DOSI: Daily Open Strategy Improved | Page 4 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Let's Improved this Strategy, Grid based System 1 reply

Improved Rsi Needs Code 5 replies

Creating a Simple Daily Open Price Strategy EA 1 reply

Improved Rsi Needs Code 1 reply

Attachments: DOSI: Daily Open Strategy Improved
Exit Attachments

DOSI: Daily Open Strategy Improved

  • #61
  • Oct 6, 2021 4:36am Oct 6, 2021 4:36am
Quoting dokopy
I suggest trying this strategy as well. It is on the daily charts and gives more signals. https://www.forexfactory.com/thread/...-candle-system
Thank you for the suggestion.

The 217-page read, and even post #1 is rather intimidating. Quickly skimming over that thread, it looked like it needed a lot of discretion and manual judgement, or am I wrong? Obviously, only a strategy with clear black/white rules can be automated in an EA.

Can you point me to a post in that thread that compactly summarizes the strategy?
  • #62
  • Oct 6, 2021 5:20am Oct 6, 2021 5:20am
Guys, you are so funny.
DOS is really great system and straight forward, and you seek "idi..o..tic" systems instead?
That guy "mastermind" uses TP of 2 pips in his published account, do you know that?
  • #63
  • Oct 6, 2021 5:53am Oct 6, 2021 5:53am
Hi Mate:

Can we add two input parameters "Profit Lock in Pips" and "Profit Lock Trailing in Pips"?

It works like Trailing Stop to lock in profit ( for both Single and Basket Trades ), as to prevent profits slipped away or dragged longer in trades.

For example, TP is set to 100 pips, and we set Profit Lock in Pips 60 pip, and Profit Lock Trailing in Pips 10 pips.

With that, we know our single/basket trades will have at least 60 pips in profits once price reached that level, and if should price divert again, we have at least 60 pips in profits locked, and, move on to next trades.

Or, if should price move in our favor, we will trail the Profit Lock in every 10 pips, until either the TP is hit or the Profit Lock is hit.

I hope my explanation is clear.

All in all, the EA is great, the strategy is good, what I trying to do is to maximize the profits and reduce the drawdown on portfolio basis.

Again, thank you for your valuable time and contribution.
  • #64
  • Oct 6, 2021 6:11am Oct 6, 2021 6:11am
Excellent idea
  • #65
  • Oct 6, 2021 6:36am Oct 6, 2021 6:36am
Quoting yoriz
{quote} Thank you for the suggestion. The 217-page read, and even post #1 is rather intimidating. Quickly skimming over that thread, it looked like it needed a lot of discretion and manual judgement, or am I wrong? Obviously, only a strategy with clear black/white rules can be automated in an EA. Can you point me to a post in that thread that compactly summarizes the strategy?
I'm trying a stupidly simple system. The problem is that without EA I have to wait for inputs at night. I pulled the indicator from the thread. I used PT from DOS settings. I put SL at the nearest peak in history. I chose the most moving pairs I tested in DOS.
Please excuse the bad English via Google Translate.
  • #66
  • Oct 6, 2021 6:48am Oct 6, 2021 6:48am
Quoting MT4-EA
Guys, you are so funny. DOS is really great system and straight forward, and you seek "idi..o..tic" systems instead? That guy "mastermind" uses TP of 2 pips in his published account, do you know that?
I agree with you. DOS only requires good pairs with good drawdrown to reduce risk. Nothing more. Even in lower TF’s it’s reliable.
  • #67
  • Oct 6, 2021 7:45am Oct 6, 2021 7:45am
Quoting amargam
{quote} I agree with you. DOS only requires good pairs with good drawdrown to reduce risk. Nothing more. Even in lower TF’s it’s reliable.
Provide here all the information we need to achieve the same results as you. Thanks.
Please excuse the bad English via Google Translate.
  • #68
  • Oct 6, 2021 7:54am Oct 6, 2021 7:54am
Quoting dokopy
{quote} Provide here all the information we need to achieve the same results as you. Thanks.
At this point I’m testing only two TF, 1H and Daily. At the moment I prefer don’t say anything about settings because it requires too much study and I’m on a 5% of the total test. The conclusion at the moment, for me, its the strategy in a few pairs it’s lot size dependant. So my opinion it’s control the risk very well.

A few posts ago yoriz shows that in three different scenarios the curve is growing because of averaging down strategy. Maybe the MM its the key, but its my opinion.

and also, this is backtest, lets do the work on forward.
  • #69
  • Oct 9, 2021 9:56pm Oct 9, 2021 9:56pm
Quoting yoriz
I have uploaded version 1.45 that does stricter input validation to avoid problems during Optimizer runs. Please give that a try! {quote} To reduce the lead-in time needed before the first trade is made, I enable trading already when have encountered only 1/10th of NrBarsLookback bars instead of waiting for the full NrBarsLookback bars. But with value 1 this effectively means 0.1 bars, which is rounded down to 0. So the minimum value is actually 10, not 1. Nonetheless, I don't expect such short lookbacks to be useful, since you need to collect enough...

Hi Mate:

I hope this message will find you well.

Mate, can you add an input parameter for Magic Number for the EA?

I am demo testing the EA with D1, H12 & H1 time frame on one same pair GBPUSD. I believe without proper different magic number, the EA might mess up the trades.

I hope you got what I meant and the situation I am trying to describe. Let me know if I got it wrong.

Mate, thx for your time.

Regards,
Bill
  • #70
  • Oct 10, 2021 4:18am Oct 10, 2021 4:18am
Quoting bdunn
can you add an input parameter for Magic Number for the EA?
Sure. I have uploaded version 1.50 which has a configurable MagicNumber input parameter.

Quoting bdunn
For example, TP is set to 100 pips, and we set Profit Lock in Pips 60 pip, and Profit Lock Trailing in Pips 10 pips.

With that, we know our single/basket trades will have at least 60 pips in profits once price reached that level, and if should price divert again, we have at least 60 pips in profits locked, and, move on to next trades.

Or, if should price move in our favor, we will trail the Profit Lock in every 10 pips, until either the TP is hit or the Profit Lock is hit.
Thanks for the suggestion. Not sure I fully understand what you are trying to do. If you are at exactly 60 pip profit, what should happen? Should the EA then start trailing at 10 pips distance? So worst case you would get 60-10 = 50 pips profit. How is this different from a trailing stop?

Can you explain more what should happen when price reaches the 60 pips Profit Lock?
  • #71
  • Edited 9:08am Oct 10, 2021 5:10am | Edited 9:08am
Quoting yoriz
{quote} Thanks for the suggestion. Not sure I fully understand what you are trying to do. If you are at exactly 60 pip profit, what should happen? Should the EA then start trailing at 10 pips distance? So worst case you would get 60-10 = 50 pips profit. How is this different from a trailing stop? Can you explain more what should happen when price reaches the 60 pips Profit Lock?
Add two input parameters: "Profit Lock in Pips" and "Profit Lock Trailing in Pips"

Should the EA then start trailing at 10 pips distance? So worst case you would get 60-10 = 50 pips profit. How is this different from a trailing stop?

Hi mate: This is exactly how the EA work. When it reaches 60pips in profit, the EA will start locking the profit 60-10 = 50 pips in profits, at least.

Then, if price move 10 pips further, then it will trail and lock 50+10 = 60 pips, till the price hit the TP, or, the Profit Lock.

It is same as Trailing Stop, but in this case, only trailing when in profit. We locked the profits and trailing it.

I hope I explain it well. Let me know if you need move examples.

All in all, I would like to have the EA not to have the profits slipped away, reduce the time on the market, reduce the risks & drawdown.
  • #72
  • Oct 10, 2021 8:40am Oct 10, 2021 8:40am
It is called trailing step.
  • #73
  • Oct 11, 2021 5:51am Oct 11, 2021 5:51am
Hi Mate:

I demo test the EA today with a number of forex pairs and different time frame.

1 weird thing I found out today is, the EA can be removed automatically, gone missing from the chart.

I noticed the problems, thought it was something wrong on the MT5 platform.

I reinstalled the platform and the same problems occurred, at least I managed to catch it.

My computer setup is top notch and up to date in terms of CPU and memory RAM.

Here are the codes for your investigations:

2021.10.11 08:31:57.814 323. yzDosi-1.50 (XAUUSD,H1) jsOrder::reset: resetting open order is not allowed. Aborting EA.


2021.10.11 08:31:57.814 321. yzDosi-1.50 (XAUUSD,H1) ExpertRemove() function called


2021.10.11 08:31:58.004 323. yzDosi-1.50 (XAUUSD,H1) 318 undeleted objects left


2021.10.11 08:31:58.004 323. yzDosi-1.50 (XAUUSD,H1) 6 objects of type BarChart left


2021.10.11 08:31:58.004 323. yzDosi-1.50 (XAUUSD,H1) 300 objects of type Bar left


2021.10.11 08:31:58.004 323. yzDosi-1.50 (XAUUSD,H1) 12 objects of type MovingAverage left


2021.10.11 08:31:58.004 323. yzDosi-1.50 (XAUUSD,H1) 46592 bytes of leaked memory


2021.10.11 08:31:58.009 321. yzDosi-1.50 (XAUUSD,H1) jsOrder::setLotSize: modifying order after opening is not allowed. Aborting EA.

  • #74
  • Oct 11, 2021 7:36am Oct 11, 2021 7:36am
Set File Attached for Yoriz

Pls rename it to .set

Attached File(s)
File Type: txt Bill Intraday.txt   1 KB | 376 downloads
  • #75
  • Oct 11, 2021 7:40am Oct 11, 2021 7:40am
In addition to the errors above, I managed to catch these lines of error moment ago:

2021.10.11 10:29:26.783 Trades '1051031422': failed market buy 0.01 XAUUSD [Position doesn't exist]


2021.10.11 10:29:26.783 Notifications invalid settings (notifications not enabled)


2021.10.11 10:29:26.789 Experts expert yzDosi-1.50 (XAUUSD,M5) removed

  • #76
  • Oct 11, 2021 7:41am Oct 11, 2021 7:41am
I reattached the EA again to the chart(s), loaded it with set file, then the EA opened and closed trades as normal. Very normal.

Then, maybe an hour or few hours later, the EA will gone 'missing' in self-removed mode.
  • #77
  • Oct 11, 2021 9:14am Oct 11, 2021 9:14am
Quoting bdunn
the EA can be removed automatically, gone missing from the chart.
Yes, that is by design. If the EA encounters a very unusual error, it will stop trading and remove itself from the chart to avoid making invalid trades.

Quoting bdunn
2021.10.11 08:31:57.814 323. yzDosi-1.50 (XAUUSD,H1) jsOrder::reset: resetting open order is not allowed. Aborting EA.
2021.10.11 10:29:26.783 Trades '1051031422': failed market buy 0.01 XAUUSD [Position doesn't exist]
These errors suggest that the internal bookkeeping of the EA gets out of sync with reality. It thinks some orders were closed by sl/tp while they were not, or tries to update an order that was closed?!

Quoting bdunn
2021.10.11 08:31:58.004 323. yzDosi-1.50 (XAUUSD,H1) 46592 bytes of leaked memory
This is normal, and not a problem with your MT5 platform. This is the result of the EA abruptly terminating itself without cleaning up used memory.

I can not reproduce this problem in the Strategy Tester with the settings you provided which makes it hard to debug it. I will open a demo account to see if I can reproduce the problem there.
  • #78
  • Oct 11, 2021 11:19am Oct 11, 2021 11:19am
I do most of my testing in the Strategy Tester, and therefore did not notice the weird behavior when switching the timeframe of the chart the EA is attached to. That would cause partial(!) reinitialisation of the EA. This could be the cause of the "resetting open order is not allowed." error.

Uploaded version 1.55 tries to fix this behavior (as always, see post #1 for the link). In addition, as a fail-safe, the EA will try to close all open trades whenever it encounters a fatal error. Note that I can not guarantee this always works flawlessly, because upon a fatal error things can be unreliable preventing the EA from closing trades.

Hope this resolves some/all of the problems.
  • #79
  • Oct 19, 2021 4:51am Oct 19, 2021 4:51am
Some users like to examine a wide range of entry criteria and therefore make the EntryStrategy input parameter also part of their optimization runs. Having RandomDirection in the selectable options obviously does not make sense, so I moved it to the begin of the range (now option A), to enable optimization runs using only options B .. D. Beware when using previously created .set files that options have changed!

However, in my opinion this strategy does not really seem to have an "edge" (see post #41) and purely relies on averaging down. This means that optimization runs should focus only on finding good TakeProfitPips, MinScaleInPips and MaxDrawDownPercent, to make sure the averaging down part of the EA is tuned. I don't think it is useful to try to optimize the entry criterium, as this will only be curve-fitting to lower the drawdowns in simulated years but you will definitely encounter them in live trading.

My advice would be to optimize only the input parameters in the "Exit Signal Settings" and "Recovery Settings" sections, and then do many random runs to estimate what the risk of ruin is, and what are realistic lot sizes. For that, I added the RandomizationSeed input parameter. To verify your optimized settings, you could select the RandomDirection entry strategy and have MT5 "optimize" only the RandomizationSeed input parameter (range between 1 and 1000000000). That way, the strategy tester will go over many different scenarios so you get a feel for what drawdowns you can expect on average and in the most extreme cases. This can help choosing the right lot sizing.

Does this approach make sense?
  • #80
  • Oct 20, 2021 7:36am Oct 20, 2021 7:36am
Quoting yoriz
Some users like to examine a wide range of entry criteria and therefore make the EntryStrategy input parameter also part of their optimization runs. Having RandomDirection in the selectable options obviously does not make sense, so I moved it to the begin of the range (now option A), to enable optimization runs using only options B .. D. Beware when using previously created .set files that options have changed! However, in my opinion this strategy does not really seem to have an "edge" (see post...
I agree. There are some pairs that makes good DD, about 3-7% with minimum 85% winner trades in 10 years at least. Of course the most important parametres to “play” and find good dd and good balance are: Pips Body Size and TP. There are some pairs that if increases lotsize, good DD goes to 35-40% DD. So the key is search good pairs with its own parametres.

im only testing it on daily tf, but Im sure that in lower tfs there are good opportunities
0 traders viewing now