One of the more interesting EAs to pass across our desk recently is the Forex Pipstack EA. Now on version 2.0, this Expert Advisor trades the gap between the North American and Asian trading sessions when market volatility is low.
It trades four pairs: EURUSD, AUDUSD, EURGBP and GBPCHF on the 15 minute time frame, and relies heavily on support and resistance - which as we all know is a good thing.
The results from previous versions have been really good, and we hope that this EA will perform well in our live test.
Our resident EA specialist Capsmart will be testing this one. You can keep tabs on progress using the link below, which shows live trading progress:
Hello everyone.
Installing the pipstack is relatively easy since you will have to download a program that will assist you locating the correct files in the correct place. 3 files are extracted but you do not have to worry about this.
Then you have to create 4 tabs for the different currencies (eur/usd, aud/usd, eur/gbp and gbp/chf). We can see that there are no yen pairs and it is more or less a "european" game. The EA will trade after the close of Us and before the Asian open when things are not moving much.
Therefore you have to "instruct" the EA about the correct time.
One more thing concerning EA authentication. You are allowed to use this EA on one real and one demo account. So after installing it you have to login the member area and enter the account number. If you decide it to use it on another demo account you will have to re-login in the member's area and change it again.
I think this is the best practice. There are developers that simply let their EA's work on any computer. This is not bad, but it could result in program theft.
Some other developers make you enter your purchase code in the EA. This is not bad, but again 2-3 people could share the same EA.
Another way of protecting the software is to use a program called CodeGuard. This is really a pain, because it will only work on one computer. No way to have it installed on a laptop and your normal computer. You have to buy two licenses.
And finally what PipStack uses where they are protected since it will only trade for one demo and one real account from as many computers.
Ok I do not want to keep the posts too long so I will close it here and continue
The idea behind the EA (and it is better explained in their site) is to determine a range with support and resistance and trade between those areas.
It sounds logical. After Us close, volatility is usually low so a range trading could be profitable.
We tell the robot how many bars to use to determine this range (the default is 10) and of course there are MM parameters where we could define the lot size or let the robot decide about it using a risk percentage. So far so good.
We also define a TP and SL that according to the manual should be 32 and 16. So a 2:1 ratio.
My only concern is the forward test that is in PipStack's site under Trading Performance.
In this one I can see winning trades of anything between 1 and 245 pips and losing trades between -1 and -125 pips
I also notice in this performance report that trades are open basically around midnight (logical) but there are trades opened at 6, 9, 17 that should not be there.
Same for the closing of the trades. Most of them close exactly at 2:00 am (logical since the Asian session starts) and this could also explain the different TP and SL. Apparently 2:00 am is a rule for the EA to close the trades no matter what.
This performance report is on demo that started the beginning of June with 5000 dollars. Till the beginning of December it shows a profit of 1873 dollars. 37% for for 6 months = 6.2% per month. But....
I see at least 4 "suspicious" trades. One at June 8th that started at 4:43 in the morning and was closed next day at 3 am with a profit of 245 pips and 490 dollars.
Same with another trade that was opened on Aug 31st at 6 am and closed on Sep 6th (after 6 days) for a profit of 245 pips and 490 dollars again.
And again on Nov 29th and Dec 1st for 300 and -375 dollars.
So if we subtract the 825 dollars of the "suspicious" trades the result is down to 1000 dollars or 20% for 6 months.
So let's wait see what this baby is going to do while we are in bed.
We have the results of the first night.
Not encouraging but the market did not do what it is supposed to do according to the system.
I am attaching a picture of the trade. I drew myself the 2 red lines indication the range. When price reached the lower line the EA opened a long trade expecting price to remain in the channel and continue up. The Money management is activated so it opened a trade of 12.5 lots. It then open another trade for gbp/chf and 4 lots and finally another one for eur/gbp and 7 lots. The stop loss was hit to the pip and price reversed and entered again the channel. So we could say that it was not lucky day. Let's wait and see what is going to happen tonight.
What I do not like is the different lot size. All EA's have the same parameters. As soon as I find out I will let you know.
I assume the 3 orders were each on different currency pairs? The money
management is based on percent risk of account equity. Since the
tick-value (in USD) of each currency pair varies, the lot size will
also vary even when the equity is not changed considerably between
trade entry times.
The stoplosses are the same across the board, but because the actual
capital risk associated with those pip values is different, the lot
sizes are adjusted to ensure that the account is still only risking 3%
(or whatever your equity risk setting is at).
You can test it by calculating the pip-value with the position size
and multiplying it by the stop loss. It should come to roughly the
same percent of the account equity when the trade was placed.
If you have anymore questions regarding the system I am more than
happy to help.
Since I was not happy with the explanation I asked for a more detailed one.
Here it is.
Since tick (or pip) values are always calculated in the account base
currency, which is usually USD. Pairs like the EUR/GBP and GBP/CHF are
actually called "crosses" because they are not traded directly from
the USD. Instead what is happening is the USD is getting converted to
CHF in order to trade it's value against the GBP.
When you buy a lot of GBP/CHF you are purchasing 100k (or in MB
trading's case 10k) of GBPs. This is worth roughly 155,000 CHF which
is worth roughly 152k USD. So for every 1 pip your USD value actually
changes 15 dollars. While the value of the EURUSD pip is only 10
dollars and the EUR/GBP is about 13 dollars.
So using your numbers the total dollar risk associated with each trade
would be. (I have divided all the lot sizes by 10, because MB trading
uses lot sizes of 10k rather than 100k that most brokers use).
Interesting... It looks like the account is not risking the same
amount per trade. Assuming the account equity at the time of the
trades was relatively the same. Do you have any screen shots, or an
account statement I can look at. We use MB trading ourselves on a live
account and haven't noticed any discrepancies in position sizing. I'm
interested to see if there is an actual problem here so I can fix it.
One more reply from Kevin. He is doing his best to explain me the system.
After looking at your statement, I was reminded of something I forgot
to mention earlier. Spread.
Something we've been struggling with is the spread on the GBP/CHF,
particularly on MB trading. While MB typically has the tightest spread
on major pairs, we're seeing spreads approaching 5 pips and sometimes
more. I've been gathering spread data on 10 different brokers for the
last 2 weeks, and I've seen quotes come in on the GBP/CHF with as high
as 9.7 pips of spread. The risk calculation takes the spread into
account when determining account risk... that said, I still feel like
that trade was a little small, but if the spread was unusually wide
during that time it could explain it. The EUR/GBP with 7.9 lots, and
accounting for spread looks about right.
We're working with some methods to improve trade entry, not just
monitoring the width of the spread... but trying to get the order
filled at an optimal level (hard to do with no depth of market).
Keep me updated on any new trades and we'll see how it sizes the
positions. Also feel free to copy/paste any of my email to your forum
members. I meant to be thorough so anyone could understand.
I am attaching a screen shot where you can see all the trades.
So last night we had 3 winning trades that recovered the losses of the day before.
But the strange thing is about the second trade from yesterday. If you see my post from yesterday you will see that the lot size is much smaller than the one I am attaching today. This is something that I have never noticed before, but of course I never go back and discuss my closed trades.
So as you can see. Yesterday for the gbpchf trade the lot size was 4,1 lots and in today's statement THE SAME gbpchf trade is for 12.30 lots.
Two more trades last night. A winning one and a losing one. Profit double than loss. We are at 10146 so 1,5% in 3 days. But the first day was bad so it recovered the second day and now it is profitable.
I am attaching the winning trade. So you will see once again the logic. I drew myself the two horizontal lines. When price hit the upper channel it went short and it closed the trade with a nice profit. I am not sure about the lower line because of the 2 first candles of the range.
I also watched the other 2 pairs to try to understand why it did not open any trades. From what I see for the first one the range was very small (less than 10 pips) and for the second one (gbpchf) price was in a slight down trend, the channel was about 20 pips and from what Kevin told us it has a big spread. So maybe this is the reason it did not trade.
LinkBack URL
About LinkBacks
Forex Pip Stack
Reply With Quote
