I’ve been noticing this a few times now when scalping GBPCHF around NY close (23:00 GMT). I use very tight stop-losses when price have been trapped in a clear channel for the day (or at least NY session). Sometimes of course this doesn’t work out, price breaks out of its channel. But now and then price have stayed in the channel, but I’ve hit stoploss anyhow on a few account, not all. The one below is a GoMarkets Demo account.
I (or rather my EA) have been scalping the channel both high and low successfully during the timeframe I scalp GBPCHF. I would say that every entry is sound and should be profitable (or close for a break even after a specified time, which is how this EA is programmed). But the first sell (red) order in the picture actually closed at it’s SL level at preciesly 23:00. A level that never was reached if you just check the chart above.
The problem of course is that the broker, GoMarkets in this case, for just a few ticks had a huge spread in this exotic pair. It was probably around 20 pips (but I don’t have the exact ticks saved so I don’t know exactly). This triggered the stoploss and a loss was realized. Just ticks later the spread was normal again and the trade would have taken profit or at least closed out for a break even result.
The solution is to move the stoploss higher, but since I risk a set percent of my total account on each trade I stand to make less profits if I adjust this. Another way would be that the stoploss is moved higher from 22:59 to 23:01, by the EA in question.
I think I will have to try the second solution actually. It would mean that I risk more than I really want for 2 minutes, but on the other hand hopefully filter out these unwanted stop-losses that occur just because there is some turbulence in pricing at NY close.
I’ve written a bit about this before. But now the live Pepperstone account have traded for a while, and I would classify the second half of September as quite hard because of the turbulence in the markets. Seems as Forex Shocker v3.0 is quite picky when to trade, which is good if it picks the wright times of course. And, well, it has. Do remember that they had losses in early August, but that they restarted the MyFXBook account to hide them. A silly move of course because no one cares about losses if they are balanced with more profits. A system that never loses doesn’t exist.
http://www.myfxbook.com/members/shockerv3/shockerv3/148370
As you may recall I entered the MyFXBook trading contest at GoMarkets that ran during September. It didn’t go very well, I lost 13%. I blamed the high spreads at GoMarkets partly (for my scalpers), but mostly the extreme turbulence we saw during the second half of the month. When the contest was over I started a new demo with GoMarkets in the same terminal, IE with the same settings. I choose a $5000 account this time thou, it is more realistic for me. Then I forgot about it, it is on a VPS over at SWVPS that I don’t use much anymore, but have payed for a year in advance.
I checked in yesterday. The account has exploded to about $11200, which equates to about 124% gained in 14 days. At first I thought that I had mistakenly started a $10000 demo, but after checking the trading history I can just say that the EAs in the account has made a great job during these weeks. The performance is not possible in a live account thou, because I forgot to adjust the lotsize of one of the EAs. It still used fixed 1.10 lots for trades, this number was based on the $50000 account size of the contest account. IE I have been taking 10 times the risk in this demo. The interesting thing is the gained pips thou, and in that respect this strategy (a midterm intraweek EURUSD strategy) have been doing very well with 6 trades resulting in 542 pips. I have corrected it to take 2% risk now, which is more realistic for a live trading environment.
All the other EAs have been doing great also, including my main scalper. I really can’t wait to get my money from JadeFX now so I can start a live account at Pepperstone (via cashbackforex of course for the kickback) and really get into this again.
Here’s the account (demo at GoMarkets): http://www.myfxbook.com/members/someguy/ea-portfolio-gomarkets/179700
It’s been quite some time since I traded accounts with any sums to talk about (I have a very small account at Alpari UK funded still, but it isn’t doing well with automated strategies, it’s okay for manual mid term strategies thou). I’m still waiting to get my money back from JadeFX to start trading automatically seriously again. The last word is that the NFA is going to get access to JadeFX accounts in November and that they then will start the distribution of funds. It will be interesting to see how much I will receive since I had quite a bit more in the account than I had deposited (I had profits of 143% during 2010H2). I’m guessing I won’t get the profits back, but at this point really anything is better than nothing. I’ve been waiting for 7 months…
In the meantime I have been testing different strategies out in demos at FINFX and Pepperstone. I was quite confident that I had a nice setup with 5 different EAs after the summer, but the last contest at MyFXBook unraveled some disturbing losses. The scalpers really suffered from the relatively high spreads at GoMarkets (they have been optimized for the low spread trading environments at ECN-brokers like Pepperstone and FINFX). More disturbing was that some intra-day strategies didn’t work well when turbulence hit the markets in mid September. All in all I lost ~13 % during the month, and I was up about 10% halfway through. I know that if this would have been a live account there would have been days I would not have traded because of the turbulence in the worlds markets, so I think a live account would have managed a break even month. I would like the EAs to take this “no trade” decision themselves thou, and it is something I am working on.
I think that you will see more activity in this blog once I get my (forex) trading capital back from JadeFX. Until then I will try to take some time and update the Forex Guide. It is long overdue.
