読み取り専用

フォーラムは現在、読み取り専用のアーカイブとなっています。.

バグ報告やプラットフォームに関する質問はこちら → [email protected]

私たちのコミュニティはDiscordとYouTubeで活動しています。ぜひ参加してください!

ご参加ください Discord YouTube

Pro適合トラッキング

18 replies

マイキー

顧客、bbp_participant、コミュニティ、856件の返信。.

プロフィールを表示

10年前 #114970

Hi SQ Team and users,

I have a question about Profit Trailing.  This is the type of trailing where the stop loss is moved X pips from the current price as it moves in your favour.  This is not the Stop Trailing feature which uses an indicator value or comination of indicator values to move the stop on the bar open.

 

My question is, Profit Trailing has been implemented in the EA (MQ4 code) to only move the stop loss on bar open, is this what is intended?

 

I would have thought the use of profit trailing (like MT4 terminal offers on an open order) is to move the stop with X pips of the current price on every tick, to capture fast moves.

 

So, for example if you have a 1 hour strategy, and price moves very quickly within that one hour bar, the stop would be moved many times.  As it is, the stop is only moved every hour.

 

よろしくお願いいたします。,

 

マイク

0

イーストピース

Customer、bbp_participant、community、sq-ultimate、298件の返信。.

プロフィールを表示

10年前 #136218

“I would have thought the use of profit trailing (like MT4 terminal offers on an open order) is to move the stop with X pips of the current price on every tick, to capture fast moves.”

 

Agree!

 

Long Position’s profit trailing is based on the highest price from entry.

Short Position’s profit trailing is based on the lowest price from entry.

0

geektrader

顧客、bpb_participant、コミュニティ、503件の回答。.

プロフィールを表示

10年前 #136221

Trailing, just like any operation in SQ, is performed at bar open only in the SQ backtests as well, not just in the created MT4 EAs! This has a reason, because you otherwise can´t create/backtest strategies accurately in SQ anymore. Remember, to trail every tick, you´d need to use TICK DATA for each generated strategy, even if it ´s a H1 one (or at least 1M data (to inaccurate though as I know from my past EAs)). So this would completely eliminate the possiblity to generate strategies on H1 in “selected time frame only mode” as the results in the SQ backtest would be trailed by each bar open only, but every tick in MT4 – the strategy will look COMPLETELY different then. Or, we´d have to use tick-data for strategy generation, even in H1 mode, all other modes would be impossible to be used anymore.


🚀 自動化された外国為替戦略開発であなたの優位性を引き出せます 🚀

1987年から開始された歴史的な外国為替データ。28ペア、M1、99%エラーなし、生涯無料アップデート

0

geektrader

顧客、bpb_participant、コミュニティ、503件の回答。.

プロフィールを表示

10年前 #136222

Personally I don´t see a reason for that, the strategies SQ finds all work on bar open and are good and intended to work this way. You won´t get any benefit if changing the MT4 versions to out of sudden to trail by the tick, the strategy will fall apart completely then from how it was created in the first place by SQ (just give it a try and you´ll see). Also, you´d have to use tick-data all the way if you want to create intra-bar trailing strategies and that would not allow us to build strategies from 1987 to 2016 anymore since tick-data is only available from 2007 and up, which is much to short. Additionally, strategies relying on intrabar-trailing and hence tick-data, will never work like this in live trading. The backtests will be inaccurate still. I once had a scalping EA which used real tick data and trailing at every tick, it looked fantastic in the backtest, every tick was modeled via Birts Tick Data Suite. Yet, in live trading, it failed, because even with TDS MT4 tick-backtesting is inaccurate and will be different in live trading if trailing by every tick. Even if using tick-data from your broker, the live trades never matched the backtests if comparing afterwards. So even if we could do it this way, the generated strategies will most likely never work live forward like they look in SQ tick-data-backtest. In comparison, the H1 or M30 strategies that I am using live, that only trail each bar open, and that have been generated by SQ, match perfectly to the backtests if comparing the lives trades with the SQ / MT4 backtests afterwards.
 
If you want fast trailing, you can use the lower timeframes though, like M5 or M1, though it will be difficult to find any longer term working strategies there, at least not if using many years of data for the creation. But SQ would then trail each 1 minute or each 5 minutes, etc.


🚀 自動化された外国為替戦略開発であなたの優位性を引き出せます 🚀

1987年から開始された歴史的な外国為替データ。28ペア、M1、99%エラーなし、生涯無料アップデート

0

イーストピース

Customer、bbp_participant、community、sq-ultimate、298件の返信。.

プロフィールを表示

10年前 #136224

Thanks, geektrader.

 

You tell the truth.

 

I didn’t find fragility in my way until today.

0

マビ

Customer、bbp_participant、community、sq-ultimate、252件の返信。.

プロフィールを表示

10年前 #136225

I gave up minute and hourly bars and are finding it much easier to build strategies using only daily bars.Even building them on only data from 1988 to 1992 i find that the best have really simple entries and more complex exits that gives nice even curves in OOS from 1992 to 2016 with drawdows that are smaller then i can find on H1. The only intrabar trades are the loosers the winners usually runs for days and weeks and even months.Looking intrabar you often see that many loosers could have been a winner but howe to catch that without destroying the winners eludes me.

0

geektrader

顧客、bpb_participant、コミュニティ、503件の回答。.

プロフィールを表示

10年前 #136227

Great to hear it works for you. H1 is the best for me here.

 

Simple: you CAN`T catch all the winners, it´s impossible and if you manually close them, the next trade entry will be wrong since the system wouldn´t have placed a new one if the previous trade would still be open. Even on my H1 strategies there are heaps of trades which would have been winners and turn back to losers. However, adding a BE doesn´t work as the backtest shows, so just stick to what SQ gives you as it is the most stable version of the system. You can´t win them all and it doesn´t matter if you win them all, what matters it that you make a profit on AVERAGE in the long run, lose less than you win overall, and you will be profitable! 🙂


🚀 自動化された外国為替戦略開発であなたの優位性を引き出せます 🚀

1987年から開始された歴史的な外国為替データ。28ペア、M1、99%エラーなし、生涯無料アップデート

0

マイキー

顧客、bbp_participant、コミュニティ、856件の返信。.

プロフィールを表示

10年前 #136263

If profit trailing is only executed every bar rather than every tick, it is no different to trailing stop such as Close[1] – 30 pips or Close[1] + 30 pips.

 

Doesn’t seem very useful to me unless it is tick based (only allowed on real tick perhaps), and maybe has a trigger, e.g.

 

IF Market Price > High[2] THEN Profit Trail on Tick Market Price – 30 Pips

 

または

 

IF Order Open > 20 Bars THEN Profit Trail on Tick Market Price – 30 Pips

0

geektrader

顧客、bpb_participant、コミュニティ、503件の回答。.

プロフィールを表示

10年前 #136268

Well, the strategies generated this way do work very well for me with trailing on bar open only:) You can just remove the bar open check in the MT4 code, you´ll see the strategy will turn to waste then if they trail by tick. The strategies SQ creates are meant to trail on bar open only, they do not work better if they trail by tick out of sudden – give it a try by editing the MQ4 code to not just trail on bar open. If SQ would generate them with tick trailing, then it´s another story. But then the problem comes in that tick trailing is SO highly broker dependent because the characteristics of tick-streams are heavily different between brokers (e.g. some brokers deliver 100, other up to 700 ticker per minute) that it will not work in live trading, even if you created the strategy with tick-data (and even if you use your own brokers tick data that you´ve collected with my MQL4 script for that). Believe me, I´ve been down that route more than a few times and it never ever worked live, the backtests simply won´t match. While if trailing on bar open only, I get a 99% match of live trades to SQ / MT4 backtests. If you need faster trailing, just create strategies for 1M timeframe then, then SQ will trail each minute and you don´t need tick-data to backtest / create them nor will they be unreliable in live trading since amount of ticks within the 1 min bar and which we can´t know during backtesting will not matter.


🚀 自動化された外国為替戦略開発であなたの優位性を引き出せます 🚀

1987年から開始された歴史的な外国為替データ。28ペア、M1、99%エラーなし、生涯無料アップデート

0

サミー・ドゥ

顧客、bbp_participant、コミュニティ、6件の返信。.

プロフィールを表示

10年前 #136300

Ok guys, but wouldn’t it be a nice option of SQ if it were capable of generating entries in a higher TF and the trailing is done on a lower TF?
For example you have the entries on H1, but the trailing stops are modified at the open of every M5 bar.
In M5 TF geektrader’s mentioned tickdata related problems should not be that relevant anymore.

This way we could have a closer trailing of EAs running in higher timeframes.
But I guess such a feature would be considered as ‘multi-timeframe’, which will probably come with SQ 4.0, right?

0

geektrader

顧客、bpb_participant、コミュニティ、503件の回答。.

プロフィールを表示

10年前 #136301

Yes, that is an idea. I though fear with all the possibilities and flexibility of SQ4, it will become a perfect curve matcher. You basically can create straight line strategies on 100 years of data if simply using all the options that SQ4 will offer, yet they will most likely completely fail once live. Just saying… I will stick to my old-school method of H1 opening / trailing etc with as few indicators as possible and only use such strategies as they´ve turned out to be most stable going forward. But I fear the “newbies” will waste a lot of money after getting excited about how well a strategy looks if they use all the options like signals on H1, trailing on M5, with custom trailing start, custom blah, etc etc…. Just something to be aware of, more options do not mean more stable strategies, it means the exact opposite.


🚀 自動化された外国為替戦略開発であなたの優位性を引き出せます 🚀

1987年から開始された歴史的な外国為替データ。28ペア、M1、99%エラーなし、生涯無料アップデート

0

マイキー

顧客、bbp_participant、コミュニティ、856件の返信。.

プロフィールを表示

10年前 #136302

Yes, that is an idea. I though fear with all the possibilities and flexibility of SQ4, it will become a perfect curve matcher. You basically can create straight line strategies on 100 years of data if simply using all the options that SQ4 will offer, yet they will most likely completely fail once live. Just saying… I will stick to my old-school method of H1 opening / trailing etc with as few indicators as possible and only use such strategies as they´ve turned out to be most stable going forward. But I fear the “newbies” will waste a lot of money after getting excited about how well a strategy looks if they use all the options like signals on H1, trailing on M5, with custom trailing start, custom blah, etc etc…. Just something to be aware of, more options do not mean more stable strategies, it means the exact opposite.

 

Absolutely true, however there’s a difference between having a flexible tool and using all the features.  Newbies might turn all features on, but experienced users will know what they are looking for and which freatures/blocks to use to try and find optimum strategies.

 

Taking your view to the extreme, remove all features from SQ, sorted.  No more curve fitting, no options to select.

0

マビ

Customer、bbp_participant、community、sq-ultimate、252件の返信。.

プロフィールを表示

10年前 #136306

One way would be to create 2 strategies with the same entry and have one trail  and one run to catch the larger profits and maybe this way try to lower the draw downs if possible. I will  try this but i think Ninjatrader minute resolution data is not supported as with MT4 ?

0

geektrader

顧客、bpb_participant、コミュニティ、503件の回答。.

プロフィールを表示

10年前 #136313

Haha yea, Mark would be happy, make a lot of money and have nothing to code into the software:)

 

What I try to point out is that I hope that newbies will be made aware that adding “another option” and yet “another option” to their strategies will not make them more stable, but I think the new documentation + new RT´s will make it easier in that relation as well.


🚀 自動化された外国為替戦略開発であなたの優位性を引き出せます 🚀

1987年から開始された歴史的な外国為替データ。28ペア、M1、99%エラーなし、生涯無料アップデート

0

Tomas Matejka

管理者、sq-ultimate、2件の返信.

プロフィールを表示

10年前 #136316

What any trailing can also do is that it gets you out of good winners often prematurely while not preventing large losses (when market turns against you right after entry). That means that using trailing does not automatically guarantees smaller drawdowns and higher profits. To some extent more space you give to market to work in your favor higher the chances it will hit your profit objectives. As geektrader said trailing triggered once per bar can work better then precise tick-by-tick trailing system since you give market more space until current bar is closed. When trading manually you have an advantage of trailing your stop-loss logically when you follow a trend for example while with auto-trading you loose that advantage and you have to rely on some simple rule that will hardly ever beat the discretion

0

ガディオ・エフエックス

サブスクライバー、bpb_participant、コミュニティ、22件の回答。.

プロフィールを表示

10年前 #136497

I tried in every way to make my profit trailing in tick-by-tick mode and not on bar close
but I could not find a solution!!
the only option may be is to change the mql file, but does anyone know how to explain this feature to convert into a real trailing without waiting for the bar?
 
I’m going crazy!
 
: /
 
GREEKTRADER said “You can just remove the check open bar in the MT4 code, you’ll see the strategy will turn to waste then if they trail by tick.” .. But I did not understand what he wants to hear.
 
GREEKTRADER, would you explain to me how to do?
 
I have to delete some code inside the file?
 
what should I do? add some code inside?

0

15件の返信を表示中 - 1 - 15件目 (全18件中)

1 2 →