The HolyGrail, want to help make it? - Page 3

elitecamper · #21 (**permalink**) 08-11-2007, 09:59 PM

Quote:

Originally Posted by jlpi

Elitecamper,

I would like to discuss these 2 very strong statements you made because they surprise me a lot.
- First "the holy grail can be mathematicly proven", could you please elaborate on that, give us more information or links to documents that demonstrate it. From my knowledge the people in Universities working in the market theories support more an efficient market theory, rather than a "holy grail" theory. So anything supporting your statement would be very interesting.
- The strategy is a martingale: again from my knowledge (I may be wrong) if there is 1 system that can be mathematicly proven that it fails it is martingale. So in my opinion it should be the last one you should think to use.
Martingale fails at least for 2 conditions that are requested in order to have it work and that cannot be met: your account needs to have an infinite some of money, and you place orders with an infinite sze of lots. (I think it's hard to put orders over 100 lots)
But it's maybe only me who misunderstood what you said.

Thank you jlpi for your reply. i will give it my best, to bring more clarity to my statements.

my statements are not based on market theory.
they are not based on live trading, nor performance.

the other day i read a story about a coin that was one of only 10 minted in 1853 and how it was the "holygrail" of coin collecting.

now that was the holygrail of a coin collector.

a trader's holygrail is different than the grail of the coin collector, of course.

my holygrail is different than the holygrail of a trader.

now, if i apply my "holygrail" to trading it works like this.

if the price ticker moves up 2 pips. then moves down 5 pips. then moves up 3 pips. my holygrail would have made 2+5+3 thats 10 pips profit.

i know thats not taking into account spread. but the example shows how my holygrail works.

so if u add spread to my previous example it would look like this.
if i was trading euro. first move 2 pips up, no profit there. 2nd move 5 pips down, 3 pip profit there. 3rd move 3 pips up, 1 pip profit there.
so (2 - 2) + (5 - 2) + (3 - 2) = 4 pip profit for my holygrail.

so this being the scientific holygrail. this will never be a system, of course not! a martingale system is the closest thing that comes to it. in pip intake.

all systems that make profit are accessing the scientific holy grail for a moment for a glimpse of it. most of the time only on one side.

elitecamper · #22 (**permalink**) 08-11-2007, 10:29 PM

Quote:

Originally Posted by omelette

Personally, I think there should be a ban on the use of HG in thread titles - if only to keep down the clutter as it attracts newbies like moths to a flame. Every month or so you find the same stuff being regurgitated, (normally be a newbie to forex), which is fustrating.
@Craig, completely agree with you, especially about Steinez strategy. I've been away for awhile and was completely amazed by the fervour being generated by Steinez. I was incredulous after spending several hours going through the threads in various forums to learn that this 'systems' basic premise is 'trade with the trend' - and that's it! No money management apart from vague references to trade multipe pairs to manage risk, no exit strategies, nothing!
It has been shown that 'entrys' have a relatively minor influence when compared to risk management over the long haul. But rather than concentrating on this, 90% of people devote their time to picking the 'best' entry and then nothing, just hoping it works out.....

yes omelette i understand your frustration. I just want you to know that i debated a long long time on weather i should use the word holygrail. It was not easy. dont be so hard on the newbies, everyone one was a newbie at one time. newbies can not come up to your level, but u can go down to their level if u wish to do so. dont be so hard on newbies they already have it hard enough.

elitecamper · #23 (**permalink**) 08-11-2007, 10:50 PM

Quote:

Originally Posted by MiniMe

elitecamper,
I would suggest that you start reading about stochastic (predicting the next move )

Now in nonlinear switched systems we can have either a common solution or a global solution, common solution means that the solution will satisfy the equations of the system over the whole region (let’s say one year) so I wouldn’t recommend that, global solution will satisfy the equation for each sub system over that small period.

In your search for a control system look for a global solution, which means you need t o cut down your system to smaller periods and hence you will end up with a new equation for each period (months or week) and adjust the feedback (current price) to change the state variables accordingly
There is a good book about this called Switching Systems and Controls by Daniel Liberzon, however I won’t recommend his approach as he use lyapunov function to prove most of the theories, the issue with this is that you can get a very good results if you got the lyapunov function but there is no easy way to find the lyapunov function

Using stochastic will be a better way to approach this problem(Holy grail ) I am sorry I didn’t made any research about this issues, but once I do I will let you know

I admire your scientific approach I wish all the luck and success

Thank you for your pointers MiniMe they are invaluable to me.

elitecamper · #24 (**permalink**) 08-11-2007, 10:55 PM

Quote:

Originally Posted by MiniMe

by the way if you took the daily setting form the flash files I send to you, you will be able to get the daily trend ( or lets say the new day trend based on the previous days and yesterday candle ) however keep this approach to your self and please don't post it here its copy right material.

Thank you for reminding me MiniMe, im very thankful for your helping hand to a new commer like me to trading. I will not repay you in the wrong way. I hope that one day i will be able to share something of value with you too. thank you

MiniMe · #25 (**permalink**) 08-12-2007, 05:08 AM

Quote:

Originally Posted by elitecamper

Thank you for reminding me MiniMe, im very thankful for your helping hand to a new commer like me to trading. I will not repay you in the wrong way. I hope that one day i will be able to share something of value with you too. thank you

your welcome
here is a way to find a global solution to a currency pair over a period of time, you need to know control theory and how to use Matlab then you can mathematically prove the Holy Grail. Don’t forget read about stochastic
Interaction between MetaTrader 4 and Matlab
Interaction between MetaTrader 4 and Matlab via CSV Files - MQL4 Articles
Price Forecasting Using Neural Networks
Price Forecasting Using Neural Networks - MQL4 Articles

elitecamper · #26 (**permalink**) 08-12-2007, 05:51 AM

HEHE time to put my laser printer to work

ya i need to find a good source on stochastic ill find one

prasxz · #27 (**permalink**) 08-12-2007, 09:19 AM

Quote:

Originally Posted by MiniMe

your welcome
here is a way to find a global solution to a currency pair over a period of time, you need to know control theory and how to use Matlab then you can mathematically prove the Holy Grail. Don’t forget read about stochastic
Interaction between MetaTrader 4 and Matlab
Interaction between MetaTrader 4 and Matlab via CSV Files - MQL4 Articles
Price Forecasting Using Neural Networks
Price Forecasting Using Neural Networks - MQL4 Articles

I like NN forecasting + fuzzy logic is more complete tools ..the problem is how many data to feed that NN and how to optimize the NN to recognize pattern it's required times and effort ...does anyone here has NN tools to forecast currency with good result ?

===================
Forex Indicators Collection

elitecamper · #28 (**permalink**) 08-12-2007, 10:04 AM

prasxz do a search for the thread called artificial intelligence.
The AI EA there, is a neural network you can train it using the mt4 optimizer.
just optimize it on the period you want then test it. u can learn to do it all in the backtester and optimizer in mt4. You have everything u need already. so if u want to learn more about neural networks download the ea and start training it with the optimizer, its not hard. check it out its cool.

prasxz · #29 (**permalink**) 08-12-2007, 03:18 PM

Quote:

Originally Posted by elitecamper

prasxz do a search for the thread called artificial intelligence.
The AI EA there, is a neural network you can train it using the mt4 optimizer.
just optimize it on the period you want then test it. u can learn to do it all in the backtester and optimizer in mt4. You have everything u need already. so if u want to learn more about neural networks download the ea and start training it with the optimizer, its not hard. check it out its cool.

Thanks ...I'll find it

===================
Forex Indicators Collection

money machine · #30 (**permalink**) 08-13-2007, 05:15 PM

Quote:

Originally Posted by prasxz

I like NN forecasting + fuzzy logic is more complete tools ..the problem is how many data to feed that NN and how to optimize the NN to recognize pattern it's required times and effort ...does anyone here has NN tools to forecast currency with good result ?

===================
Forex Indicators Collection

I THINK YOU SHOULD USE MONEY MANEGMENT