Hi Steve,
Thanks for re-opening this thread again. It is much appreciated. I do want to publicly apologise for at least being one if not the only one who was responsible for having this thread locked. Main reason being I had posted on this thread, discussions of other EAs which I should have posted elsewhere if at all.
MOVING ON... Thank you again for re-opening this thread.
Right! Over the weekend I took this EA to the test. I did some.... OH NO! ... wait for it... BACKTESTING!
Let me point out straight away that I too am a firm believer that backtesting and especially on Empty4 is rubbish and an absolute waste of time. HOWEVER, with this EA, why would testing pending orders be a problem? There are no indicators or room for many errors during test, they are simply pending orders and closes. THAT's ALL!
Soooo... what's the problem? I'll tell you what may be a problem. PROPER DATA FEED.
After further investigation I have noticed that my old broker Tallinex does not provide quality tick information to do an Empty4 backtest based on every tick! Rubbish.
However :youknow: more brownie points and applauds
to GLOBAL PRIME!!!!
Their tick information is QUALITY! I've been backtesting Griddy Wotsit on the GBPJPY flawlessly!
Sooo... What sort of results have I found?
1. The setting of Take Profit Pips is only good if the pair remains in volatility or Replace Opposite Grid at L4 to false.
Imagine the pair trending in a direction and not experiencing any pullback and you are not in a hedge. PA will open your pendings in one direction and will close them once your Take Profit Pips is achieved.
What next? Unless you have set the Replace Opposite Grid at L4 to true, you will have created a massive gap between where your orders have opened and closed in profit, to the other side of your first pending on the hedged side. In a trending market as I have tested this and there is a point where the gap between buy and sell was so far that the EA simply would not open any other trades and may have never done so for months.
So, why not set the Replace Opposite Grid at L4 to true? The main reason is DD. Too many unnecessary open positions in all sorts of directions clogging up the screen and eating away at your equity balance until a hedge closes. I am convinced that leaving this setting to false is the way to go UNLESS... you want to increase the size of the grid which I will leave that for another discussion.
So, how did I resolve this problem? Heh heh, cheekily. I decided that I didn't want Take Profit Pips at all and set it to "0". "So what, you waited for the EA to be in a hedge for a TP?" No. Because the EA does not have the feature of a Global TP option, I have modified the EA to think that it was in a hedge even when it is not. But this was only a temporary solution to my problem. My question is to Steve.
"Can you add a cash TP option in a non-hedge so we can benefit from setting Take Profit Pips to "0"?"
2. Anybody who has been using my setfile, I would chuck it in the bin, burn it, delete it, smash it to smithereens IT IS USELESS!
Right, DISTANCE BETWEEN TRADE PIPS. Unless you have deep pockets, in my backtest I have needed to re-evaluate this setting. I have seen on the chart, jumps of near £750 and that is pounds not dollars, when closing baskets. This was why I asked Steve for a function that would close positive positions first. I won't get into my personal sob story but WHEN I start trading live, I will only have about £500 to trade with so it is crucial I find something consistent that will not blow my account.
All seems well when GW works. "Oh, another basket! Yay!, Oh, another basket! Yay!" But what about when you have 20 positions on each side of the grid and your thinking how many positions will it take to close the hedge? Well, the answer is actually around 55 positions against 20 but that is besides the point.
So even with the setting of Replace Opposite Grid at L4 to false, there is still too many trades open! Well, as opposed to just hoping I would have the right settings to make GW work, I've actually decided to study Grid trading and Grid EAs and have seen various methods including martingale and etcs that I care not to go into. WHICH BY THE WAY, it is more of a consensus amongst traders that Grid trading doesn't work! It is something to chew on even though I am of the opposite opinion.
So after studying, at least the main opinion of Grid traders and Steve also makes a mention somewhere in the first few pages of this thread if you have bothered to ready the entire thread (it is a short one), that DISTANCE BETWEEN TRADE PIPS is best kept larger rather than small. Although, I seem to have had the largest distance amongst us testers, IT IS STILL NOT GOOD ENOUGH. It blew the £500 account which was at around £650 at the end of the month of March 2015 (which is the year I am testing) with however 70-80+ positions open.
I have now moved this distance to 20 pips and will see what results yield on the GBPJPY. This will not only result in fewer positions open but hopefully as well will result in unnecessary hedges every so often while the EA trades.
So there! I'd like to thank Steve again for unlocking this thread and request a tiny change in GW.
I will keep positing my findings but I suggest that you open yourself a Global Prime account and take advantage of their valuable quality tick information that results in highly successfully backtesting GW EA on any pair that you may choose. If anybody care to try a different pair rather than the so infamous GBPJPY which is known to be one of the most volatile amongst the munch, try GBPAUD. Surprisingly it was pointed out to me and I am curious to see this pairs results.
Happy Pipping!
SSL (I could have gone longer about other things but I think it was more than enough for 1 post.)
Daniel
