The 3-plan | Page 11 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Trading plan template 53 replies

Plan the Trade / Trade the plan 74 replies

Plan the Trade and Trade the Plan: Swing Trading the Euro 5 replies

do you have a written trading plan? 31 replies

my first trading plan (very primitive) 5 replies

Attachments: The 3-plan
Exit Attachments

The 3-plan

  • #201
  • Dec 6, 2013 8:14am Dec 6, 2013 8:14am
Quoting myself at post 171. The entry and exit suggestions are way too simplified. Back testing using a number of combinations did not provide much success. I need to re-think the logic process for this double stoch method.

The original idea presented by Eurusd is actually simple to apply but there is always the issue with people's ability to compare the 1ema and xstoch to look for a difference, this was encountered in the "Similarity" thread but an indicator was developed to help "spot" the difference. The same indicator or a striped down version of it might be useful here.
I want more brain power!
  • #202
  • Dec 6, 2013 12:29pm Dec 6, 2013 12:29pm
Another approach could be to wait for the stoch to get close to top/bottom with a high volume and then trade in the direction of the price, se pic.

Waez

http://i43.tinypic.com/2u42cn7.gif
  • #203
  • Edited 2:35pm Dec 6, 2013 1:54pm | Edited 2:35pm
Quoting Eurusdd
{quote} Yes. You must keep up with price dynamically. A value for stoch may work for 5 hrs or 1 week and the fail. The moment it failed, another will be in perfect sync. That should give you a massive edge.
Exactly, I have added the slope of an SMA to the calculation so that you can see which is the optimal Stoch value to use.

In the below, the optimal stoch (the one with the highest correlation to the close price and the slowest slope value) would have been somewhere around Stoch(200).

The correlation % is shown in Blue with the slope of the SMA shown in Red.

Attached Image (click to enlarge)
Click to Enlarge Name: StochCorrelationWithSlope.png Size: 30 KB


I stepped forward in time 10000 minutes and ran the same analysis and you can see that the optimal slope changed to be around Stoch(1000).

Attached Image (click to enlarge)
Click to Enlarge Name: StochGPUWithSlopeLater.png Size: 26 KB


I'm not sure if any of this is useful but at least you can see that the optimal stoch value is dynamic and needs constant recalculation.
  • #204
  • Dec 6, 2013 1:58pm Dec 6, 2013 1:58pm
Attached Image
  • #205
  • Dec 6, 2013 2:01pm Dec 6, 2013 2:01pm
Quoting Ralome
Same thing, but this time I applied a window having the same size as the stoch itself. So, in case of a 1440 stoch, I needed 2880 bars. I don't know if this is useful at all, just felt like doing it. Although, not having a nice extra GPU, it's kinda slow for me {image}
The reason this is required is due to the actual formula used to calculate the stoch. A stoch is basically calculated as

(price - lowest of x bars) / (highest of x bars - lowest of x bars)

So, to calculate the value of Stoch (1234) for price at time x, you need to have the calculation look at the past 1234 bars to find the highest and the lowest values. The period that I call "window" is the walk forward period after this time. For example, if I wanted to check the value of the correlation for stoch (1234) over a window of 300 bars, I would take the first 1234 bars, find the high and the low and the get the first stoch value. Keep moving forward another 299 bars and then compare the 300 stoch values you have against the last 300 prices and you will have a valid correlation figure.

Hope this makes sense, let me know if you need more details...
  • #206
  • Dec 6, 2013 2:15pm Dec 6, 2013 2:15pm
Quoting Eurusdd
{quote} haha. meaning that level could hold till NFP on friday. That is the top for the 1ema with 80% chance.
Yep, you called it, the level held flat as a pancake..... Where did the % figure come from?

Attached Image (click to enlarge)
Click to Enlarge Name: LevelHolds.png Size: 55 KB
  • #207
  • Dec 6, 2013 2:20pm Dec 6, 2013 2:20pm
Quoting Eurusdd
{quote} The stoch in this case is used to pick tops or bottoms and that is all.
I've been thinking about this as well...

Given the formula used to calculate the stoch value, the stoch alone cannot be used to pick tops or bottoms. The stoch going to 0/100 is only confirmation that the price is a new low/high price during the last x bars. So seeing a new 0/100 level reached is not an indication that we have hit a top or a bottom, just the possibility that it is a top or a bottom. I guess it is how you calculate the probability of the possibilities that is important.

Would you recommend another indicator that helps this process?
  • #208
  • Dec 6, 2013 2:49pm Dec 6, 2013 2:49pm
Quoting 64f_v2
{quote} Yep, you called it, the level held flat as a pancake..... Where did the % figure come from? {image}
got the mean value right. but little deviation here and there.

In other words, I called eurusd to fall when the news is released. After that, anything goes.
  • #209
  • Edited 3:50pm Dec 6, 2013 2:51pm | Edited 3:50pm
Quoting Eurusdd
Hello: The value you choose for STOCH at any given time will give a curve whose shape is exactly the same as the 1EMA in the main chart. As time goes on there will be a moment when the curves do not look the same anymore. "Look the same" means exactly what it means - look the same in shape, slope etc.... When the two curves are not similar, it means there must be another value of STOCH that will give a curve similar to the 1EMA, currently. search for that value and discard the current one ... Continue this process and you will be right there with...
Can't really understand the need to find new value for STOCH when "THE 1EMA and the current stoch-curve will start to look different as soon as there is a strong underlying force AND stoch is very close to one of the extreme values".

I understand that that we only need to change the value for STOCH ONLY if the stoch-curve look different at the extreme top/bottom; which I suppose is to mean when "stoch-curve is very close" to 0 or 100 ?

BUT, what could be the end result once we get a new perfect value which is perfectly-in-sync with the 1EMA?

Is it not enough to understand that "there is a strong underlying force" beneath such difference in stoch-curve at such extreme top/bottom THAT is is not safe to enter a counter-trend trade yet, for as long as the stoch and 1EMA are not similarly curved yet, sir? And perhaps, we should just ride the trend.

The attached chart might be able to illustrate the "underlying strength of current trend" though I only used the value of 200 for that stoch.


Thanks
Attached Image (click to enlarge)
Click to Enlarge Name: diff.PNG Size: 23 KB
  • #210
  • Dec 6, 2013 4:29pm Dec 6, 2013 4:29pm
Next week, I will be watching how the price play out if it could go below areas marked blue both on the line chart and stochastic window.

My hypothesis is: below either or both those blue areas, the price is normalized and the so called underlying uptrend strength is neutralized. So, stoch is back in business without us having to change its value
Attached Image (click to enlarge)
Click to Enlarge Name: diff.PNG Size: 23 KB
  • #211
  • Dec 7, 2013 1:10pm Dec 7, 2013 1:10pm
Quoting 64f_v2
{quote} The reason this is required is due to the actual formula used to calculate the stoch. A stoch is basically calculated as (price - lowest of x bars) / (highest of x bars - lowest of x bars) So, to calculate the value of Stoch (1234) for price at time x, you need to have the calculation look at the past 1234 bars to find the highest and the lowest values. The period that I call "window" is the walk forward period after this time. For example, if I wanted to check the value of the correlation for stoch (1234) over a window of 300 bars, I would...
Yes, I understand the window term, I meant that choosing a 1440 window for a 1440 stoch was too much for my processor :-)

Speaking of that, which programming language do you use?

And about your other post with the slope values included: how did you calculate the slope value exactly? Did you just take the difference of the last two MA values?
Even if you fall on your face, you're still moving forward.
  • #212
  • Dec 7, 2013 1:49pm Dec 7, 2013 1:49pm
Before committing too much of your life to EURUSDD threads read this post http://www.forexfactory.com/showthread.php?p=7142705#post7142705 you have been warned.

I too have wasted 100s of hours on his last four threads only for him to "cut and run".

Challenge him strongly and ask for evidence before believing and wasting your valuable time following an over inflated ego hiding behind someone else's forward predictions!!.
  • #213
  • Dec 7, 2013 4:01pm Dec 7, 2013 4:01pm
Quoting go4it
Before committing too much of your life to EURUSDD threads read this post http://www.forexfactory.com/showthread.php?p=7142705#post7142705 you have been warned. I too have wasted 100s of hours on his last four threads only for him to "cut and run". Challenge him strongly and ask for evidence before believing and wasting your valuable time following an over inflated ego hiding behind someone else's forward predictions!!.
Relax man. What is it about this challenge him stuff.
Are we at war. It is a forum for God sake. I have serious things to do apart from posting here, so do not expect me to spend my entire day replying to every post or taking part in a challenge contest!
lol

I give other members the change to take over the thread because I do have all wisdom!!!
  • #214
  • Dec 8, 2013 5:30am Dec 8, 2013 5:30am
Quoting Ralome
{quote} Yes, I understand the window term, I meant that choosing a 1440 window for a 1440 stoch was too much for my processor :-) Speaking of that, which programming language do you use? And about your other post with the slope values included: how did you calculate the slope value exactly? Did you just take the difference of the last two MA values?
I use c# mainly as it's quick and easy for me to plug into mysql, cuda (via cudafy.net) and dukascopy (via ikvm).

For the slope calculation, for each stoch length I created a SMA of the same length eg Stoch(1234) and SMA(1234). On the GPU I the calculated the values over the whole data set and used the last x bars as a window, in these examples the last 300 bars (minutes). For the slope calculation, I had the GPU calculate the slope of the linear regression of the sma. You can't use the first and last values because they will not accurately represent the windowed data set. The slope of a linear regression quickly and easily captures the trend of the slope of the window period.

Hope that helps?
  • #215
  • Dec 8, 2013 7:53pm Dec 8, 2013 7:53pm
Quoting 64f_v2
{quote} I use c# mainly as it's quick and easy for me to plug into mysql, cuda (via cudafy.net) and dukascopy (via ikvm). For the slope calculation, for each stoch length I created a SMA of the same length eg Stoch(1234) and SMA(1234). On the GPU I the calculated the values over the whole data set and used the last x bars as a window, in these examples the last 300 bars (minutes). For the slope calculation, I had the GPU calculate the slope of the linear regression of the sma. You can't use the first and last values because they will not accurately...
Keep up the good works guys.

I am less fortunate in the sense that I just can count 1,2,3..., and not the technicality of maths as u guys do.

I strongly believe that, it is enough that I know the facts that market is overbought, oversold or at critical level 50 of stochastic and how price react at those areas to make a sound decision. If it move higher even though overbought and I ride the trend, there must be a reference price if I were to decide to cut loss should the price reverse

Cheers
  • #216
  • Dec 9, 2013 1:58am Dec 9, 2013 1:58am
eurusd closed below a critical number for the first time. 1.36985. So today, could turn out to be a fairly-bearish day. The nest few hours could close beloe that number .1hr chart
  • #217
  • Dec 9, 2013 3:26am Dec 9, 2013 3:26am
failed. Broke the top.
  • #218
  • Dec 9, 2013 5:33am Dec 9, 2013 5:33am
Quoting 64f_v2
{quote} I use c# mainly as it's quick and easy for me to plug into mysql, cuda (via cudafy.net) and dukascopy (via ikvm). For the slope calculation, for each stoch length I created a SMA of the same length eg Stoch(1234) and SMA(1234). On the GPU I the calculated the values over the whole data set and used the last x bars as a window, in these examples the last 300 bars (minutes). For the slope calculation, I had the GPU calculate the slope of the linear regression of the sma. You can't use the first and last values because they will not accurately...

Thank you!
Even if you fall on your face, you're still moving forward.
  • #219
  • Jan 30, 2014 3:05pm Jan 30, 2014 3:05pm
Guys, here is my little contribution to wake up this thread, maybe not .. Also, I added "difference" option so 2 stoch's doesn't need to be in perfect sync to show vertical lines in OB/OS areas.
Attached File(s)
File Type: ex4 2 x stochs diff.ex4   7 KB | 542 downloads
  • #220
  • Mar 4, 2014 4:36am Mar 4, 2014 4:36am
Hi Eurusdd,

I have a question for you if you don't mind. In your last thread, The Similarity system, you have grown your account from 3.000 USD up to 20.000 USD within five days.
And my question is this: You have used just the similarity principle in order to achieve that or you have used and the other principles that you talk about (revolution etc.)? Because that performance was astonishing and I would ask you again if you traded the system in both ways (the similarity and the dis-similarity)?

Thank you!
0 traders viewing now