RE: [margrids-battle-chest] Re: 3.5%

[dan...@danielalexander.info]

 

I have been continuing to use Mark's file with some manual intervention, lately switching from Euro to Aus.

 

Am now closed down for the week for 9.11% profit over the last 12 trading days. That's within my 3% weekly target.

 

I have to admit that the Chameleon would have reduced the amount of manual control needed.

 

Rick, how close are you to circulating this version?

 

Best regards

Daniel

[Rick Drake]

Excellent Daniel. I think this whole weekly target is an excellent idea! It's a great way to break up developing trends as the system is always being restarted each week! That is one feature I'm going to implement in Chameleon... a weekly target that when reached will stop opening new baskets until the next week begins.

I'm getting the groundwork laid for Chameleon. I've figured out how to communicate between the DLL and Metatrader, and set up the importing of all the various data from Metatrader. Now I'm working on the system that will do the work of opening up trades and managing them in data structures. Once I'm done with that, I'll be ready to code the actual logic. I plan on implementing a simple Martingale first... and get it out to the group for bug testing while I continue implementing the Anti-Martingale features. 

Each step along the way, I am verifying that my code works as expected, which is slowing me down a bit but in the end will result in a system with much fewer bugs. I have a long weekend coming up, and my wife is going to visit the in-laws for a week and a half, which will give me lots and lots of time to get the system coded. So I'm hoping to have the basics for Chameleon implemented in a couple weeks. Then I'll work on summarizing each basket in the log files. Then after that I'll work on getting the front end working for Forex Tester so people can practice in the past with manually intervening when necessary, as well as being able to run a whole portfolio while backtesting.

- Rick

--
You received this message because you are subscribed to the Google Groups "Margrids Battle Chest" group.
To unsubscribe from this group and stop receiving emails from it, send an email to margrids-battle-...@googlegroups.com.
To post to this group, send email to margrids-b...@googlegroups.com.
Visit this group at http://groups.google.com/group/margrids-battle-chest.
For more options, visit https://groups.google.com/groups/opt_out.
 
 

[dan...@danielalexander.info]

 

That sounds brilliant, Rick.

 

Feel free to post any interim versions that you need us to test!

 

Best regards

Daniel 

[Rick Drake]

I will. The idea is that Chameleon will be released and tested in stages so that we know it is robust and error free... I'm following more of the Extreme Programming mantra. I figure... this thing needs to be bullet proof and error free. It's challenging enough to successfully run a Martingale... it would be horrible to lose good money because of programming errors. Of course, there will be bugs. But hopefully they can be kept to a minimum, and be discovered during the testing phase so that when people start using real money, there are no surprises!

- Rick

[Keith Krazee]

+1 Daniel.  I've been just lurking and reading lately but wanted to pop in and say I really like the idea of running a riskier system like a margrid in a very controlled manner.  Start and stop it frequently and avoid news etc at all costs.  When I first started grid trading I was thinking 'sky's the limit, I'll run it and withdraw until it blows, hopefully after 100% ROI'.  But I like this way much better now.  Set a reasonable goal and stick to it.  3% weekly is still >10% per month with weekly compounding.  If you can sustain it for a couple of years, you are on your way to riches!   

Just a quick example, 3% compounded weekly for 50 out of 52 weeks a year gives 438% ROI (assuming 2 full weeks of no trading per year for vacation etc).  After 2 years of this it is almost 2000% ROI.  : D  This compounding is always the dream, the trick is how to get it without a margin call!!

[Keith Krazee]

p.s. in live trading, how fast on average are you able to shut down trading?   2-3 hours or does it usually take longer?

[EOGORFX]

Rick

This is great news and you approach to realizing a solid and safe algo trader is A1.

As you say running an MG is scary.

For me anyhow every time a basket starts to go beyond 10 levels or so I think will it ever stop growing and since I don’t have massive capital to cover a very large open positions it seems that the risk far outweighs the reward. I am even slow to put in large capital in case I might blow it all away.

I have seen my open position go >> 1000 and my eventual close for < 100 and the angst of hoping for that retrace makes it difficult to stick with the system.

We need a sure fire way of controlling DD and capturing and retaining profits from all that wave energy in the FX seas which can turn into violent storms without warning.

I will do all that I can to help with testing your betas.

All the best

Ed

[dan...@danielalexander.info]

 

That's right Keith, on a file like Mark's it usually takes a few hours to close, on average.

 

Specifically if you have a breakout and want to follow the trend for the duration, it can take longer.

 

The only way to shut down in very rapid time is to have a file such as EP=3, TP=1 for every level doubling the lot size for each level (if you put that into my excel you will see extremely tight retraces occurring at each level, significantly less than 5 pips all the way). Running this on a demo only on USD/CAD I made 4% today and shut down took 3 minutes! Survived all the news. DD mostly below 0.2% the whole time. The only problem is that a max basket will easily take that 4% away from you, so I am sure we would breakeven or worse if we simply turn on and let it run the entire time. I am looking into running for a discrete period of time, and closing some of the higher level baskets for some loss here and there, which can then be rapidly made up in ensuing trading. I did this a couple of times today, removing level 6s where the level 6 trade was going into some profit but market was struggling to give me more. Bad news is you do have to literally stare at the screen the WHOLE time in order to cut out some DD, depending on how you view the market. But it's always satisfying to get rid of a not too big of a DD loss and then see the trading restart at level 1s. It's harder to run on more than one pair because you will need to "massage your FX chart sensitively" at the just right moment before a loss gets too big, if you feel you might be headed into too much DD. But with a rapidly increasing equity tracking the increasing balance for the whole the day, but for one or two potentially unlucky moments, this does offer a way to trade provided that you have some babysitting hours available. The Chameleon could also help on this system too, by partially hedging at times on the AM side, again this would have to be manually controlled in addition to the EA operation.

 

It might be fun to have people here actually run this file for themselves, on demo or otherwise. For just one hour, say, each day Monday to Friday. As a project! Turn on....watch for one hour.....then turn off. Then we can collect people's results and experiences. If there is interest I can also email the set file I am referring to here.

 

Daniel

 

-------- Original Message --------
Subject: Re: [margrids-battle-chest] Re: 3.5%

From: Keith Krazee <krazeee...@gmail.com>
Date: Wed, July 31, 2013 8:28 pm
To: margrids-b...@googlegroups.com

p.s. in live trading, how fast on average are you able to shut down trading?   2-3 hours or does it usually take longer?

On Wed, Jul 31, 2013 at 3:26 PM, Keith Krazee <krazeee...@gmail.com> wrote:

+1 Daniel.  I've been just lurking and reading lately but wanted to pop in and say I really like the idea of running a riskier system like a margrid in a very controlled manner.  Start and stop it frequently and avoid news etc at all costs.  When I first started grid trading I was thinking 'sky's the limit, I'll run it and withdraw until it blows, hopefully after 100% ROI'.  But I like this way much better now.  Set a reasonable goal and stick to it.  3% weekly is still >10% per month with weekly compounding.  If you can sustain it for a couple of years, you are on your way to riches!   

Just a quick example, 3% compounded weekly for 50 out of 52 weeks a year gives 438% ROI (assuming 2 full weeks of no trading per year for vacation etc).  After 2 years of this it is almost 2000% ROI.  : D  This compounding is always the dream, the trick is how to get it without a margin call!!

On Wed, Jul 31, 2013 at 10:44 AM, Rick Drake <drak...@gmail.com> wrote:

[Rick Drake]

Daniel,

I think that would be an awesome experiment! I can't really participate because I'm using my time to code... but other's could. What a concept though... to be able to run the system and gain 3% a week with minimal intervention... or 3% a day by baby sitting for a few hours. What trading session would probably work the best? Asian? I wonder if the trends during the New York trading session would be difficult to deal with... of course there are often reversals to the trend at the end of the day... so perhaps the last 3 hours of the New York trading session would work well.

- Rick

[EOGORFX]

Daniel

Please post the set file as I would like to see how it runs on a demo.

How about trying it with the close at max basket = true! Where max maybe level 5  or 6 say.

I use this sometimes when I have to quit the screen and the market is looking wild.

Perhaps I should use freeze instead.

I am out of the market this week as it feel like the only safe place to be.

 

Another issue I have noticed is if you place Chicken out  = true it can be very slow (several minutes) to close out all the trades. I have closed some manually while waiting for the CO to do it duty.

It seems so tardy compared to when a basket closes normally.

 

Thank you

Ed

[protokop]

Hi Daniel,

I have often wondered if a strategy like this could be worthwhile.

How much would you use in the account? Obviously the % daily/weekly return is dependant on that.

If you are happy to share the set file I am happy to test and share results. :)

Kev

[Maria M]

Hi Rick/everyone

I've managed to continue where I left off 2 months ago with backtesting set files with pipspan to 800 and 900. What is repeatedly being confirmed to me that even with such large pipspan and low DD a pair will get into strife always at times of major news events.

Rick if I would like to make some suggestions/requests if I may:

1) With server news set to true, could we have the ea freeze all trading not just once a basket is closed? This would prevent large baskets/DD from happening on existing open baskets and trading can be continued under normal trading conditions.

2)  Could you also pls include Interest rate statements news releases for all pairs? Not sure this is happening for AUD and NZD, CAD?

3)  Could you also pls include Retail Sales news releases for gbp, cad, usd? Add any other pair you feel is also at risk.

4)  Is it possible to have mbc close at TP when a hedge is in place pls? I would like to start forward testing sliding hedge with this feature it would close baskets quicker than current closing at SL.

The abillity to switch trade direction at trend/support/resistance lines will add major risk protection even though it would require once daily monitoring which is not a problem. Therefore I think this is a biggie!

With the 3% pw shut down, I wonder how you could bactest that to check that you're not losing all your profit?

It's all very exciting developments which we are all so much looking forward to.

Thanks for all your hard work.

Maria

 

It's not the strongest of species that survive, nor the most intelligent, but the ones who are the most responsive to change - Charles Darwin

 

As described by the Saints and Sages that came before, this is the time for Self Realisation. Your time is now! Go to http://www.sahajayoga.com.au/level_1/meditation.shtml click on experience it now.

---

From: Rick Drake <drak...@gmail.com>
To: margrids-b...@googlegroups.com
Sent: Thursday, 1 August 2013 7:11 AM

[Eogorfx]

Rick

I believe the M grid should be dynamic or should have that option over fixed.

Dynamic = when price is moving away from the basket set a buy or sell stop behind the price which will only fill when a pullback occurs. This order will be auto adjusted to remain X pips behind the price. X is a user input. The size of the order could also be auto adjusted to close the basket at a fixed profit also set by user. This will stop a large number of levels opening when a big pip movement occurs. It will reduce DD and make basket closing much quicker and less costly as there is less profit required to pay off the lower levels.
In parallel the AM side is catching the BO and locking I some profit! 
Ed

[Gosta Hulden]

Hi all, 

My apologies if this has been discussed before, but could someone explain the benefit of a weekly profit target. I can understand a maximum weekly loss limit, but profit target? If for example I manage to do my 3 percent before end of Wednesday, then - if anything at all - it tells me that the current market conditions are very beneficial for my strategy, so why stop trading?

I have heard people say that you don't wanna give back the earned profit to the market, but the risk for doing that should be just as high the following Monday as I see it. 

And if it has to do with stopping the baskets before the weekend then it should be done no matter what the result is, shouldn't it?

As I see it, this strategy seems to be; 

If the market conditions are great, then stop trading. 

But if the market conditions are not so good, then keep on trading. 

How can this be a good idea? What is it that I am not seeing?

Thanks

Gosta

Sent from my iPhone

[krazeee...@gmail.com]

The advantage is that it controls your greed on a system (grid martingale) that can easily MC your account in the wrong market conditions. The idea is to just slip the trading in during intraday or intra-week periods that are most suitable for the margrid to make money with low risk of a trend forming. It still won't save you from sudden, sharp and unexpected market moves. It does however save you from long, slow forming trends that still end up exceeding your setfile pip span.

Sent from my BlackBerry device on the Rogers Wireless Network

---

From: Gosta Hulden <go...@hitparkmusic.com>

Sender: margrids-b...@googlegroups.com

Date: Thu, 1 Aug 2013 12:26:06 +0100

To: margrids-b...@googlegroups.com<margrids-b...@googlegroups.com>

ReplyTo: margrids-b...@googlegroups.com

Cc: margrids-b...@googlegroups.com<margrids-b...@googlegroups.com>

[Gosta Hulden]

Anybody else who has something to add about the benefits of a weekly profit target? Cause I can't say I got it.

I am all for restricting the trading to what you believe is the best trading conditions, but what the profit target rule says is rather the opposite. "Now the conditions seems to be great, so stop trading".

I also can't see how a profit target rule can save me from "long, slow forming trends that end up exceeding my set file pip span". If the trend is against me I will never reach my profit target, right?

I'm not trying to be difficult, just trying to understand this rule that many of you seem to like and that seems all backwards to me. So my question remains, if I have reached my target by Wed evening for example, why stop trading?

Thanks

Gosta

[krazeee...@gmail.com]

No problem, I'll answer more what I'm thinking. First of all it's not a golden rule, you might be right and it's not the best way. Have you traded grid martingales much? They tend to blow up pretty quick in the wrong conditions and you lose your whole account. You can see a nice steady gain for months even a year or two then within a few hours or a few days it's all gone. So that's why you want to stop the trading and lock some profits before the losses show up.

Btw in case there's any confusion my original comment about this related to martingle, not anti-martingale mode (which has its own dangers). So when you say you can't see how it makes money going agaist the trend, exactly how a martingale works is to build baskets against the trend until a reversal occurs. If the reversal does not occur (ie a trend has formed) before your pip span is exceeded then your account is gone.

This auto hedging in a grid martingale is pretty new for commercial margrids so personally I'm still being cautious from my older martingale days of trading. Once I get more history, if it seems really safe in live trading I may be more inclined to just let it run and compound for a long time.

But I'm with you, let's hear what some others think on it because I'm no expert, just someone who's experienced nice gains and also catastrophic losses with martingales before.

Sent from my BlackBerry device on the Rogers Wireless Network

---

From: Gosta Hulden <go...@hitparkmusic.com>

Sender: margrids-b...@googlegroups.com

Date: Fri, 2 Aug 2013 23:16:08 +0200

To: <margrids-b...@googlegroups.com>

ReplyTo: margrids-b...@googlegroups.com

[dan...@danielalexander.info]

 

Put another way, it is never a good time to trade a martingale.

 

Therefore the less market time exposure you give it, the better. Better to hit 3% in short bursts in less market time, than 10% running all week and subjecting your account to all manner of nasty.

 

The test I just circulated attempts to do 2 things:

1) Limit the risk of hitting a bad move by trading for less than 5% of available market time

2) Limit DD by hedging at 3.5%DD initially

3) Allow you to actually trade a market knowing that you risk 3.5% per switch on, and are likely to obtain x% in return for that risk.

 

This then starts to feel like real trading because you have your risk:reward set up from the very beginning. The trick is then to trade at times where you feel most comfortable - and that's your trading strategy - instead of trading breakouts with TP and SL - you're trading a market you think should retrace at certain price points. Same trading mentality as a pro trader, just different system.

[Rick Drake]

Daniel,

You put it very well when you state that it's never a good time to trade a martingale. As I look at it, the more market exposure you have, the more likely you are to blow the account. If the system trades 100% of the time... you have a 100% chance of blowing your account eventually. By limiting when the system trades, the you might miss the times that would have blown your account. Of course, the flip side is that you might not miss blowing your account... and you could have made a lot more by just letting the system run. So it has some give and takes. As I see it... it's an excellent way to break up the baskets to help prevent them from maxing out.

- Rick