Basket 14 EA System

Only that the pairs traded are limited to EURCHF, GBPCHF and EURGBP.

You could take AddSix 02 and modify the logic to only trade EURCHF, GBPCHF and EURGBP and you'd have the Add_04.

The good results are partly due to using a bröker with tiny spreads. The downside with that bröker is that you have to pay a monthly fee. I tried the same EA with IBFX and it didn't open enough trades to be worthwhile. Big spreads on those pairs get in the way.

A good thing to keep in mind is that the policymakers, bankers and brökers over the last few years have added little annoying things to make it hard to do well as retail traders. Swap/interest rates were adjusted to keep us from accumulating swap money on certain pair combinations, same-pair hedging was eliminated unless you follow FIFO (or however it is now), ticks are heavily filtered to keep us from exploiting distortions, leverage has been decreased (probably good for the many clueless traders who haven't bothered to self-educate about margin, leverage, etc.), brökers adertise "low spreads" but freely increase them when it serves their purposes, brökers flag and disallow various trading styles and trade against our positions... and that's just a partial list.

MC,
If I were to try to implement a trading time feature into one of my mods, would just having it set Stopafternotrades input to true automatically work?

Stopafternotrades just doesnt reopen the cycle again once a global closure is hit, but it doesnt prevent new trades from being added every hour as normal right?

If internally the stopafternotrades was set to true by trading time code, will it reset the ea as it currently does when you set that to true manually?

Thanks

Some brokers have limitation on max number of open trades (usually 100) I got there several times. Then the EA will not open any new trades. After reaching this point the EA should close shop and start over otherwise it can result in big DD.

Hehe, thought I would get such an answer and I didn't see it in the code either, but maybe overlooked it so wanted to know it for sure.

I have an $5000 demo-account at FXCM which only allows 0.1 lotsizes as minimum. Running it with Ph14_AddMore_01a. For now it has 36 trades open for a total of -$400. Dollar profit target is 0.5 % -> $25.
Lowest margin is 100.14% Max floating drawdown was -8.37%.
It still survives.....
I'll edit this post when it "died"

By the way. I get very nice results with the Phantom14_Add02
I'll post a statement next week. I just let it run a bit longer to get a objective result. The Phantom14_Add02.taz version is also doing nice.

Thanks all for the contribution here.

A good reason for limiting the pairs traded.

Aahh, the market maker...

And i think that (maybe you) someone tried to base an EA on the profit, and not on the pips ? which can solve the problem of the spread.

Do you please remember which one is it ?

Thank you MC for the details

The Collective FX is the bröker.

Cleanest solution would be to remove the spread from the profit calculation, just look at bid movement and apply lot size and tick value.

My new mod idea (which might have been what thegame was playing with not sure) is working nicely so far. Most legs are opened and close within 15 mins, was pretty fun to watch last night.
However, I do think a weakness to this one, as well as the add_02.taz mod, is the timing of the new legs. From what I see so far in testing, I am of the opinion that a cycle that is opened toward the end of the us session is typically the one that creates the dd. Obviously things slow down quite a bit, asian retracements occur, etc. So what ends up happening alot of times, about 1900 or after, a cycle starts up. Then just flounders around until a bit before london open, but is adding trades all during that time. Then at london open, we have to just pray like hell that it doesnt take off in the wrong direction of our basket, when momentum and volume takes off.
So, what I propose, and maybe someone can help me with this.

At 0500GMT, the ea can turn on and start taking basket trades.
At 1900GMT, the ea continues what it is doing if it is in a basket cycle already, continuing to add to trades as it's coded to etc. However, the stopafternotrade flag is set to true, without affecting or refreshing the ea. Once the basket leg is closed in profit, the ea sleeps and takes no more baskets, until 0500GMT, when it awakens and does it all over again.

If someone would be kind enough to help me out in adding this code, I am more than happy to share the latest mod I am working with.

P.S. Yes, this can easily be done manually, just change the flag to true at 1900GMT, however this action resets the ea, so all phantom calculations are reset to 0 and it affects the current basket cycle continuing to do what it needs to do.

Here's my chart for the 18 hours of testing i've done so far. The 6% dd happened during the time I am trying to avoid with the above, and the second dip happened because my damn pc crashed due to another crazy ea I am playing with while I was sleeping last night. Regardless, 10% gain in 18 hrs isn't too bad And most baskets are closed before I even notice they were opened lol

My version was like yours except I was checking the pair's order type (buy/sell) before taking the trade in the same direction (when they were in profit obviously).

I think because you haven't done that it actually saved you because you may find that there was buy and sell trades during a leg on a pair (or several pairs), effectively hedging itself and reducing the risk.

Basically your mod was letting the winners win and the losers lose OR even better hedge themselves (possibly).

I think we are on to a winner here.

Hmmm, I wonder what would be the advantage in doing this rather than letting the EA reset the Phantoms say 15 minutes before the active time?

Yeah, I thought about that too, which is probably fine, as long as it doesnt attempt an entire new basket cycle right before the asian session starts. The current basket I am in started right after that magical time I was referring to (which I had already decided on logically last night), wierd coincidence

Anyhow, my mod basically is the addsix with following changes.

1. my normal mod, which is reversing the biggest loser, and adding the biggest winner.
2. doing it every 15 mins
3. only looking at the 6 pairs in the basket I am trading, so there will ALWAYS be a biggest loser/winner. So the other 8 pairs are completely removed from the ea's calculations.

Probably pretty close what you were trying

Yep I also did point 3... I just overkilled it with the same direction code LOL!

It was doing alright then the whole market (it looked like) just changed direction.

So you had it taking more trades than the norm? (other than the obvious every 15min check instead of every hour)

What time gmt did the basket start that got you in trouble?

No, mine was like yours but I checked the existing trades direction before going into that direction.

Sorry can't answer that as I deleted the demo account and restarted with another strategy.

ah i see. well mine has still never opened the same pair in opposite directions. Never had a basket cycle open long enough yet for that to be possible I guess.
Even on my add_02 mod, I don't think i've seen it, so maybe there is a check in there for that already? dunno

Too weird... maybe it's that timing issue then.

e.g. my EA took the initial trades in the "wrong" direction and yours was right because you activated your EA at a different time than mine?

taz, when you set StopAfterNoTrades to true, I believe it stops the EA from funtioning at all. The EA stops and makes no trades until that switch is turned back on(set to false).

the_game, yeah most likely. I believe I started mine AFTER you had mentioned you had an issue.
My overall point is, imo, a basket with this method that first initiates that close to the us session close, to me is based on unreliable information. The pairs that are deemed the biggest loser/winner have no momentum to continue their move. Once asian session kicks in all bets are off usually (not always), momentum drops to nothing, volume drops to nothing and alot of the time the typical asian session slow crawl retrace occurs. Compare that to the london open, and as you have seen, if a pair starts to move, it has no problem doing it, with a vengeance, and pops us out very quick.

frazerd:
ok, then that's going to cause a problem with my idea then, so I guess I am going to have to find a coder that is willing to help me make this work the way I would like to see thanks for the answer about that. although the name of the input itself is a bit misleading, it should just be called stoptrading, not stop after no trades

Hey taz,

I'm too busy today at work (Friday!!!), but I'm sure you can manage to hop on to docs.mql4.com and have a look at the Time and Date documentation.

You've gotten this far, you should be able to do it with ease after a bit of reading up.