@Eursudd
For math we like, how we can do this mathematically?
Many thanks
For math we like, how we can do this mathematically?
Many thanks
- #111
- Dec 4, 2013 8:04am Dec 4, 2013 8:04am
- Joined Aug 2009 | Status: (I have a huge) Member | 5,555 Posts
dreams do come true.
{quote} This is as near to the "HOLY GRAIL" as you can get. Paul Tudor Jones says he's made most of his money from calling Tops and Bottoms. I think we can do so too in our "small fish" way. Going thru EURUSDD's other thread -"The Similarity System", I have come to adopt/adapt this template. The thread is worth every minute of study. {image} {image} {file}
- #114
- Dec 4, 2013 8:19am Dec 4, 2013 8:19am
- Joined Aug 2009 | Status: (I have a huge) Member | 5,555 Posts
dreams do come true.
- #115
- Dec 4, 2013 8:27am Dec 4, 2013 8:27am
- Joined Jan 2012 | Status: Loving Nature* | 3,985 Posts
{quote} Thanks so much for taking the time, X. so, when we see after we see the dissimilarity one was to wait for the stoch to come to the same level without the dissimilarity. is this correct? i understood from what i read that one had to wait for the price and stoch to be in sync again but i didn't know that we had to wait for the same level to be tested. makes sense. the way i see it i would expect 3520, 16 or even 10 to long so i hope i will see some sort of dissimilarity there so i can enter a trade using this system too. do you trade...
In my case, I did the opposite. since I saw that dissimilarity, I waited price went far away from that point, and after I trade expecting the approach to that point again...
you can trade it both ways, depends on your willing and proposals to the trade you are going to open!
ps: only to udpate, eurjpy is almost there again, where the dissimilarity was shown for first time...
will be testing this principles...
grateful child!!
{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.
In the example below, I took 6000 samples, applied all the stochs from stoch(400) to stoch(2500) then performed a correlation over the window of samples 4100 to 4400 (a window of 300 samples).
You can see in the chart below, the best stoch values were between 1900-2000 and about 2445.
I applied the stock values and the following chart shows the stoch overlaid against the price for the windowed period.
It's interesting that if I step forward in time, you can see that the correlation falls apart, this could either indicate that the stochs need to be recalculated, or perhaps a period of dis-similarity has started.
So the above shows that it possible to calculate the stochs dynamically. In my case I use a GPU to calculate all the stochs and correlations in parallel, so I can process a few thousand combinations per second. I need to think about the complete system further before continuing, I suspect it will be critical to select the correct window period for the stochs, but to understand that I need to consider the whole system.
As always, thanks to eurudd and any thoughts or guidance would be appreciated....
- #118
- Dec 4, 2013 8:41am Dec 4, 2013 8:41am
- Joined Aug 2009 | Status: (I have a huge) Member | 5,555 Posts
dreams do come true.
- #119
- Dec 4, 2013 8:55am Dec 4, 2013 8:55am
- Joined Jan 2012 | Status: Loving Nature* | 3,985 Posts
Dear X, still trying to understand. sorry for being lazy: i know i could/should go read the similarity thread instead of bothering u... i have draw the attached chart. the vertical line show the beginning of the dissimilarity. that's the number i should keep in mind, right? 1.3570. my question is (again): - wait for stoch to get back down to 0 without dissimilarity and long to 1.3570 or - wait for price to get to 1.3570 and short to the low? the first seems to be top-bottom picking strategy (which fits my style better) and the second trend trading,...
the second one, itīs not like I said, or like I traded. but it can be used to catch few pips, imo.
what I was tolding you was, I saw the dissimilarity when stoch was at 0, waited to price recover and going above, waited to overbought and stoch at 100, and at that time, price was similar to stoch (at 100), so I shorted expecting the approach again to stoch at 0, to fix the dissimilarity...on that way, I catch all the ride from the top to the fixing problem which occured at stoch 0. so itīs different than starting short only in oversold(stoch at 0), but it can also work to catch few pips, like I catched right now!
