Monthly Archives: June 2014

Back to pivots

I’m going to half archive the idea behind a delicate secret. In essence, it’s kind of like an early sign of trading the wick. Wait for extreme (given by the projection line), and trade away from it (with a TP at the other end of the bar). One of the issues I think it has is the fact that statistically, the bar hits the target in the next bar. This makes it difficult to create a setup or trigger after the first signal bar has been given. Additionally, as attached to my pivots as I am given all the time I’ve put into it, there are plenty of other pivot indicators that are capable of producing the same signals as mine do. This means that the ‘secret’ that I discovered is just a part of the market, more than it has to do with my indicator discovering an edge that is difficult to produce elsewhere. Nevertheless, it’s a good thing to keep in mind when trading.

I spent a lot (a lot) of time yesterday thinking and talking about what to do next, and I think adding onto the pivots is where I want to go. The more recent statistics have definitely been more eye opening about how the market functions, although I’m still clueless about the important parts. There are two premises that I believe to be true: the market must retrace by x% in a given ‘wave’ and has a potential of Y% during it’s moves. I’m starting to see every word in that sentence in a slightly different meaning. “Significant” is a word that I’ve been thinking about lately, because at the EoD, if I don’t care about it, I don’t deem it significant and the chart, statistics, and everything else should follow that guideline

s

 

Hit bars expanded

I took the idea of the hit bar and did some work on it taking in the delicate secret theory.

However before getting into the data, the thread has been an eye opener in two ways. First, it has allowed me to get a glimpse of what Rel mentioned a long long time ago about the types of traders: Wave, Level, and Volatility. This has shown me a little bit about the types of things level traders probably look at. Secondly, it’s given me a target for EoT (End of Trajectory) trading. One of my ‘mantras’ if it could be called that, is that if you want to trade profitably, you can do it in two ways: either know when the current trend is ending, or know if the current trend is continuing. Statistics seem to favor the latter, but this project has given me some insight to the first one. In the ideal of ideals, I want to wait until the market shows me a clue, and then have a pattern edge that favors the trajectory either continuing or ending.

There are two possibilities for a hit bar: either the bar closes through the level (above for longs, below for shorts), or it doesn’t (like a wick on a R/S level). I thought there would be some difference between the two types of bars and the effect that would come after it. Turns out to be right, and for a good reason: bars that wick the top already have momentum to go back to their original level. Remember, I’m looking to short at the green level and long on the red level. Very important filter. Bars closing through have momentum!

probability

I added the 30 hours just as a rough estimate of giving the bar until EoD to complete. (used the not through criteria) Comparing the 1 day, 30 hours, and 1 week, there’s clearly diminishing return on probability to time, but even so, nearly 90% chance to hit the level during the course of the next week is pretty darn good odds if you ask me. Of course, a good chunk is completed by the end of 24 hours, but it’s still nice to know a level might be valid if price hasn’t touched it yet.

 

Then, the usual house keeping:

Of instances where the level is returned to within 30 hours, half the time the level is hit in the very next bar.

time

Yet, potential is pretty good from the High of the hit bar (or whatever high comes after that) and the low of the bar that hits the level.

pot

I don’t have to check but I know results are slightly skewed to the high side given how markets have changed over time. This has some good clues on where entry should be. However, there aren’t any clear stop signs in the data yet…

 

MQL4 log and blog

Log:

1. Draw something. Anything
2. Draw lines that makes sense
3. Draw Daily data on 1HR TF
4. Fix lines not appearing until EoD
5. Fix lines not appearing on smaller TF  ×
6. Draw projection data, Daily data from yesterday being projected today
7. Draw lines using function calls
8. Implement pivots
9. Enable more than 19 days of pivots to be viewed

 

6/12: successfully drew SOMETHING on the chart:

6/19: successfully drew scribbles, next step draw something that makes sense

success

Later that day, got levels to follow H/L, next step get lines to draw D1 datasuccess06/22: Got D1 data to appear on my H1 chart! Fix lines not appearing until EOD

AW YEA

 

06/23: It’s peeking out! Got lines show at before EoD. Also got data to be created from yesterdays data, not current data. At this point recreating traditional pivots is not too difficult I don’t think. Time to work on using data other than MT4 given iHigh and iLow to give values.

Capture

next step get from using iHigh/iLow to custom array formulas

6/25: Beta version complete: Adds/subtracts pivot from open, but PP formula needs tweaking to better match excel output.

Capture

6/26 Finished! It’s a bit buggy, but it works! At least in the scope that I need it to for now.

Capture06/27:

After completing the indicator, I “leveled up” again by being able to appreciate what creating indicators does, and I kind of see the process like this:

1. Observe market pattern
2. Find a way to test for market pattern in a quick fashion to see if something might be there (bulk of my recent work in excel)
3. Build indicator to view on the chart
4. See if a new market pattern emerges with the indicator on the chart (What I recently gained in leveling up).
5. Rinse/repeat.

While I love excel and have grown very comfortable with it, having a visual indicator is not just for making work easier, but for allowing new patterns which otherwise would have been hidden to be shown. I realized this after the teamviewer session and asking about the process of getting from point A to B.

 

I don’t browse forums much since the majority of it is not applicable to me, I do enjoy reading them and always have since my gaming days. I’m a particular sucker for anything with a flashy or cryptic title, and while I rarely learn anything that I can apply directly in my own work, I try to follow the thread to get a sense of how other people view the market. I don’t really care if there’s no proof of the method, as that is something that I will undertake myself. Up until now, everything has been about waves waves waves. And while waves are the fashion in which price moves, having other forms of “market analysis” allows me to get the “same signal from 2 independent sources” idea.

One example of an idea I read about was in a thread the Forrest pointed me to towards about half a year ago “the Ultimate truth”. In it was  basically the idea of not just roulette-ing, but classified roulette-ing among multiple types of “input”. The thread died and became the project of “Fractals, ZZ, and the pissing dog”. A basic version of the ‘Ultimate truth’ idea was done in one of my statistics back then, where I found nothing (not surprisingly). I thought that I could take the concept from the second thread to be applied to my waves, but I had too many issues in the design phase that I never went through with it. The creator of the original thread somewhat recently made another thread titled “A delicate secret” which i won’t explain, but the idea behind it has unintentionally (or perhaps subconsciously) influenced what I may have found in the charts post indicator.

In taking Rel’s advice, I spent some time looking at post bar hit, rather than pre-bar hit.

levelsI noticed that when 1) there is only 1 hit of the day and 2) the hit bar fully closes through the level, that price returns back to the pivot, or rather back through the original hit bar within about a 24hr period. This is somewhat similar to both the idea in the ‘delicate secret’ thread, as well as Rel’s “accepted level” which I have been trying to find a way to implement. There may be a pattern, lets see if I can rinse and repeat and come back with a better edge.

 

 

Just a note

I’ve had a number of people comment and try to get in contact with me now but my email replies have not been answered. I’m not sure if this is due to people flaking, or an error sending on my end, but if you’d like to get in contact with me(and you’ve already attemped), you can email me at totallybogus434@gmail.com, and then tweet me @D5Air. Or ya kno. check the inbox.

The weekend gap

Here we go!

A lot of things said about the gap.. lets see what the numbers say:

Odds to fill? Fantastic really. Gap forecasting ability of the days MM? Not so much,

gap1

Differences between gaps that fill vs never fill:

gap2

On the top is gap size. Little blip there for gaps between 60-70, however considering it’s not in the extreme (80+), I see it as more of a fluke than anything else. Besides, with 16 occurances total over 10 years.. not really what I’d consider an edge.

On the bottom move size (H-O vs O-L). Perhaps those that don’t fill make larger moves on Monday in direction of the gap (‘propelled’ by it). Not quite the case.

Lastly, Time:

Here’s what I thought would be a more fun way to show the relationship:

I would think the larger the gap, the more time it takes to fill, as indicated by the yellow line. However, this is not completely the case…

gapgraph2

Along with the normal output:

gap3

Seems to me that the suggestion is that the bigger the gap, the lower the probability to fill, and the larger the gap, the more time it takes to fill, but not quite so linearly. 24 hours is pretty much the expectancy.

 

 

Overview June 2014

About a year ago was when I first started attempting this new approach to the markets. I have become convinced that the markets are not random, however finding patterns that are actually trade-able to exploit these non-random movements is still out of my reach. One of the perhaps coolest things I have learned is the process and progress of “leveling up”. I had faith that waves, statistics, and programming were keys, but I had to really work through it to see and appreciate the value in it.

Here’s how I see some of the groundwork, my overview:

1. Waves

  • Waves aren’t purely trade-able patterns. You don’t just trade the current up pattern which will turn into a down pattern. They’re meant to show price.
  • The reason why there’s so much work that goes into them being “correct” and “accurate” is because in a sense, they replace the standard OHLC time frame candle chart.
  • Nothing is more pure than OHLC data coming in every time there is a change in one of those elements. The question is how best to record these movements? OHLC every 5 minutes, hour, 24 hours, is a way to do this(standard). Waves are another. However, since the OHLC TF data is all that we have access to (as it’s the universal norm) it appears that waves are derivatives of price, with OHLC being the ‘true chart’ when this is not the case. OHLC and waves are both derivatives of true price movement which is almost intangible.
  • Thus when one trades CCI, stochastics, RSI, etc, they’re really trading patterns that someone else discovered. Do they still work?? Maybe. But one really should know the actual formula and how these double derivative indicators work before simply trading a divergence on one or above/below a certain level on another.

2. Statistics

  • A lot of people seem to say they don’t trust statistics because they change. Actually first, a lot of people don’t think there are any statistical edges in the market… and they’re just wrong. But statistics never lie. They’re just numbers. It’s how we interpret them and make conclusions using them that creates lies and other false information.

3. Programming

  • This is just more of a self thing but damnit is it difficult. Originally I set out to try it, ultimately to quit because it was too difficult. I attempted a second time, quitting again because although I could appreciate it’s value, I had no value to gain from it. Now, I am on my third attempt. I see it’s value for me personally, but the difficulty has not changed. I write all the code which looks ok to me but nothing appears on the chart. Why? I haven’t a clue. Clearly it’s zeroing out somewhere, but not having the step by step debugger like I’m use to in VBA makes it very very challenging.

As for actual trading and progress, I’m at the point where I finally fully appreciate the need to move to waves(I think). Not to zigzags or barros swings or signal bender waves, but rather just a generic term for trading a price structure that isn’t single time frame. It’s also not the case that one can’t make bound-time frame data work, but instead, it’s that moving to waves broadens the range of possibilities for which an edge can be found. This is one of the cases when you realize what you were missing out on and now you can never go back. Perhaps its time for a real haitus for blogging? I really think I need the programming skills now and at the rate that I’m learning(zzz), it’ll be quite some time before I can program anything, really. I’m considering bumping my excel based research down to maybe just 1 day a week for a sake of uncovering statistics that I’ve thought about but never really had the motivation to do, and spending the rest of the time trading to code.