Spider: an auto-trader for spiderforex's VQ system

Gave me an excuse to hang in the bohio and drink a lot of rum with good friends and eat great bbq

I whole heartedly agree with the 1hr trend on 15 min time frame. The daily and weekly trend are fine, if you have deep pockets, because the unscheduled news releases and quick changes in risk sentiment can cause strong pullbacks against the trend that will hammer shorter term time frame trading with the longer trend.

Steve,

I am setting them up now. NO AA, No Martingale, using Hanover, one using 1hr, the other using daily to determine the trend. Trading size of .1 lots on FXCM.

For the trade exit inputs i am using 30 SL and 30 TP. Is that what you want to test to start of with? No trade management features like JS etc. I just want to be sure before i run them.

Or do you want to run it with OOTB except for no AA, and no M, and just alter the Hanover settings to determine the trend?

Mike, I can say with total honesty that my brain has fried and I haven't a clue.

Just try some experiments. Sorry, but I really don't know.

Cheers

Testing Asian M1 scalp setting, 10/10 TPSL

TF M5
AA off
M off

Same here - any chance to mute the terrible sound on this?

Thank you
josch

SORRY FORGET THIS - NO SOUND WITH ERRORS

After one week of testing here is what i got:

TF15 (Hannover H1), AllAverages: ON, No Marti (detailed settings in attachment)

http://img338.imageshack.us/img338/6...1105100703.jpg

TF (Hannover D1), Averages: ON, No Marti (detailed settings in attachment)

http://img577.imageshack.us/img577/1...1105100823.jpg


Clearly Hannover kept me away from many trades and those that were fired happend too late in my opinion. While VQ signals where pretty nice during this week my spider took far less trades.

Maybe it would be better to use a higher ratio for hannover: 15m trading w1 hannover. Probably setting Han. slope to 0 would help as well, and keep me away from going against the markert (entering short at reversals for example).

What i could do next week, would be setting up EAs with a combinations of:

 

1. 0 slope
2. playing with threshold
3. w1 or even m1 TF for hannover
4. figure out better money management

What do you guys think?

Fabulous blazko. Many thanks.

I suggest you abandon further testing. Here is what I think:

 

1. spiderforex posited a perfectly acceptable trading model.
2. we have done the usual thing, and filtered it until it no longer works
3. what was once KISS is now MIRCS (Make It Ridiculously Complicated, Stupid)

Here is how I think spider's system will work, and how the EA can do the same:

 

1. trade the EA only on an account dedicated to it, on EU as originally posited
2. turn off all filters that are not directly a part of sf's original system
3. trade a tiny starting lot, so small that M will have to go ludicrous distances to blow the account
4. watch the account grow slowly

After all, the system only requires the EU to move 30 pips in one direction every once in a while to net 30 pips overall. Keep on netting 30 pips and the world belongs to you.

If I remember tomorrow, I will post an Elementary Spider EA with all the extra stuff removed, so people can try this easily.

sf is probably applauding warmly whilst reading this. He knows how to make his system work. Never mind; no harm in experimenting, especially when we can have a robot do the work for us.

interesting conclusions. I was wondering what the outcome would be.

This is the most promising SH EA indicator wise, interesting that filters be it hanover/all averages have no positive effect.



simple is best eh?


Should generally mean hanover will be its own beast.

Couldn't agree more Steve on your last post, so often we see a good simple system destroyed by over filtration, like any system it's going to have bad days due to the current market conditions.

Keep it simple, use good money management and be prepared for the losses....

Thanks Steve. Please do post the "scaled" version. Many thanks, you are a jewel!

I have put Simple Spider into post 1. There is no difference between using this and using Spider with the extra filters disabled, but the inputs list is vastly shorter and easier to read.

Thank you so much Steve

After 1 week without internet access, I can now restart my demos
with this version

Regards
Alex

Encouraged by Steve Hopwood live trading Spider, and the simplicity of the Simple Spider, I dared let it loose on the following 34:

AUDCAD, AUDCHF, AUDJPY, AUDNZD, AUDUSD
CADCHF, CADJPY
CHFJPY
EURAUD, EURCAD, EURCHF, EURGBP, EURJPY, EURNOK, EURNZD, EURUSD
GBPAUD, GBPCAD, GBPCHF, GBPJPY, GBPNZD, GBPUSD
NZDCAD, NZDCHF, NZDJPY, NZDUSD
USDCAD, USDCHF, USDDKK, USDJPY, USDNOK, UDSEK
XAGUSD, XAUUSD

when the market opened for the week. Demo account, M30 charts and OOTB. Later I disabled the Martingale.

Very encouraging results so far.

Khalid

I know demo is only monopoly money, but the consensus is that you should trade it like real money. As such trading Danish, Swedish, Norwegian currencies and Gold and silver is not really suitable for this type of strategy.
Kind regards

Thank you for you feedback, astral77.

I am a seat-of-the-pants trader who is trying to figure out systematic trading. Still learning the elementary stuff; at the “what if/why not” stage, and won’t rule out, or rule in, anything.

Rather self-centric, perhaps. But I am also prepared to run up to 4 additional EA demos Sunday to Friday for the general benefit if advised by experienced members what to do.

Khalid

However, many a potentially profitable trades closed at loss.

There should be a way to attach some utility to an EA recording trade data similar to the Account History, but also including Maximum Favourable Excursion and Maximum Adverse Excursion so all (forward) tests would yield data to better determin TP anf SL -- Excel example attached.

http://www.fibonaccitrader.com/help4...AE/default.htm

Khalid

Steve, You are a mind reading alien, simply the world is not enough for you

I don't know what to say...you know it already ....yes...very true...I was checking the thread now and then, but could not post it to write a long view on what i think and i knew the usual direction it was taking with all inputs "TO IMPROVE" the system.. to make it perfect etc.

We have seen 100's of system going this way by improvement...we have all been there...I wanted a very basic system..which will work it out in the long run...If we just accept few losses and let martingale play its part once in a while...The system will be profitable ...I wanted a simple EA to work on what i was doing manually, which was the first EA we did then contributions started coming in and with filters etc..there was no breathing space for the EA to work it out....

I feel we cannot MAXIMIZE profit by filtering trades and trends, some losing trades are part of the "system"...Martingale should take care of it...IMHO There are no "improvements" that could be made in this system for it to "Work Better". This is the max it can give, We can't milk this system more by applying more filters and additional indicators etc etc ..it will take some time with this system to slowly built your account balance..

What we can do instead is to work on the indicator length to find it best for a particular pair..some pairs work well with 8 length some with 12 and also tweak it for Time Frames which suits ones trading style...

Just use Martingale when a losing trade comes, it won't cross more than 4-5 trades, use jumping stop loss or trailing stops with break even pips ...you can't loose in the long run...the danger with Take Profit is we don't know for how long the trend will continue...

For example ...if we have one losing trade - one winning trade and again one losing trade - one winning trade and so on with 5 SL and 5 TP...we will be still losing due to spread, now if the trends break at 4 pips , we make 4 Pips profits and the next trade takes a SL of 5 Pips and so on, we are still losing..by using tp ..we won't be able to take full advantage of the trend, if the trend changes before it hits TP...The best way to use this system is to go with Jumping Stop Loss or Trailing Stop Loss with Break Even...Let the system run on 5-6 pairs and with break even profit of 1-2 pips, we can build up the account but it will be gradual process, if we don't use break even..the profit will be faster but with losses as well and the swing in the account balance will rise and fall rapidly..


I think we are now taking the right direction by keeping it Simple.....

Steve, Thanks a lot for not giving up on the system and being persistent with it...I think Simple Spider is the way to go...it will built the account balance gradually...

=========

P.S...As other members would also be reading this post...This system is free to modify and use as you may seem fit...I'm not trying to step on anyone's toes as im learning also, the contributions you provided with filters etc were in all good intentions...to make the system better..I just gave my opinion on how best to work with the system. No harm meant if you feel i have criticized your contribution to the system. We can all work together ...thanks

Spot on, sipderforex.

Except, in my humble opinion, about Martingale.

The only thing Simple Spider needs are effective exits.

I only ran the the EA for about 48 hours; and very choppy 48 hours those were. And also ran many SQ Index simualtion on MultiCharts. Here is what I think:

Spider can trade any market, both major and exotic. Indeed, XAUUSD was the star performer amoungst 134 EA trades and more than 70 additional Indicator signals the EA did not take; followed by EURAUD and AUDUSD.

Probably better to trade 20 odd of the more cost-effective pairs -- ones with more favourable (Spread/ATR) ratios -- practising uncompromising regime of cutting losses and proctecting profits.

Building equity on the few trades which will generate nice profits every day, and position size on basis of (Total Equity*%Risk per Trade)/Average Loss or ATR-based Volatility.

Khalid

Dropping the other filters helped a lot.
I got a few nice trades and what was a red balance is now a green one with some fat pips inside


There is one thing a could think about (to improve exits) - a candletrailing stops with an option to set 2-3 candles look back - now it works on the last one and it got me out a couple of times (and it was good) but also there where situations when holding SL back to 2 candles would improve profits.

The only filter i might look into and test it with different VQ settings is AA.

Best regards,
blazej