stevehopwoodforex.com
https://www.stevehopwoodforex.com/phpBB3/
Print view

Shelley's multi-pair basket trader
https://www.stevehopwoodforex.com/phpBB3/viewtopic.php?t=583		 Page 51 of 87
Author:  Jemook [ Thu May 31, 2012 3:21 pm ]
Post subject:  Re: Shelley's multi-pair basket trader
feucht wrote:How about looking at each pair that would be in the basket, reading the top and bottom of the TMA true band at that point and getting a percentage value of where the current price is. So a potential sell just above the midpoint between the bands would be a 0.55 (and all reversed for a potential buy). You could average across all trades and set a threshold before the basket would trigger, or only include pairs above the threshold.

Just bluesky thinking, excuse me if this is impossible/impractical to code...
Sounds smart. I was also thinking waiting for each pair in the basket to be within X pips of opposite 1H TMA band before triggering.
Author:  John [ Thu May 31, 2012 3:32 pm ]
Post subject:  Re: Shelley's multi-pair basket trader
Ahem...
I respectfully disagree with the notion that the entries need to be tweaked with regard to the Shelley EA/strategy. Before you flame me, just hear me out. (I hope Shelley wouldn't mind pitching her two cents in on this, as she has observed the market behavior as it relates to this far longer than I have.)

That said...

We are trading at the beginning of each session when volume, liquidity and hence, movement is at it's peak, right? The idea is, we are scalping... yes, SCALPING... across many pairs that are in a strong trend, trying to accumulate many pips whilst only shooting for a few on a per pair basis. Thus, I'm not sure we need an optimal entry, as has been suggested. In the past 6 sessions, 2 of each, I have lost just this past London session... That is Phenomenal! Consider the issue of "drawdown" for a moment... I am shooting for the same basket TP as I am risking basket SL... so, to have over an 80% winrate for the past two days is quite remarkable, no matter how you slice it!

Also, as Caillou pointed out in a previous post, if you wait for a retracement, you may miss the boat (profits) for XYZ Session altogether! I know that based on what I've observed, only the most recent London session had said retracement, and so I lost the basket. However, the other 5 sessions that I did bank, didn't have much of a retracement to speak of if any at all... So... in theory at least, with the retracement logic in place, I would be 1-0, whereas I am actually 5-1 without it. See the difference? Losses are INEVITABLE! Why try to fight that and try to make a system that is impervious to losses, since it is IMPOSSIBLE anyway?

Now, I can offer two tidbits that will improve the system, in my opinion... First, I would like to humbly suggest a setting that closes the basket at an expiry, ONLY IF THE BASKET IS IN DRAWDOWN, as opposed to only when it is in profit... Why you ask? Well, simple... Every single winning basket I have had has closed in under 3 hours... but, I would rather let those RUN! On the contrary, if I am sitting in DD after 3 hours, I would rather just close up shop and wait until the next session to try again... Thus, I am limiting my exposure (and losses) when the basket goes against my favor, but I am maximizing my profit when the basket shoots off in my favor, and besides, I'm covered by my jumping stops, breakeven, etc. in the "good" scenario anyway...

Second... And I have suggested this previously, but it seems to go ignored or unnoticed... I think that if we could have a Percentage Risk setting Per Basket, or a Dollar Value per Basket risk setting, this would afford us the opportunity to risk the same amount of money on each basket, regardless of the number of currencies we are trading for each basket... This would help significantly in my opinion to maximize the profit potential.

Ok, there's my .02... Now lead me to the slaughter...

:)
Author:  SteveHopwood [ Thu May 31, 2012 3:35 pm ]
Post subject:  Re: Shelley's multi-pair basket trader
I moved the code around, so the ltf filters are working - V 1p in post 1.

What I am seeing is this: the ltf Slope is filtering out a fair few trades; the ltf Tma True is filtering out the lot.

:D
Author:  SteveHopwood [ Thu May 31, 2012 3:41 pm ]
Post subject:  Re: Shelley's multi-pair basket trader
jahill wrote:Ahem...
I respectfully disagree with the notion that the entries need to be tweaked with regard to the Shelley EA/strategy. Before you flame me, just hear me out. (I hope Shelley wouldn't mind pitching her two cents in on this, as she has observed the market behavior as it relates to this far longer than I have.)
FWIW, I agree. This is why all the tweaks can be disabled. I have added the tweaks only because people want them, and the great thing about an EA is that it allows experimentation so easily.
Now, I can offer two tidbits that will improve the system, in my opinion... First, I would like to humbly suggest a setting that closes the basket at an expiry, ONLY IF THE BASKET IS IN DRAWDOWN, as opposed to only when it is in profit... Why you ask? Well, simple... Every single winning basket I have had has closed in under 3 hours... but, I would rather let those RUN! On the contrary, if I am sitting in DD after 3 hours, I would rather just close up shop and wait until the next session to try again... Thus, I am limiting my exposure (and losses) when the basket goes against my favor, but I am maximizing my profit when the basket shoots off in my favor, and besides, I'm covered by my jumping stops, breakeven, etc. in the "good" scenario anyway...
I will do this; it is an excellent suggestion.
Second... And I have suggested this previously, but it seems to go ignored or unnoticed... I think that if we could have a Percentage Risk setting Per Basket, or a Dollar Value per Basket risk setting, this would afford us the opportunity to risk the same amount of money on each basket, regardless of the number of currencies we are trading for each basket... This would help significantly in my opinion to maximize the profit potential.
Someone else can code that if they want to. Life is too short for me to try to get my head around the logic required by that algorithm. :lol:

:D
Author:  Jemook [ Thu May 31, 2012 3:43 pm ]
Post subject:  Re: Shelley's multi-pair basket trader
jahill wrote:

:)
It's all good, you go do what you're doing and I'll do what I'm doing and we'll see which works better. We are all here to help each other so it's kind of nice that everyone is doing different things because we'll know quicker what works and what doesn't.

All you gotta do is honestly report your findings!
Author:  feucht [ Thu May 31, 2012 3:49 pm ]
Post subject:  Re: Shelley's multi-pair basket trader
jahill wrote:Ahem...
I respectfully disagree with the notion that the entries need to be tweaked with regard to the Shelley EA/strategy. Before you flame me, just hear me out. (I hope Shelley wouldn't mind pitching her two cents in on this, as she has observed the market behavior as it relates to this far longer than I have.)

That said...

We are trading at the beginning of each session when volume, liquidity and hence, movement is at it's peak, right? The idea is, we are scalping... yes, SCALPING... across many pairs that are in a strong trend, trying to accumulate many pips whilst only shooting for a few on a per pair basis. Thus, I'm not sure we need an optimal entry, as has been suggested. In the past 6 sessions, 2 of each, I have lost just this past London session... That is Phenomenal! Consider the issue of "drawdown" for a moment... I am shooting for the same basket TP as I am risking basket SL... so, to have over an 80% winrate for the past two days is quite remarkable, no matter how you slice it!

Also, as Caillou pointed out in a previous post, if you wait for a retracement, you may miss the boat (profits) for XYZ Session altogether! I know that based on what I've observed, only the most recent London session had said retracement, and so I lost the basket. However, the other 5 sessions that I did bank, didn't have much of a retracement to speak of if any at all... So... in theory at least, with the retracement logic in place, I would be 1-0, whereas I am actually 5-1 without it. See the difference? Losses are INEVITABLE! Why try to fight that and try to make a system that is impervious to losses, since it is IMPOSSIBLE anyway?

Now, I can offer two tidbits that will improve the system, in my opinion... First, I would like to humbly suggest a setting that closes the basket at an expiry, ONLY IF THE BASKET IS IN DRAWDOWN, as opposed to only when it is in profit... Why you ask? Well, simple... Every single winning basket I have had has closed in under 3 hours... but, I would rather let those RUN! On the contrary, if I am sitting in DD after 3 hours, I would rather just close up shop and wait until the next session to try again... Thus, I am limiting my exposure (and losses) when the basket goes against my favor, but I am maximizing my profit when the basket shoots off in my favor, and besides, I'm covered by my jumping stops, breakeven, etc. in the "good" scenario anyway...

Second... And I have suggested this previously, but it seems to go ignored or unnoticed... I think that if we could have a Percentage Risk setting Per Basket, or a Dollar Value per Basket risk setting, this would afford us the opportunity to risk the same amount of money on each basket, regardless of the number of currencies we are trading for each basket... This would help significantly in my opinion to maximize the profit potential.

Ok, there's my .02... Now lead me to the slaughter...

:)
I don't think anyone expects to get to a point where this or anything is impervious to losses, we all know they will happen. But that's not to say that is isn't worth trying. In simple math terms if this continues to work and make x profit 80% of the time, we need to make sure that it loses less than 4x when it goes against us. Using my previous idea of an index of where the basket is on average in its TMA True bands, how about using this to scale the SL? So if we were near the edge of the bands we could set the SL relatively small, larger if not. In the case of a Black Swan and a move well outside the bands against us we wouldn't get as hammered.
Author:  John [ Thu May 31, 2012 3:53 pm ]
Post subject:  Re: Shelley's multi-pair basket trader
Steve,

I appreciate your reading my short story... Thank you for your willingness to implement the "bad basket expiry" feature as well... I think that will go a long way toward making the good baskets make a killing while limiting the injury caused by the bad ones. In addition, no worries on the Basket Risk thingy I proposed... Truth be told, I fully expected that it was likely a pain in the arse to code... Anyway, I think this thing is pretty damn fun if not wildly impressive... My sincere gratitude.
Author:  Jemook [ Thu May 31, 2012 3:56 pm ]
Post subject:  Re: Shelley's multi-pair basket trader
jahill wrote:Thank you for your willingness to implement the "bad basket expiry" feature as well... I think that will go a long way toward making the good baskets make a killing while limiting the injury caused by the bad ones.
Just out of curiosity, what time would you set on this feature? Tonight it took a couple of hours in DD before the trades continued with the trend.
Author:  SteveHopwood [ Thu May 31, 2012 3:58 pm ]
Post subject:  Re: Shelley's multi-pair basket trader
jahill wrote:Steve,

I appreciate your reading my short story... Thank you for your willingness to implement the "bad basket expiry" feature as well... I think that will go a long way toward making the good baskets make a killing while limiting the injury caused by the bad ones.
I like the idea so much that I have added it immediately. V 1q in post 1. The input is OnlyCloseLosingBasket.

You are safe for now, guys. If I see one more line of code today, I shall scream.

:D
Author:  John [ Thu May 31, 2012 4:00 pm ]
Post subject:  Re: Shelley's multi-pair basket trader
Also, just to clarify, I in no way was intending to offend those or discount those who have offered suggestions to improve the entries. Truth be told, I was only trying to ensure that the EA wouldn't be drastically changed such that I would be forced to adopt those changes in my own trading. Yes, by all means, let's see what those extra tweaks can provide for the strategy, and if they prove to be better, I can assure you I will be an early and eager adopter of the evolved strategy! Thanks for understanding... If there is one thing I've learned about my years in the forex game it is this: There is no "Right" way and there is no "Wrong" way... There is only Profitable or Not! :)
All times are UTC Page 51 of 87

Powered by phpBB® Forum Software © phpBB Limited