SAR system based on QQE Adv and MACD Platinum

Hi, Jim,

I have recorded 3 videos to demonstrate how to set all possible combinations you have requested based on my understanding. Please take a look at those videos and see if I have covered everything.

I use Jing to make those videos so I don't have to upload to youtube, this is why same video split into 3 since there is a 5 minute limitation. This is the method I use for demonstration to all my clients as you already known. If I upload all my videos, then my day will be all spent in uploading videos. Sorry for the inconvenience for everyone who is watching.

http://screencast.com/t/9bkSkhGB

http://screencast.com/t/r83u69LTkh0X

http://screencast.com/t/jxcYiKxtq8

Also one more thing to add, if you are using a time frame that is more than 1h, session settings are automatically disabled as well.

Let me know if anything is not clear, if it is, I can start making this EA. Thanks.

Christina,

If I may say so — one programmer to another — I'm VERY impressed with the level of detail, and the quality of the explanations, in your videos.

IMO they should be mandatory viewing for anybody who's contemplating having an EA developed. At face value, a strategy like Jim's might seem simple enough, but your videos illustrate exactly how an analyst/programmer must work through every potential combination of possible scenarios.

Consequently, traders tend to significantly underestimate the amount of work that's needed to create a correctly working EA.

@everybody else: Hey folks, we programmers are not mind readers! I've seen literally dozens of threads here where (unlike Jim) a trader obviously spends less than 5 minutes explaining what he thinks he wants, and then expects a programmer to spend several hours somehow figuring out how to best put the nondescript pieces together...... and all without having to pay for it.

David

Hi, David,

Thanks for your comments. These comments are obviously from a fellow programmer, a very experienced on I might add based on the quality of your posts.

Without talking too much about this, I agree most clients do not present their strategy in a very clear and detailed fashion, some are incapable of doing so, some not spending enough time to do their part of the work. Can't begin to tell you how time consuming and frustrating when that happens. But without getting to the bottom of every detail, the programming can't really starts. Clients often think they are paying for the time spent on coding, in fact more time is spent to verify the logic and testing and on going support, this is the reason most clients tends to underestimate the amount of work put in to develop a strategy. Especially if you want to get high quality piece without bugs from the beginning.

As a programmer, all I want from the clients are 2 things: explain their system clearly and respect the work and pay the reasonable amount for the time spent on it. A programmer is likely to be more motivated if they see those 2 qualities from a client.

On the other hand, I also see a lot of not so responsible programmers. Some gives out buggy codes, most likely because starting the projects without spending time thoroughly understand the system. I frequently bid on projects where most programmer quote a price without even see the detailed description of work scope, which is completely ridiculous.

Specific to this project, at first glance this is with a simple enough logic, in fact I have at least coded several variations of it in the past, but with all the other factors putting into mix, this does needs some careful planning. In fact, careful planning and considerations should be given to all projects in order to achieve the correct results with the most efficient codes. Of course, what separate good and bad coding is not always obvious in the eyes of the clients.

David, you are right. Christina knows her stuff and presents it very well. I have always found the videos a great tool to help understand the EA. Something she has done with all of my EAs to date. Like I said, and Christina touched on it in her reply to you, I did try other Programmers through oDesk and they let me down badly, hence me seeking out Wise-EA.

Hey Christina,

Thanks for the videos. Had to watch the 1st one a couple of times just to get an understanding of what 'Instant Sequence' meant. Am I right that if it is true, then a new sequence will begin immediately? eg the current 3x long sequence hits the break even target of 1.3540, then a new 1x long trade opens at that same price beginning a new sequence. If it was set at false, what would happen in this example?

Also on the 2nd video between the 3.05min and 3.50min marks, I just got a little lost. It appears from what you said as soon as the EA is in profit, it closes completely for that session. Maybe I misunderstood, but shouldn't that be based on a profit target only? If profit target was set at -1 (nil target), would the EA just continue to trade throughout the nominated session and close where I elected to close at either 1,2,3 or 4. Or is it that theses options only come into play when there is a profit target set?

Other than that, looks good. I might have to do a bit of maths though with regards to working out profit targets based on deposit currency, but that shouldn't be hard as I'll just have a spreadsheet made up.

So just to clarify what I am looking for:

1. The original concept with no trading times and nil profit target,
2. As above but with a profit target,
3. Trading times and nil profit target, and
4. Trading times with a profit target.

I'm just having problems seeing point no.3 in your videos. No big deal if this one can't be done. Or did I miss it?

I like how you have simplified the number of inputs also. Glad you don't listen to me when it comes to that. Thanks. Jim

Thanks for the videos. Had to watch the 1st one a couple of times just to get an understanding of what 'Instant Sequence' meant. Am I right that if it is true, then a new sequence will begin immediately? eg the current 3x long sequence hits the break even target of 1.3540, then a new 1x long trade opens at that same price beginning a new sequence. If it was set at false, what would happen in this example?

If this is set to false, in your example, EA will wait until next sell signal to start a new sequence, instead of opening a new sequence right away with a buy. I thought you want this option, but if you always only want to start a new sequence right away, this variable can be deleted.

Also on the 2nd video between the 3.05min and 3.50min marks, I just got a little lost. It appears from what you said as soon as the EA is in profit, it closes completely for that session. Maybe I misunderstood, but shouldn't that be based on a profit target only? If profit target was set at -1 (nil target), would the EA just continue to trade throughout the nominated session and close where I elected to close at either 1,2,3 or 4. Or is it that theses options only come into play when there is a profit target set?

Other than that, looks good. I might have to do a bit of maths though with regards to working out profit targets based on deposit currency, but that shouldn't be hard as I'll just have a spreadsheet made up.

So just to clarify what I am looking for:

1. The original concept with no trading times and nil profit target,

This can be done with one of the time leave to blank, and target set to -1.

2. As above but with a profit target,

This can be done with one of the time leave to blank and target set to x. As soon as sum profit of all trades in sequence (can be only 1 trade or multiple trades) reaches x, sequence is finished.

3. Trading times and nil profit target, and

This can be done with setting start and end time, with profit set to -1.

Behavior of EA is this: During the same session there can be multiple sequences, each sequence finished within the session is the same as your option 1, it'll end with either first trade in profit, or overall is break even. There will always be 1 sequence left not finished at session end time, then EA proceed with closing methods. If this case, you can choose method 1,2 or 4, but not 3, since there is no profit target.

4. Trading times with a profit target.

This can be done with setting start and end time, with profit set to x.

Behavior of EA is this: During the same session there can only be 1 sequence, sequence must finish with sum profit of x if finished within the session, then EA will pause until next session. If by end time still not finished, you can choose to close with method 1,2,3 or 4. Notice you can still choose method 2 which is break even, but this should not be confused with profit target if finished within the session time.

I maybe giving you more option than are wanted, for example downgrade the target from x to break even after session close time, this is because I see this as a valid option. If there is anything you don't want, we can leave it out.

It's also possible with this option you also allows multiple sequence, then you can have multiple sequence finish with profit of x, only leave the last unfinished one to the closing method of choice. You can decide on this, it'll be easier to code with multiple sessions instead of just 1 sequence allowed.


I'm just having problems seeing point no.3 in your videos. No big deal if this one can't be done. Or did I miss it?

I like how you have simplified the number of inputs also. Glad you don't listen to me when it comes to that. Thanks. Jim[/quote]

Christina,

Thanks for the quick response. Feel like I am back in school with all that red ink! That what my exam papers use to look like when I got them back. But it does make it quite clear and easy to follow. And you have covered everything that I needed clarifying. No problems leaving that 'instant sequence' in as it is something else I can have a play with later down the track.

If you are happy with everything, then so am I, and therefore good to go. Thanks. Jim

Oh, I didn't thought of that, next time I will use green.

By the way, we use multiple sequences during the session period, correct?

Give me 2-3 days I will send the EA over. Thanks.

Christina,

No, the red was fine. Certainly couldn't miss it!

Yes, there could be multiple sequences during the session period. If there was a target nominated and the target was hit on the first trade, then that session's trading would be completed. But if the target was not hit on the first trade, then the sequence would be complete when (and if) it hit the break even target, then a new sequence would begin (if time permitted) in an attempt to hit the same target, and so on. If no target was nominated but I was trading during a session period, then the EA should just keep on taking trades as per the concept until the end of the session or however nominated by me to end.

Looking forward to seeing what you come up with. Thanks. Jim

As you can see folks, there is a fair bit of communication between the parties involved when building an EA. Both Christina and David stated it is not as simple as it looks to someone that may have a good trading idea. You have to be able to cover every scenario and possible outcome. It can be a bit of a process initially, but trust me, it is all worth it when you more than likely will get a good working model right from the get go, that may only require some minor tweaking. The videos certainly help and it is certainly a joint effort at the end of the day. And just for your info, using a Programmer like Christina (and I'm sure David does the same), they will not run away with your ideas or use them for their own personal or commercial use unless you give them express permission to do so.

Trade update:

Nothing to report as all three pairs still short on their 1st trade of the sequence. Cheers. Jim

Start Date : Jan 10th 2013 on 6x pairs / Jan 30th 2013 on 3x pairs
Closed results to date : +2876 pips

Ok it looks obvious to me that my understanding of your idea is completely wrong. I also don't quite think what you wrote above makes complete sense to me.

So let's say you have a profit target of $100. Looks like you are only looking to get a profit from the first trade of the sequence. If so:

It's possible first trade gets to $100 before next signal. Do you close the trade here and finish the sequence? Or you only close at next signal? If you close right away it's easier for me to understand, but if you don't close, it's possible for this trade turn out to be profitable but less than $100 at next signal, for example, $40. It's also possible for the first trade never reach $100 profit and by the next signal it's profitable such as $40.

Now looking at the part I highlighted in orange, you seem to assume as long as profit target is not hit, it's at a loss, that's not true with my example above, it's $40 and it's above break even point, what is the point of opening another order for this sequence? When you say break even you do mean the sum of profits of all orders in the same sequence is >= zero, do you?

My previous understanding was the profit target is for the entire group of trades of the same sequence, if first gets $40, my next trade goal is to get $60 etc, it certainly doesn't look at this is what you are thinking, it's fine if you can tell me the answer of my questions here.

This also makes the closing methods I designed to be wrong as well. The rules needs to be completely re discussed. But I can only look at that after this basic concept to be cleared. Thanks.

Christina,

I think you are close with your understanding. So if there was a profit target set and that target was hit at anytime during that trade, then that sequence (trade) is immediately closed, and the next sequence (1st trade) would not begin until the next reversal signal. Of course this would depend on trading times, if nominated. So you are right, if a target is hit, the trade is closed immediately and no new trades are taken until the next signal.

Now here is something I didn't consider, that you have pointed out. Say I nominated a $100 profit, and the 1st trade closed at a profit of $40 only, and never made it to the $100 level. Now what? Would it be too difficult to do some maths here where the EA would continue the fib position size sequence and take the next trade as per normal, and when it reached +$60, closed out the trade for an overall profit of $100, and then wait until the next signal before commencing again? Or if that first trade is a break even or better result, just finish the sequence there and then, and then begin a new sequence on the next signal and start from scratch? Open to ideas here and obviously what is easier for you.

In the third scenario, where the first trade finished at a loss, the EA should just continue on as per normal, but now it is just aiming to clear the current sequence at break even. If and when that is achieved, then a new 1x trade would be opened immediately which would then have a go at the profit target.

So possible outcomes when a target is nominated are:

1. 1st trade hits target, sequence immediately closed, and wait for new signal in reverse direction to start a new sequence,

2. 1st trade ends up at break even or in profit short of target:
a) look to make up difference on next trade in sequence, or
b) close sequence completely and take whatever result and start new sequence completely on next signal

3. 1st trade ends up as a loss, the EA goes into normal strategy looking for overall break even result, and if that is achieved, continues on immediately with a new sequence looking for profit target, and so on.

I realise that adding profit targets would confuse the issue and create so many outcomes. These outcomes above are also going to be affected by trading times, if they are selected. See what you can come up with Christina, and if it is too difficult to code, then I am happy just to go with something a bit more basic along the lines of the original strategy. We can always revisit this EA with modifications at a later date. Thanks. Jim

So possible outcomes when a target is nominated are:

1. 1st trade hits target, sequence immediately closed, and wait for new signal in reverse direction to start a new sequence,

2. 1st trade ends up at break even or in profit short of target:
a) look to make up difference on next trade in sequence, or
b) close sequence completely and take whatever result and start new sequence completely on next signal

Both of these are possible but in principle very different. Option a) is what I had in mind, is profit target based on entire group, although I think it's clear your idea is centered around that profit is only achieved with the first order of the sequence. Since you designed this system, I think we should follow your thinking and go with option b). With this design, the difference between using a profit target or not become very minor, only if first order reach the target before next signal will make a difference, all other aspect is the same. With the idea of option a) there is significant difference.

Option b) is easier to code.

With this in mind, now we should ignore those videos I make earlier. Regarding close method when session closing time reaches, with your new principle, you effectively can only have 3 choices:

a). Close the last order immediately at session close time regardless of profit or loss.
b). Close the last order at first signal after session close time regardless of profit or loss.
c). Let the current sequence run it's course as normal. By this I mean if the order is first order of sequence and by next signal, if it's profitable, close order (this will be the same regardless if you use profit target or not as explained above), if not, then open next order and sequence won't finish until break even. If the order is not the first order of sequence, obviously only finish when break even.

3. 1st trade ends up as a loss, the EA goes into normal strategy looking for overall break even result, and if that is achieved, continues on immediately with a new sequence looking for profit target, and so on.

Don't you ever sleep Christina? I have an excuse for being awake as it is only 4.45pm local here and I am on days off.

If you think option 2. b) is the better option, I am happy with that. No problems.

I also understand the closing choices and I am happy with them also.

At the end of the day, I'll probably just trade this EA of the 1hr, 4hr or daily charts with no profit target in mind, and more than likely over a 24hr period. Especially on the 4hr and daily, as I would be able to manually interfere with trade management with regards to moving stops etc, if I so choose to. For example, the EA opens a nice long trade on the 4hr chart, I could easily place a pretty close trailling stop on this trade if I saw potential to lock in profit. The EA could then be removed from that chart and placed on a completely separate chart of the same pair so it could then pick up the reversal signal when it occurs. Plenty of options for me.
The reasoning behind the profit target and trading times was an attempt to lock in small profits using the same concept on the smaller timeframes during the more volatile trading sessions, so I didn't get whipsawed in the usual sideways markets during the quieter times. So no big deal really if the target/trading times is not that fined tuned at this point in time. Again, I could run the smaller timeframes manually once the EA picks up the entry if I was so inclined to be at my desk for hours on end.

But I think we are close to being on the same page here Christina. Thanks. Jim

Trade update:

Big reversal happening on all three pairs. EUR/AUD has now gone long 1x at 1.2985. There was a profit of +21 on the previous short. This is now the start of a new sequence. The other two pairs will probably both turn in about 92mins if they continue up.

Start Date : Jan 10th 2013 on 6x pairs / Jan 30th 2013 on 3x pairs
Closed results to date : +2897 pips

OK we will go with what's being discussed. Will send to you soon.

Thanks Christina, looking forward to it.

Trade update:

The EUR is now long 1x at 1.3542. There was a profit of +94 on the previous short. A new sequence has now begun.

EUR/JPY is now long 1x at 126.64. This is the 2nd trade in the sequence as the previous short had a loss of -48. So there is now a break even target of 127.12. Remember if this break even target is achieved, then that sequence is closed out for an overall break even result, and a new sequence with 1x long begins immediately at that same break even price.

These markets have been kind to me so far with nice and steady returns, and limited big losses. Really need some sideways stuff to put this method under a bit of stress. Cheers. Jim.

Start Date : Jan 10th 2013 on 6x pairs / Jan 30th 2013 on 3x pairs
Closed results to date : +2991 pips

Trade update:

EUR/JPY did the right thing by me with price hitting the overall break even target which closes the current sequence. There is now a new 1x long trade at this same price of 127.12 for the start of a new sequence.

No other changes. Cheers. Jim

Thank you for your system! There was a problem: 2013.02.06 17:54:52 MACD_Platinum GBPUSD, H1: unknown subwindow number -1 for ObjectCreate funtstion, help to fix!

Trade update:

Nothing has changed, but I'd imagine in about 2hrs, a couple of these pairs are going to reverse if nothing changes too much with their price.

bratcevdg,
Thanks for following. With regards to your problem on MACD Platinum, I'm not sure I understand what you are asking. If you could explain it more, even a screenshot, that would be appreciated. Thanks. Jim

Just had a bit of a thought when I was laying in bed last night.

Now I haven't really tested this out so it is just an idea that needs further exploring. What if, using this same trading idea, instead of closing out each trade on a new signal (reversal), I just left the last trade open until there was an overall profit reached.

What I mean by this, is if I started a sequence with a 1x long and then on the next sell signal, if the long was a loss, I would keep that trade open and also short 1x. So there would now be two trades open, one in each direction. Obviously this would only work on platforms that allow hedging. If that 1st long trade had ended up in profit at the reversal, then it would just be closed as per normal and the short would have been the start of a new sequence.

Now if the short 1x was also a loss, then I would go long 2x as per the fib sequence, and I would keep the previous long 1x and short 1x open. So basically I would be nett 1x long. I would then just wait for the next sell (reversal) signal and do some maths. If at this time the sequence was in an overall profit, then that would be the end of that sequence and a new one would begin on the this new short. If it was still at a loss, then the short would be 3x, and leaving all other trades open, so overall I would be nett 1x short. Then just wait for the next reversal, do the maths, and if in profit, close the sequence and start again. If an overall loss, continue with a long 5x, keeping all other trades open so I would be nett 4x long, and so on. I would also add that if the sequence ended up at break even overall, then that would be enough to close it.

Could also possibly look at hybrid system of the one mentioned above and the original concept, where if I have multiple trades open in both directions as above, instead of waiting for a new reversal signal, just close the sequence whenever it reaches an overall break even point and start a new sequence at that price.

This is just a new idea I was thinking about and has nothing to do with the EA I am currently building. It will need some further testing to see if it is even viable. I have had a quick play with some charts and it doesn't look too bad. So why would I look at this? As already stated, it will only work if your broker allows you to hedge. This would reduce your margin requirements somewhat due to this hedging. For example if you were on the 7th trade of the fib sequence using the original method, this would be for 13x. However using this method, even though I would still have a 13x trade in play, because of all the other open trades, I would only be open nett 9x. Remember the sequence is 1,1,2,3,5,8,13 so every 2nd trade is in the same direction. Also not sure about this one, but it is possible it may produce more profit if I go from signal to signal to reverse, instead of closing out immediately I hit break even, due to the fact that I will be continuing on with a larger position.

This is just a brief overview and will require a lot more research on my part to see if it is even viable over the longer term. This I will do over the next few days and get back to you with some further info. In the meantime, I would appreciate any feedback, and if you think it sucks, feel free to tell me and the reasons why as I may have missed something. Cheers. JIm

Trade update:

EUR/JPY is now short 1x at 126.42. This is the 2nd trade in the sequence and there was a loss of -70 on the previous long. Now have an overall break even target of 125.72. No other changes. Cheers. Jim