Exceptional day, will it continue ? | Page 4 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

The Cable double day / the two day "rule" 14 replies

Exceptional or Normal Results? 27 replies

A Day Traders Day? 7 replies

Day by Day 9 replies

Attachments: Exceptional day, will it continue ?
Exit Attachments

Exceptional day, will it continue ?

  • #61
  • Jul 15, 2009 3:11am Jul 15, 2009 3:11am
Hi all,

Yesterday was an extremely productive day as far as the coding is concerned. I coded for a solid 16 hours and made a breakthrough which I think will make some coders on this forum very exited.

All multiple time frame indicators always had one inherent problem and that was the fact that they repainted to a certain extend. The reason for this is because you have candles closing over various lengths of time and an indicator showing values for shorter periods of time i.e. on a 15 minute chart the indicator will show the hourly value every 15 minutes even though the hour candle has not closed yet. Should the hour candle change value 1 second before it closes it meant that the last 4 bars on the 15 minute indicator would change thus in effect repainting the past.

The breakthrough I made last night resolved this whole issue and provides a more accurate measure of price movement over time. I can now guarantee that there will be no repainting on this indicator. Not even if you go into the history and delete some candles.

The EA as it stands now processes just under 40 million (if pushed to max history) values per tick. Don’t believe me: DTM – Do The Math: 100,000 candles (on 1 min this is around 3 months - the max for the indi) x 11 conditions x 4 values per condition x 9 time frames = 39.6 million values. And these are crunched every tick. Realistically however only about 1% is crunched per tick.

What else has been done?

As you know I prefer naked charts. If there is one thing I hate it is chart clutter. So to resolve this I have coded the EA to hide everything i.e. pivots, fibs, take profit lines etc.. When loaded onto a chart you will see nothing, duh that helps…

You then enter a pip value example 40. When ever price is 40 pips above or below a certain point i.e. a pivot line, it will alert you and only then show the line as long as price remains within the +40/-40 area. As soon as price moves out of this area the line will be hidden away again. In that way you have all you want on the chart, just without the clutter. You can of course turn everything on and off manually as well.

This enables me to add more to the EA like normal pivots, camarilla pivots, fib pivots etc… you just pick your flavour and away you go. It does however make the settings section of the EA a bit cumbersome but I will change this to load from a file as well saving you all the hassles. Just need to set it up once and as the saying goes, Bob’s your uncle.

That’s it for me for the day, off to the airport to welcome my wife and son back from their short holiday.

PS: Regarding the signals, I did not run signals for the whole of yesterday as I was coding inside the signal module. I can thus not say whether it was a good day or not. From what I have seen I think the number of losses might have been a bit higher with the ranging that went on but that the overall outcome would still have been far into the green.
  • #62
  • Jul 15, 2009 3:24am Jul 15, 2009 3:24am
Sounds great Mike. Like the fact you can turn things on and off and as you say the clearer the chart the better.
Boxing clever since 76.
  • #63
  • Jul 15, 2009 3:34am Jul 15, 2009 3:34am
Greetings Mike

I am exited to hear what is brewing, even though I am not much of a coder

kos
John 3:16 - From THIS love no power on earth can me sever.
  • #64
  • Jul 15, 2009 3:49am Jul 15, 2009 3:49am
Sounds like it will be a nice bit of kit Mike. Although I am very sceptical about EA's over manual trading, I think it will benefit many traders who are still unsure about how to enter around SR levels and will reduce the emotional effect of freezing up when it comes to pulling the trigger, when traded manually. And of course being able to trade multiple pairs without having to have multiple pairs of eyes. Great stuff.
  • #65
  • Jul 15, 2009 4:03am Jul 15, 2009 4:03am
Just had to post this before I leave.

As a more of a price action trader I have never really relied on Expert Advisors to do any trading for me, they just cannot read a chart the way a human does. I do however believe that they can provide great assistance in trading i.e. order management.

As the framework of the indicator I’m rewriting now is inside an EA I have decided to add all the normal trading and money management functions to it as well. This morning I enabled it when I got out of bed to see whether it could trade and to watch for errors. Went for a shower and when I came back still nothing.

I just opened the history and to my amazement there was actually a trade. While I was showering, having breakfast, cleaning my koi dam filters etc… the ea went on with its business. It took a buy on EU (only chart that was open) at 1.3994 and closed it out just below TP3 on 1.4054. A nice 110 pips to kick start the day . Unfortunately it was on my demo account .

So the auto trading part shows promise as well. At this rate it seems like everyday will be an exceptional day.

Cheers
  • #66
  • Jul 15, 2009 4:13am Jul 15, 2009 4:13am
Quoting mike_85
Just had to post this before I leave....
Really looking forward to this.

Keep up the good work Mike ...
  • #67
  • Jul 15, 2009 4:59am Jul 15, 2009 4:59am
Quoting mike_85
Just had to post this before I leave.

.

So the auto trading part shows promise as well. At this rate it seems like everyday will be an exceptional day.

Cheers
Amen. Well done Mike.
  • #68
  • Jul 15, 2009 5:34am Jul 15, 2009 5:34am
Quoting mike_85
The EA as it stands now processes just under 40 million (if pushed to max history) values per tick. Don’t believe me: DTM – Do The Math: 100,000 candles (on 1 min this is around 3 months - the max for the indi) x 11 conditions x 4 values per condition x 9 time frames = 39.6 million values. And these are crunched every tick. Realistically however only about 1% is crunched per tick.
Firstly, is it necessary to process this many values every tick? Surely you would not need to process that many values after the initial application of the EA/indicator as historical values would already have been processed at first run, no?

Secondly, what kind of performance are you getting from the product? If the robot/indi is crunching that many numbers every tick, then what is your recommended rig setup to avoid performance degredation of both the system and the EA?

Quote
This enables me to add more to the EA like normal pivots, camarilla pivots, fib pivots etc… you just pick your flavour and away you go.

So your 'walls' or SR is based on pivots or fibs? Would you also consider adding SR based on fractals figured on candle highs and lows and candle closes?

Regardless, well done on producing an alpha stage product. I look forward to your responses and the beta tester.
  • #69
  • Jul 15, 2009 5:58am Jul 15, 2009 5:58am
Quoting Porkpie
entering the market in and around SR levels produces high probability trades compared to entry using indicators such as stochastics and MacD, MAs etc, as all of these lag and will often force you to enter into SR levels in the wrong direction.
I'm no advocate of leading or following indicators and don't use them personally, but that doesn't mean that they don't work for those that know how to use them. In some instances, its the lag that you need from the indicator which keeps you in a trend. Also, choosing the correct indicator for the current market is crucial.

I might add that MA's have received bad press, probably more so than any other indicator. However, if you ignore the MA cross strategies and focus on using MA's purely as Support/Resistance, then they become a whole lot more valid and consistent, eg. Daily TF, 21 SMA (1 Forex Month), 65 SMA (Forex Quarter), 261 SMA (Forex Year).

I will also add that indicators are a support tool and should not be used as the only signal for entry without looking at price/volume over time, a fact that all successful indicator users agree on.

Price is king, no doubt.
  • #70
  • Jul 15, 2009 7:16am Jul 15, 2009 7:16am
Quoting Limstylz
I'm no advocate of leading or following indicators and don't use them personally, but that doesn't mean that they don't work for those that know how to use them. In some instances, its the lag that you need from the indicator which keeps you in a trend. Also, choosing the correct indicator for the current market is crucial.

I might add that MA's have received bad press, probably more so than any other indicator. However, if you ignore the MA cross strategies and focus on using MA's purely as Support/Resistance, then they become a whole...
The problem with oscillators, MA's etc is that they do not recognise ongoing supply and demand relationship at work every second at every price level. They are simple math calculations derived from price. Most are averages of price which means they lag price. Any indicator that lags price adds risk and decreases profit margins.

Those who use a MA as a trend filter would never buy when the trend is down as the MA tells them the trend is down. As a result you will ignore the lowest risk/highest reward opportunity to buy (or sell) each time it is offered.

Moving averages are calculated from past data. They can only turn higher after price does. If you took a 200 day MA and waited for a cross. First of all price would already have to be high to achieve a cross as you would be buying far from the supply/demand imbalance of the original upmove that caused the MA to cross. This decreases profit potential. The MA only appears to work when a moving average lines up with true demand or supply. But believing that it works as a SR tool alone is pure illusion in my opinion.
  • #71
  • Edited 8:21am Jul 15, 2009 8:05am | Edited 8:21am
Quoting Porkpie
Those who use a MA as a trend filter would never buy when the trend is down as the MA tells them the trend is down. As a result you will ignore the lowest risk/highest reward opportunity to buy (or sell) each time it is offered.
Agreed.

Quote
Moving averages are calculated from past data.

So are pivots and fibs.

Quote
If you took a 200 day MA and waited for a cross. First of all price would already have to be high to achieve a cross as you would be buying far from the supply/demand imbalance of the original upmove that caused the MA to cross.

Obviously you need to choose the relevant MA for the time frame you are trading firstly and secondly, isn't that the nature of breakouts? Many, many successful breakout threads here in FF testify to this, albeit, they may not use MA for confirmation, yet buy at breakout of a tested high or sell at the breakout of a tested low.

Quote
The MA only appears to work when a moving average lines up with true demand or supply. But believing that it works as a SR tool alone is pure illusion in my opinion.

I don't think I was clear enough in my last post. I wrote 'if you ignore the MA cross strategies and focus on using MA's purely as Support/Resistance', in this instance I should perhaps have stated that at the support and resistance provided by the MA, you would be looking for mostly bounces and only taking breakouts if you have indication from price/volume over time, as with all price action. The point about MA's is that lots of people/institutions do use them as SR. If there are enough entities using them at a given time(as with fibs, pivots and other ilk), then it becomes self fulfilling. Yes, to use them purely by themselves is a mistake, no question.

I was not disagreeing with you on any point, merely stating that to dismiss indicators out of hand just because you or I don't use them is narrow minded.
  • #72
  • Edited 9:30am Jul 15, 2009 8:09am | Edited 9:30am
Quoting Limstylz
Agreed.



I don't think I was clear enough in my last post. I wrote 'if you ignore the MA cross strategies and focus on using MA's purely as Support/Resistance', in this instance I should perhaps have stated that at the support and resistance provided by the MA, you would be looking for mostly bounces and only taking breakouts if you have indication from price/volume over time, as with all price action.

I was not disagreeing with you on any point, merely stating that to dismiss indicators out of hand just...
yes I agree, if only to know what the 'herd' are doing or use SR to filter the indicator's signal.
  • #73
  • Jul 15, 2009 9:24am Jul 15, 2009 9:24am
Limstylz, let me tackle your questions first,

a) You are correct, it is not necessary to process that many values, when the indicator starts up it does an initial run and then only update per close of 1 min. bar, so every minute. Volume of values processed then drops to around 0.5% of initial load.

b) Performance is exceptional but that is because MT is not doing the processing. It is currently still being done by my VB app. So at this stage MT throws the values, VB does the cruncing and throws results back to MTand finally MT does the drawing. I'm starting today on transfering the VB code into MT, hold thumbs...from visual basic to c++...nightmare at it's best.

c) Walls: They are not based on anything but price action. If you however add pivots and the like you will find that many of the walls do happen in these areas, that is the nature of support & resistance. So in short, no, walls as I trade them is 100% price action based. As mentioned, I trade my hammers & walls on a crispy clean chart with not one single line on it.

Quoting Limstylz
Firstly, is it necessary.....
  • #74
  • Jul 15, 2009 9:43am Jul 15, 2009 9:43am
Quoting mike_85
Walls: They are not based on anything but price action. If you however add pivots and the like you will find that many of the walls do happen in these areas, that is the nature of support & resistance. So in short, no, walls as I trade them is 100% price action based. As mentioned, I trade my hammers & walls on a crispy clean chart with not one single line on it.
You mentioned in a previous post about turning pivots on and off... In your indicator/EA if we choose to trade naked (all pivots, fibs off), how does your indicator then draw your walls? Will it draw walls based purely on PA then? If so, then could you briefly explain your code behind this?

Are you taking the close of a bar in each time frame and provided that the closes/high/lows of the bars in each time frame are say within a distance of 0-5 pips, this constitutes a wall? Are you also programmatically defining what a hammer is?

Apologies for all the questions. I am hoping that your answers will help me to complete something that I have been working on for a while.

TIA
  • #75
  • Jul 15, 2009 9:48am Jul 15, 2009 9:48am
Quoting mike_85
Here is the same chart but with the hammers and the walls added. Can you see what the hammers and the walls are?

See how easy it can be to trade a clean chart?

I'm busy putting a document together now explaining the hammers and walls in detail together with a bunch of other stuff I know you guys will appreciate.

Off to bed for me, after 2:30 am here.

Cheers
If you don't mind me asking, what time frame do you usually trade these hammers and walls off of? Just to clarify, I mean when you trade manually, I noticed you said the indicator is looking for these across all time frames, but when you do it indicator free, are you just scanning many time frames all day looking for these hammers?
  • #76
  • Jul 15, 2009 10:00am Jul 15, 2009 10:00am
Porkpie & Limstylz,

It seems we are all on the same trace but just expressing it differently so let me just clear some confusion here, if it may exist.

a) I trade price action only i.e. supply and demand, support & resistance, hammers & walls, call it what you like. I trade from a clean chart, I add nothing to my charts because if I see it I want to trade it . So with nothing there I'm forced to trade the only thing left and that is price. So first and foremost I classify myself as a naked trader.

b) So what is all this about the indicator then? As mentioned from the start; IT IS A TOOL ONLY. I trade to achieve an annul target of 3,000 pips. My daily target is 15 pips and I only trade every other day (all in all about 180 days a year). If made I'm financially fine and have no worries about the future. This however does not mean that I only want 15 pips per day, just that I do not put pressure on myself with targets like 100 pips etc..

Currently I get up in the morning and take up position in front of my notebook waiting for a trade. Sometimes I'm lucky and it's all done within an hour or so, other times I sit for hours with no trade. Time is one of the most valuable commodities around so the idea I came up was:

I come from a programming background (self taught) and programming runs in the veins just as badly as forex and alcohol. So I'm always tinkering with the one or other thing, call it my hobby. So the idea came up, I have all the programming skills required, good knowledge of the financial and forex markets, know the workings of each indicator inside out....so why not put something together that will do the following:

i) Analyse the market across all time frames and scan for certain conditions,
ii) If found alert me to pull up a chair and take note of possible trading opportunities.

With this tool in place I will be able to run around the garden and house doing other chores without worrying about the markets, I will be called if there was anything worthwhile happening.

And this is where it all started. It only went to an EA because I had better control of all the code when running it as an EA instead of an indicator (personal choice) so the fact that it is now in EA form does not mean that I’m heading down the EA road, as everything on this forum does, eventually.

I am a naked trader and for ever will be one, nothing beats price full stop. This indicator is a tool to assist with trading and not to do your trading for you. The fact that it can does not mean that it must. I will never ever take an entry provided by any indicator, not even my own, if it is not confirmed by price action.

For any new traders stumbling onto this thread take note. No indicators and or EA will do it for you if you cannot read a chart. Not being sarcastic here, I’ve been down the indicator road and saw my you know what more than once. Since going back to basics I have been earning easily ten times as much with a quarter of the effort.

Once you are able to read a chart you will see what you have missed. All of a sudden all indicators and systems start making sense in that you see them against price action. This is when you are in the zone, you will then be able to be profitable no matter what system you use, AS LONG AS you base everything still on pure price action and use the rest as assistance or as tools.

Hell, only see now how I went on, sorry for the long post I just felt that everybody needs to understand where I’m coming from and what the aim is of this indicator/EA. Please fire away with any questions and keep the pm’s coming. I will answer all.

Quickly on the way forward. I will start Monday with the hammers and walls. I will however not post it but upload it in pdf. In that way you can download and keep/delete it. That will be followed by three or four basic strategies showing how price action interlinks with ‘normal’ trading i.e. using indicators, basically how you can build a system for yourself using price action as a base. And once that is done I will release the EA for testing which I do not think is required. It works, further testing is just curve fitting and this is not required. You will see what I mean once you understand and grasp the concept of the h&w's and you look at the indicator/EA.


On another note thanks for all the interest. I really did not expect this thread to reach even page 2. (I think this post will take it to page 8 now )
  • #77
  • Jul 15, 2009 10:08am Jul 15, 2009 10:08am
Quoting mike_85
Quickly on the way forward. I will start Monday with the hammers and walls. I will however not post it but upload it in pdf. In that way you can download and keep/delete it. That will be followed by three or four basic strategies showing how price action interlinks with ‘normal’ trading i.e. using indicators, basically how you can build a system for yourself using price action as a base. And once that is done I will release the EA for testing which I do not think is required. It works, further testing is just curve fitting and this is not required....
Lovely... Will wait until Monday for your PDF before I post anything else then. Thanks for your patience... I will try to follow your virtue.
  • #78
  • Jul 15, 2009 10:12am Jul 15, 2009 10:12am
Quoting Limstylz
....
Limstyles,

Sorry would have answered after the other posts but the power went off as I typed. In any case herewith answers to your quest.

a)As mentioned I trade from a naked charts. What I do now is have one blank chart and one with pivots, fibs etc…I trade and enter on the blank then check the other for possible TP points. So the thought came up, put it all on the chart but just turn it off or make it interactive so that it comes on by itself and hides itself again.
b)Wall: Walls are not drawn by the indi. That you should be able to spot without any effort on any chart that is the basis for price based trading. Check my previous post for more details on upcoming document. That will clarify a lot of information on the h&w’s.
c)Hammer: search for hammers, hanging man, inverted hammer. There are candle stick formations.

You won’t know until you ask so fire away.

Is it just me or am I confusing the hell out of everybody...lol...

Let me just clarify, it will all be in Monday's doc. I trade naked but have 2 charts? Yeah yeah....but it will become clear on Monday why. I do trade naked i.e. enter etc. I can also manage my trade on a naked chart. BUT..I trade for 15 pips a day, so I enter naked, make my 15 plus...get my stop up to BE or more, then flip the chart, check for possible TP's, enter this on the trade and walk away.

I made my 15 for the day, if it goes another 200 then I get a bonus, otherwise I made my target for the day, thank you. So the pivots etc... came in to give my trades a longer run than to just take profit at 15 and get out. Hope this clarifies some.
  • #79
  • Jul 15, 2009 10:22am Jul 15, 2009 10:22am
Quoting mike_85
Hammer: search for hammers, hanging man, inverted hammer. There are candle stick formations.
Yes, I am aware of candlestick patterns and formations... I was asking you if your CODE searches for hammers and if so, how have you PROGRAMMATICALLY defined these? Same goes for the walls. I can visually see Support and Resistance, but how do you CODE this in your indicator?

I have been working on this problem myself for the past 5 months so I don't need to sit in front of the machine all the time, and the closest I have come to defining Support/Resistance PROGRAMMATICALLY is with MTF fractals with a deviation of between 0-5 pips. However, because of the nature of fractals, it either lags or repaints when you use MTF.
  • #80
  • Jul 15, 2009 10:26am Jul 15, 2009 10:26am
Hey Mike no need to thank us for the interest, we should be thanking you for sharing it wiht us! so by monday you think we will be able to give it ago and try it out? thanks in advance.
1 trader viewing now