Advertisement
2nd Research Continue of Salvation EA - SaveMe.
Sep 14, 2010 at 05:05
538 개의 뷰
8 Replies
Apr 20, 2010 부터 멤버
게시물808
Sep 14, 2010 at 05:05
Last week has been busy study the coding of EA and the theory to a martingale hardly blow account has close to success.
Strategy use:
1. Calculated lot size of leverage so it adequate to survive.
2. Follow martingale base on trend instead of hedge.
3. Base on 1 minute data back test.
4. Open price base on MA.
5. Minimum to cut loss as follow hedge will lock your loses most time.
6. Basket profit all when reach $10.
7. Cut loss profit $10 when open 7 order trade.
8. Initial deposit also reduce from 3k to 1k.
This is version 3 - SaveMe3. Which has some error that if it not able to cut loss at $10, it continue open.
And we can see that on Nov and Dec is a dangerous market trend where price up and down did not follow technical analysis.
Strategy use:
1. Calculated lot size of leverage so it adequate to survive.
2. Follow martingale base on trend instead of hedge.
3. Base on 1 minute data back test.
4. Open price base on MA.
5. Minimum to cut loss as follow hedge will lock your loses most time.
6. Basket profit all when reach $10.
7. Cut loss profit $10 when open 7 order trade.
8. Initial deposit also reduce from 3k to 1k.
This is version 3 - SaveMe3. Which has some error that if it not able to cut loss at $10, it continue open.
And we can see that on Nov and Dec is a dangerous market trend where price up and down did not follow technical analysis.
첨부 파일:
Information is Gold when come to organised.
Oct 28, 2009 부터 멤버
게시물1409
Sep 14, 2010 at 08:05
You're wasting your time running backtests with 25% modeling quality.
11:15, restate my assumptions: 1. Mathematics is the language of nature. 2. Everything around us can be represented and understood through numbers. 3. If you graph these numbers, patterns emerge. Therefore: There are patterns everywhere in nature.
Apr 20, 2010 부터 멤버
게시물808
Sep 14, 2010 at 08:37
All the back test are same, no matter 90% or 25%, is just a speculation of logic. It will sure have some different later.
But my first open logic only base on indicator, after that is pure mathematic grid trading, which mean the tick of indicator no longer important as martingale will just keep on open trade at certain pip of price.
And from the data provide by Alpari UK, seem 2005- 2007 only have hundred trade compare to 2008 and 2009 has thousand over trade. Hmmm... probably data corrupted.
If I increase the modeling quality , which is to test on a higher Time Frame, the trade will be even less.
Model quality increase but it only give you bogus result as it trade less.
Look at Forex Bullet Proof, it only design to run on 1 minute. It result is worst than mine, but it probably sell Hot as marketing.
Is it the EA logic problem or the Data problem ? Can 1 minute data actually get 90% model quality ?
But my first open logic only base on indicator, after that is pure mathematic grid trading, which mean the tick of indicator no longer important as martingale will just keep on open trade at certain pip of price.
And from the data provide by Alpari UK, seem 2005- 2007 only have hundred trade compare to 2008 and 2009 has thousand over trade. Hmmm... probably data corrupted.
If I increase the modeling quality , which is to test on a higher Time Frame, the trade will be even less.
Model quality increase but it only give you bogus result as it trade less.
Look at Forex Bullet Proof, it only design to run on 1 minute. It result is worst than mine, but it probably sell Hot as marketing.
Is it the EA logic problem or the Data problem ? Can 1 minute data actually get 90% model quality ?
첨부 파일:
Information is Gold when come to organised.
Oct 28, 2009 부터 멤버
게시물1409
Sep 14, 2010 at 08:41
You need to download tick data, you can get it from Dukascopy.
Here's how to do it
http://eareview.net/tick-data
Yes Forex Bulletproof will sell well and no doubt will bankrupt hundreds if not thousands of people one day.
Here's how to do it
http://eareview.net/tick-data
Yes Forex Bulletproof will sell well and no doubt will bankrupt hundreds if not thousands of people one day.
11:15, restate my assumptions: 1. Mathematics is the language of nature. 2. Everything around us can be represented and understood through numbers. 3. If you graph these numbers, patterns emerge. Therefore: There are patterns everywhere in nature.
Apr 20, 2010 부터 멤버
게시물808
Sep 14, 2010 at 08:43
After re-coding some minor parameter. SaveMe4 version has come out.
첨부 파일:
Information is Gold when come to organised.
Apr 20, 2010 부터 멤버
게시물808
Sep 14, 2010 at 09:05
(편집됨 Sep 14, 2010 at 09:07)
Problem with Tick Data Download.
1. It download day by day. 2000-Jan-Day 1
2. It got Ask and Bid separate to download.
So you need to download 365 Days X 2 for Buy and Sell to get accuracy, and sure my Platform will crash after loaded so many history data.
This is the download table for Day 1 Jan 2009.
1. It download day by day. 2000-Jan-Day 1
2. It got Ask and Bid separate to download.
So you need to download 365 Days X 2 for Buy and Sell to get accuracy, and sure my Platform will crash after loaded so many history data.
This is the download table for Day 1 Jan 2009.
첨부 파일:
Information is Gold when come to organised.
Oct 28, 2009 부터 멤버
게시물1409
Sep 14, 2010 at 09:08
Yep, it certainly isn't an easy thing to do but then nothing worthwhile ever is.
11:15, restate my assumptions: 1. Mathematics is the language of nature. 2. Everything around us can be represented and understood through numbers. 3. If you graph these numbers, patterns emerge. Therefore: There are patterns everywhere in nature.
Apr 20, 2010 부터 멤버
게시물808
Sep 14, 2010 at 09:15
So are there easy way ? Like using Dukascopy platform to download 1 year data first, then export it to CVS.
Then using it to import to MT4 platform ? Or need also a converted to convert the CVS to MT4 format ?
Then using it to import to MT4 platform ? Or need also a converted to convert the CVS to MT4 format ?
Information is Gold when come to organised.
Oct 28, 2009 부터 멤버
게시물1409
Sep 14, 2010 at 11:20
Not sure. I've never come across another way.
Of course you could just write your strategy in Java and test it with Jforex to start with
Of course you could just write your strategy in Java and test it with Jforex to start with
11:15, restate my assumptions: 1. Mathematics is the language of nature. 2. Everything around us can be represented and understood through numbers. 3. If you graph these numbers, patterns emerge. Therefore: There are patterns everywhere in nature.
*상업적 사용 및 스팸은 허용되지 않으며 계정이 해지될 수 있습니다.
팁: 이미지/유튜브 URL을 게시하면 게시물에 자동으로 삽입됩니다!
팁: @기호를 입력하여 이 토론에 참여하는 사용자 이름을 자동으로 완성합니다.
