3 swings. Relativity only considers 4. But I never thought before about taking the relationship of the 1st and 3rd swings end points into account. It brings the possibilities to 11 configurations. My personal interest was in the probability distribution of the length of the 4th swing given the 3 current ones. I'm curious to know if there is a statistically significant departure from the Gamma distribution of the length of one swing taken alone.

Another interesting question is of course to see if it is possible to forecast a trend change given the (estimation of the) current trend and the sequence of the last N swings.
No greed. No fear. Just maths.
  • #34
  • Aug 15, 2014 10:24am Aug 15, 2014 10:24am
Quoting Relativity
-Market Conditions (part 2) Dow Theory says an uptrend moves with higher highs and higher lows, which leads to; -2nd swing low does not exceed 1st swing low -3rd swing high exceed 1st swing high which produces Trend state Therefore logically, a failure to uptrend means either; -2nd swing low does not exceed 1st swing low -3rd swing high does not exceed 1st swing high which produces Squeeze state OR -2nd swing low exceed 1st swing low -3rd swing high exceed 1st swing high which produces Expand state Opps so there is a 4th state? Where is it? Here...
So let me get this straight...basically, the major difference in your methodology is in the algo you use in constructing your 'waves', which as you mentioned earlier is a modified form of Barros Swings?
  • #35
  • Aug 15, 2014 10:28am Aug 15, 2014 10:28am
Quoting Relativity
{quote} I don't want to provide the statistics to prove the success rate and optimal entry levels, thats all
Once the 'structure' is understood, I think any serious trader can map out his or her own simple probability distribution curves with respect to any given permutation.
  • #36
  • Aug 15, 2014 10:29am Aug 15, 2014 10:29am
Quoting PipMeUp
{quote} I asked because I distinguish 3 cases: inner swing, outer swing and equal swing. There are then 9 possible sequences of 3 swings. Relativity only considers 4. But I never thought before about taking the relationship of the 1st and 3rd swings end points into account. It brings the possibilities to 11 configurations. My personal interest was in the probability distribution of the length of the 4th swing given the 3 current ones. I'm curious to know if there is a statistically significant departure from the Gamma distribution of the length...
Yes I see where you are going. My older work ended up with... 16 configurations. I decided to work back and simplify, then let the statistics do the rest.

Yes it possible to forecast a trend change given the (estimation of the) current trend and the sequence of the last N swings. I answered this question partially in iDoubleStoch's thread here ;
http://www.forexfactory.com/showthre...48#post7643548
http://www.forexfactory.com/showthre...20#post7646220
  • #37
  • Edited Aug 16, 2014 9:24am Aug 15, 2014 10:31am | Edited Aug 16, 2014 9:24am
-Trading the Resume/Reverse state aka Funny Looking Rectangles & Useless Drivel

To use 3 swing based states is to attempt to anticipate the 3rd swing's length and trade it, in relative to the 1st and 2nd swings. We now speak in context of the Resume/Reverse state.

You need to know 2 things in logical order, to trade this pattern successfully.
1-Where is the 2nd swing going to end
You must answer this first to eliminate risk and determine where your stop loss should be placed.
http://i60.tinypic.com/2gy4tmv.png

Logically speaking, this means looking backwards at what are the current 3 swings. We know that there are 4 market states. We can eliminate Squeeze pattern in this situation. This leaves us Expand and Trend.

If Trend = You would be caught at the wrong end of the market, attempting to trade up in a downtrend, of its 3rd swing pushing price down further. Statistically, this 3rd swing is very dangerous to trade against, especially if it wants to break 1.20 extension. Frequency distribution wise this is a very dangerous slippery slope! However... it doesn't mean this is not tradable.

If Expand = You have a good chance of price moving in your favor if the overall context is actually a big uptrend, that this expansion pattern is just part of it. Also, the statistics are in your favor; the length of 1st and 2nd swing combined as the reference. The movement of the 3rd swing in this case is much more predictable, given the density and slope of frequency distribution.

2-How far can this type of 3rd swing can go
This is much easier to answer now, with risk eliminated and hence now we can determine an optimal exit.

If Trend = Accumulative Frequency Distribution wise;
90% of time, the 3rd swing is at least 0.60 length of the reference swing and downtrend resumes
50% of time, the 3rd swing is at least 1.00 length of the reference swing and downtrend resumes
10% of time, the 3rd swing is at least 1.50 length of the reference swing and downtrend resumes
90% of time, the 3rd swing is at least 1.00 length of the reference swing and reverses trend
50% of time, the 3rd swing is at least 1.70 length of the reference swing and reverses trend
10% of time, the 3rd swing is at least 3.20 length of the reference swing and reverses trend

http://i58.tinypic.com/jg5tsm.png
Thats why I said, its actually tradable, but you really have to be very careful. As you can see, the odds can be really good.


If Expand = Accumulative Frequency Distribution wise;
90% of time, the 3rd swing is at least 0.40 length of the reference swing
50% of time, the 3rd swing is at least 0.70 length of the reference swing
10% of time, the 3rd swing is at least 1.10 length of the reference swing
http://i58.tinypic.com/14sp2sl.png

In conclusion : To be able to anticipate almost half and even full length 90% of time, I would take it. Combine this with overall context (e.g. higher price frames)... odds do stack up in favor.
  • #38
  • Aug 15, 2014 10:35am Aug 15, 2014 10:35am
Quoting LITEchild
{quote} Once the 'structure' is understood, I think any serious trader can map out his or her own simple probability distribution curves with respect to any given permutation.
You are right. But thats their job to figure it out plus also see if this style fits them. I can only give 1 example and that is the one related to iDoubleStoch's work. It is exhausting typing all these out, I got to admit, because it describes my thought process in detail.
  • #39
  • Edited 11:03am Aug 15, 2014 10:52am | Edited 11:03am
Quoting LITEchild
{quote} So let me get this straight...basically, the major difference in your methodology is in the algo you use in constructing your 'waves', which as you mentioned earlier is a modified form of Barros Swings?
Sorry for missing your question!

Yes, this major difference matters. It is indeed a modified form of Barros Swings, to the point I didn't dare to call it Barros Swings sometimes. Yet I do wish to credit Ray for making a difference in my trading.
  • #40
  • Aug 15, 2014 10:56am Aug 15, 2014 10:56am
A summary of where we are now :
1-Price Data :
M1

2-Chart :
Modified Range Bars

3-Wave Tool to map out key price points :
Modified Barros Swings

4-Tool must be able to help identify all market conditions :
Four types of 3 swing states; Trend, Squeeze, Expand Resume/Reverse
To consider 5 swing states

5-Gather statistics as per market condition :
Done, can improve further if more swing VS swing relationships are found, especially those beyond the 3 swing range.

6-Use the statistics to create my own wave theory :
Done, can improve further, dependent on #4 and #5.

Other important considerations to model and test :
GMMA levels confluence with swings
TCD relationships with swings



Updated this in the 1st post as 'Status'
0 traders viewing now