No need to worry about the coding thingy in my last post, guys. I have received a couple of pm's from whf2006 with solutions. I will post them later.
Hanover: an auto-trader using Recent Strength
- #302
- Nov 4, 2011 5:57pm Nov 4, 2011 5:57pm
- | Commercial User | Joined Apr 2007 | 8,331 Posts
Thanks to whf2006, the attached appears to work with the correct data. There might yet be the odd bug lurking in the undergrowth, but the overall basics are in place.
Guys, following another coders' logic is really hard, so a warm round of applause to whf2006.
I attach the EA here for the backtesters to play with over the weekend. Next week, we need to work out what are the best settings for TableSize and Threshold. Once we get these right in StuffTheNews, we have a module I can isolate for us all to use in all our future EA's.
Unbelievable stuff we are coming up with here guys. I often remark to my lot when one of them manages to string a few notes together coherently, "See? All it takes is gob-smacking amounts of talent and it is easy."
For sure, we have gob-smacking amounts of talent contributing here.
Guys, following another coders' logic is really hard, so a warm round of applause to whf2006.
I attach the EA here for the backtesters to play with over the weekend. Next week, we need to work out what are the best settings for TableSize and Threshold. Once we get these right in StuffTheNews, we have a module I can isolate for us all to use in all our future EA's.
Unbelievable stuff we are coming up with here guys. I often remark to my lot when one of them manages to string a few notes together coherently, "See? All it takes is gob-smacking amounts of talent and it is easy."
For sure, we have gob-smacking amounts of talent contributing here.
- #303
- Edited 12:24pm Nov 5, 2011 12:13pm | Edited 12:24pm
- | Commercial User | Joined Apr 2007 | 8,331 Posts
I have had a lot of fun in the last couple of hours.
Here is a pm I received from wfh2006 earlier today:
"Steve,
When the CBI conversation started a day or two ago, I put together an indicator that calculates the average (SMMA) of the last x ticks and displays it on the chart as a horizontal line. I also included a line above and below the average line as a threshold of sorts. The lines adjust with each incoming tick.
I watched it for a while, and boy is it boring to watch the average of tick data!
The idea would be to compare current price to the average of the last x ticks instead of comparing the current price to the price x ticks ago. Would the smoothing make much difference, probably not, but I thought I'd toss it it out anyway. If nothing else, it demonstrates how to calculate a moving average without keeping a buffer of the data you are averaging.
BTW, the indicator is really boring when the market is closed, as it shows nothing at all until after it receives enough ticks to show the average.
Here's the code: "
I have subsumed the code into the EA. We now have two methods of spotting a breakout, named after the contributors who inspired their creation. I list them along with any inputs specific to them:
Here is a pm I received from wfh2006 earlier today:
"Steve,
When the CBI conversation started a day or two ago, I put together an indicator that calculates the average (SMMA) of the last x ticks and displays it on the chart as a horizontal line. I also included a line above and below the average line as a threshold of sorts. The lines adjust with each incoming tick.
I watched it for a while, and boy is it boring to watch the average of tick data!
The idea would be to compare current price to the average of the last x ticks instead of comparing the current price to the price x ticks ago. Would the smoothing make much difference, probably not, but I thought I'd toss it it out anyway. If nothing else, it demonstrates how to calculate a moving average without keeping a buffer of the data you are averaging.
BTW, the indicator is really boring when the market is closed, as it shows nothing at all until after it receives enough ticks to show the average.
Here's the code: "
I have subsumed the code into the EA. We now have two methods of spotting a breakout, named after the contributors who inspired their creation. I list them along with any inputs specific to them:
- UseAstral: the original table tick method. To the best of my knowledge, astral has never posted his real name here, so I use his FF name.
- TableSize: you are familiar with this
- UseGeorge: wfh200's real name is George.
- Per is the number of ticks to calculate the moving average of the Bid.
- Channel_Delta is the distance above and below the moving average of the bid to draw the lines beyond which the bot will start CBI trading.
Other inputs are common to both methods.
You can create George's lines even though the crims are not generating any ticks by using the attached tick generator. Most of you have this by now but I attach it for newcomers.
The quickest and easiest way to backtest the bot is to set the 'From' box to 31st October and the 'To' box to today on the M1 GJ chart; you will quickly catch the BOJ intervention. It is quite good fun to do this visually.
I fixed some bloops in the code that caused the bot to immediately hedge a trade it had just sent, so I think all the kinks are worked out here.
There is still a bit of a problem with the Jumping Stop feature. It only works once the stop has moved to breakeven, so for now use the breakeven feature as well. It is quite entertaining to send a zero tp, and use the JS to manage the trade.
Here is what we do guys. In time for market opening tomorrow, set up a demo and drag the bot onto representative charts that in total involve all of the currencies. Heck, it probably doesn't matter if you drag the thing onto every chart you can find.
Then sit back and let the week unfold and allow some decent news releases/CBI's to happen and see if the bot is actually profitable.
Should be a blast. We shall see.
- #304
- Nov 5, 2011 5:39pm Nov 5, 2011 5:39pm
- | Commercial User | Joined Apr 2007 | 8,331 Posts
In idle moments, I like to read back over my threads to see what I have forgotten. For sure, I had forgotten this post here: http://www.forexfactory.com/showthre...30#post5064830
So, sharpen your pencils and start writing down your fav M15 trading systems that can be automated and would benefit from Hanovering. Hang back from actually posting for the time being - let's get the CBI module running properly first.
Soon, I shall ask you to contribute your trading systems.
So, sharpen your pencils and start writing down your fav M15 trading systems that can be automated and would benefit from Hanovering. Hang back from actually posting for the time being - let's get the CBI module running properly first.
Soon, I shall ask you to contribute your trading systems.
- #306
- Nov 5, 2011 5:57pm Nov 5, 2011 5:57pm
- | Commercial User | Joined Apr 2007 | 8,331 Posts
wow, Steve - burning the mightnight oil on Staurday night...? Thanks for your continous coding efforts!
So, why not do some coding that I will have to unravel tomorrow in light of the ai?
Time to catch up on the cartoons thread, I think, to stop me doing anything dangerous.
In idle moments, I like to read back over my threads to see what I have forgotten. For sure, I had forgotten this post here: http://www.forexfactory.com/showthre...30#post5064830
So, sharpen your pencils and start writing down your fav M15 trading systems that can be automated and would benefit from Hanovering. Hang back from actually posting for the time being - let's get the CBI module running properly first.
Soon, I shall ask you to contribute your trading systems.
- #308
- Nov 6, 2011 6:08am Nov 6, 2011 6:08am
- | Commercial User | Joined Apr 2007 | 8,331 Posts
I decided that the CBI stuff merits its owm thread. http://www.forexfactory.com/showthre...97#post5106097
There is nothing new there; just letting you all know that I shall be posting further developments there to avoid confusion here.
There is nothing new there; just letting you all know that I shall be posting further developments there to avoid confusion here.
- #309
- Nov 6, 2011 4:55pm Nov 6, 2011 4:55pm
- | Commercial User | Joined Apr 2007 | 8,331 Posts
I set up Carsten's latest WatchAndCopy version tonight, then closed it down to see if it would 'remember' my inputs on restart.
It worked perfectly.
Vast round of applause to Carsten. You have the warmest appreciation of us all.
It worked perfectly.
Vast round of applause to Carsten.
You have the warmest appreciation of us all. - #310
- Nov 6, 2011 4:59pm Nov 6, 2011 4:59pm
- | Commercial User | Joined Apr 2007 | 8,331 Posts
I had an idea earlier.
We have the MA Cross using Hanover to guide our trading in the direction of the overall RS of the currencies involved in the pair.
Time to try a currency cross. For example, eur crosses usd downwards, so we sell eu.
Any comments/suggestions before I code this?
We have the MA Cross using Hanover to guide our trading in the direction of the overall RS of the currencies involved in the pair.
Time to try a currency cross. For example, eur crosses usd downwards, so we sell eu.
Any comments/suggestions before I code this?
- #312
- Nov 6, 2011 5:05pm Nov 6, 2011 5:05pm
- | Commercial User | Joined Apr 2007 | 8,331 Posts
Yay! It's working! I'm useful. *dance*
By the by, I put the wrong smilie in the post. Proper one up there now.
I had an idea earlier.
We have the MA Cross using Hanover to guide our trading in the direction of the overall RS of the currencies involved in the pair.
Time to try a currency cross. For example, eur crosses usd downwards, so we sell eu.
Any comments/suggestions before I code this?
I am currently checking the gradient to be > or < a threshold value before entering the trade. Up to now the ea has been so so, i am hoping the gradient check will improve the winners over the losers.
This should give us more opportunities for trades that prove to have legs hopefully.
Mike
- #314
- Edited 7:22am Nov 7, 2011 6:22am | Edited 7:22am
- | Commercial User | Joined Apr 2007 | 8,331 Posts
Here is the first version. I thought I would post it here first and will add it to the post 1 zip when it is complete.
The cross-spotting code is at the top of void LookForTradingOpportunities().
strength1 is the first currency in the pair, whilst prevstrength1 is the same pair 1 period ago.
strength2 and its parter are the second currency in the pair.
I have included a gradient check to ensure that the strong currency is rising and the weak falling. Without the gradient check, it is possible for currencies to be going in the same direction but still cross because, for example, one is weakening more quickly than the other. Mike is finding this not to be a good thingy, so regard this bot as highly experimental.
The cross-spotting code is at the top of void LookForTradingOpportunities().
strength1 is the first currency in the pair, whilst prevstrength1 is the same pair 1 period ago.
strength2 and its parter are the second currency in the pair.
I have included a gradient check to ensure that the strong currency is rising and the weak falling. Without the gradient check, it is possible for currencies to be going in the same direction but still cross because, for example, one is weakening more quickly than the other. Mike is finding this not to be a good thingy, so regard this bot as highly experimental.
- #315
- Nov 7, 2011 6:53am Nov 7, 2011 6:53am
- | Commercial User | Joined Apr 2007 | 8,331 Posts
If you happened to download the latest bot before the time on this post, then go back and re-download, as I have added the gradient check.
- #317
- Nov 7, 2011 10:28pm Nov 7, 2011 10:28pm
- | Commercial User | Joined Apr 2007 | 8,331 Posts
Steve,
I am not sure if you could add the module to the one of the Basket EA you created. Below is the link.
http://www.forexfactory.com/showthread.php?t=263548
Thank you.
Edd
I have no plans to do so.
I have found that TF M5M30 is very effective for last two weeks. I have set SlopeConfirmationCandles 2,2 and TP 50 SL 30. Some of this trades was also with Jumping TS (so I should correct my statistical calculations (but later)
Winners 75 % Loses 25 % . I have experienced almost all kind of systems and vast quantity of EA. At present time I have seen enough trades and trading system to be impressed for these data with so simple system.
Obviously, experienced traders inmediatly recognize that this TP and SL is just a guess, mainly because H. lock them for each currency pair (I hope it is correctly expressed in English, unless mid correctly). In other words TP and SL are always the same and it is not possible to change them when the trade oportunity is at work.
Now one demand: it could be fantastic if you include in your Hanover code (pure Hanover EA) an ATR TP and SL. I think that should improves unless the ratio pips per trade due to the adaptative behavior of ATR to each pair.
Thank you for your valuable time that generously
share with us.
Regards
