What to do…

A while ago I beta tested a product called OrderBook FX that basically gives you a tool to understand the current market order depth (or depth of book) for currency pairs. The company behind the product collects its underlying data from multiple institutions and aggregates it into what it calls a “global depth of book for all major currencies“. During the beta test there where many problem with non-us installations and such, but all where fixed and they are now live accepting customers. I very much like the chance to see market depth in MT4 as this is very poorly implemented in the standard MT4 terminal (better in MT5 if your broker offers the service, but still not a “global” view, just the brokers book or the ECNs book). Using the global book you can take advantage of selling or buying pressure to speculate in coming price movements in a totally different way than with other indicators. I only had very little time to experiment with this during the beta test and following 7 days free trail, but I will definitely work on some filters for my current EAs to use OrderBookFXs API. It is excellent for a Asian session scalper to know when there is a selling / buying bias in the market to wait for channel bounces and even break outs. And of course this API could be used as base of a new strategy also, buying/selling when others do and chances are that prices will move.

I will very likely (90+ %) sign up for the service once my major trading capital is returned from JadeFX by the helping NFA and CTFC. Hopefully that is soon…

OrderBook FX offers a 7 day free trail. Make sure to use it well. After that there are two choices depending on the amount of pairs you would like to follow at the same time for $49 and $79 per month. Money that better entry and exits by just fractions of a pip will get you back over a month of successful scalps I think

Just a quick note that Birt has updated his excellent collection of tools for tickdata backtests in MT4 using Dukascopy tickdata (some time ago, but I’ve missed it until a few days ago). The big news is the new compact FXT format (deletes duplicate ticks) which really saves on file size allowing you to test for longer periods and a extension of the 2 GiB limit to 4 GiB on Vista and Windows 7 systems. I’m still on Windows XP SP3, so I haven’t been able to test the last feature, but with the new scripts I still can test 2+ years on any currency without hitting the 2 GiB barrier. The tests also run faster on the same time period because it contains fewer ticks, which is always very welcome.

If you are new to tickdata backtests or run older versions you have to check it out.

It was of course inevitable, there had to be a loosing month sometime. January started of rather bad (like December ended), picked up a little and then fell through on the 24-25th. And then some more yesterday the 31st. So for a total loss of 14.53% it ended rather badly.

It is trades in EURGBP, the main currency traded that have been bad, and mostly then by The Channel Scalper. Other EAs have been small winners, with the exception of XBars MA Scalper trading GBPUSD that ended a small loser with 3 wins and 1 loss (and the ratio between TP and SL is just over 3).

When I’ve been looking at the performance day to day I can’t but see that there is some portion of “bad luck” involved here, or at least bad settings.

This picture is from yesterdays session. First off we can see of course that it started trading a little too early. The market was drifting upwards the entire session though, so the trades should have been taken “badly” either way. The really unlucky part is in the timeout value of 4 hours thou, when if 4h30min would have been used instead all trades would have taken profit instead of closing for about three times the TP in losses. Three positions survived the peak and took profit here. One EA, trading with a experimental SL of just 11, just exactly hit that before the reversal.

4 hours instead of 4,5 is a little better in backtests thou, so I will stick with it. Hopefully the Market will behave more like I expect it the coming month.

There have been even some more bad news for this account thou, as JadeFX have announced a 33% increase in commission for GBP based pairs. As 85% or so of my trading on the account is in EURGBP and a few more percent is GBPUSD this is really bad news. The commission on these will be $8/lot starting February 14th. Perhaps it is time to look at some of the alternative brokers, which are quite few, but perhaps worthy of testing out either way. FinFX is a ECN broker that I’ve been demotrading with successfully. They have about the same spread as JadeFX and they will after this change have lower commission. A HotForex Currenex account is another possibility, this one has the advantage of requiring just $500 to start out live testing with.

But I really like JadeFX and the setup I have there with their ultra low latency VPS. We’ll see, I won’t make any hasty decisions.

Just a quick graph of the 6 months of trading that made up the second half of 2010 at JadeFX. The account have during this period exclusively been running scalpers and 85% of the trading have been by The Channel Scalper trading EURGBP.

Performance July 1st - December 31st 2010

The funny thing is that the graph above states about 104% gained when in reality 143% was gained. MyFXBook calculates something very differently here and I have reported it to them. Not that 104% would be shabby but the account really have 143% more dollars now and I haven’t deposited any. $150 have even been taken by JadeFX from the account to pay for six months of their VPS. Strange.

143% gives a average gain of 16% per month, which I am satisfied with. Over 12 months this would give about 500%

Update of my JadeFX accounts scalping performance under December. Prior years I have shut down EAs in early to mid December, but this year I decided to trade almost as usual. I ran quite a few backtests using Dukascopy tickdata and concluded that it wasn’t riskier to trade this month than any other (with my strategies that is, others might perform differently). Up until December 16th I thought I had made a excellent decision with a gain of 30% during the month, but that day and the also the following I lost about 17% of the account and was down at 7% gained during December. The latter half of the month added a little pips here and there and ended up at 12.53% gained.

December performance at JadeFX

The biggest losses came from XBars MA Scalper trading GBPUSD and RSI MA Scalper trading EURUSD. A version of settings for The Channel Scalper trading EURGBP also got hit hard. Prices was trending significantly during the normal quiet Asian market. Going forward I have made adjustments that I hope will reduce drawdown like this without penalizing profit taking during periods when prices are range bound during my trading hours.

A few weeks ago over at Donnas forum a thread discussing HotForex gained momentum. I didn’t think much of it at the time, there are so many brokers out there no one can keep up with all of them. And it is rare that they are anything worth thinking about either. Since then I’ve been checking the thread now and then and it seems people are pleased with their experiences. There are also very favorable reviews at Forex Peace Army, a site that usually is on the negative side (common human behavior to only report negative events of course).

The last thing that really spurred my interest was something I saw just today thou. They are now featured on CashBackForex and give 0.4 pips / lot back. I’ve asked how this applies to the Currenex account type at HotForex but still haven’t gotten an answer, hopefully it is valid regardless of account type (Update: Yes it was). The Currenex account is a ECN with $10 / lot traded ($5 per side) in commission, that with a rebate of 0.4 pips would be $6 / lot. The spreads and execution are supposed to be very good. Minimum deposit is $500 and minimum lot size is 0.1.

In order to get the rebate you have to register at CashBackForex.com of course and follow the instructions posted there. I’m thinking about checking it out.

November is over and it is time to look at the performance of my scalper EAs at JadeFX. It was the worst performing month since May this year with “just” 2.88 % gained. There we’re two massive loosing streaks on the 3rd - 5th and the 18th - 19th. In both cases news pushed prices out of the usual channels during the Asian session basically affecting all scalping strategies.

Performance outside of these dates was good. Risk taken is clearly quite high, but it is intended as a high risk account. I am experimenting with tighter stoplosses and adapted timeout values as a way of reducing the swings.

Most profits (and losses) are made by The Channel Scalper trading EURGBP, but the performance of XBars MA Scalper on GBPUSD this month is notable. It took profit in 9 of 11 trades (82%) and made 99.5 pips in total because of it’s sensible TP vs SL ratio of 1:3.

I’ve started updating the Forex Guide published here by me. It is a little bit dated now (untouched since August last year). Hopefully it will be more useful in a couple of days.

Stay tuned.

Just a short update with the trading result for October at JadeFX. The setup used is basically the same as the previous months, but I have been experimenting a little with timeout values (testing a change of TP to 0 pip after 150 minutes, which is still under evaluation).

18.65% up over the month with just one big negative day, Oct 7th, of -153 pips. So I must say that I’m pleased with this performance. It is very much inline with the two previous months (of 16% and 17% equity growth).

This is interesting. Apparently it is not OK if the small guys take advantage of a big guys faulty algorithms. Good to know since the markets are being manipulated on a second to second basis by HFT firms frontrunning and flash trading. We have to be careful so that our trades doesn’t make the computers trade badly, because it isn’t supposed to loose. If it does it is clearly because it was manipulated. Come on, this is ridiculous. Clearly the algorithm used wasn’t tested enough before being put in production. Thank the Norwegian guys for finding the flaws and fix them, don’t take legal action like a sore loser.

Check it out: http://www.computerworlduk.com/news/security/3244186/norwegian-traders-convicted-for-outsmarting-us-stock-broker-algorithm/

I hope my trades never “cheat” a big players software so I get sued too…