Ray Robot II™ Spread Betting Live Test – Alpari UK versus GKFX
We were just gearing ourselves up to gently launch Ray Robot II™ onto an unsuspecting world when Alpari UK announced that they had started offering MetaTrader 4 spread betting accounts. In contrast to some reports elsewhere on the web, here at the Trading Gurus we are well aware that automated spread betting via MT4 has been available for well over a year from another broker – the recently renamed GKFX. We figured we'd kill several birds with one stone by running RR2 in identical fashion on live accounts from both MT4 spread bet brokers. While we're at it we thought we'd also try him out on demo accounts from both brokers, to see if there is any noticeable difference between automated live trading and pretend trading at our small sample of MetaTrader brokers. Whilst we're at that we thought we'd also compare some backtest results with all those forward test results, to see whether that made any difference to the outcome(s).
Here's the experimental setup. The minimum deposit for an Alpari spread bet account is £250, so we're starting off with 4 different accounts, all funded with £250, although of course only 2 accounts are using real live cash. To start with we're trading the minimum possible size, which (coincidentally?) is 10 pence per pip at both brokers. We set four identical clones of the son of Ray the Random Robot™ trading daily bars of cable using his default settings. Here's how things looked earlier on this morning on Ray's two live accounts. Can you spot the difference?
If not here's the answer. The GKFX chart shows a small candle representing Sunday evening's trading, which is absent from the Alpari chart. We burned a bit of midnight oil last night, and we've already modified the code of the "robot" to take account of that fact. The next thing on the agenda was to wait and see which pending order got filled first. As things turned out it was the one just below Friday's range. Both Alpari accounts got filled at 1.58931 at 16:48, whereas both GKFX accounts got filled at 1.58940 at 16:47. Despite that slight difference all four accounts finished up the same way a few minutes later. The stop losses were taken out for -10.0 pips and a loss of £1.00 in every case. Not too many differences to be seen in the one and only trade so far then, but the expert adviser logs tell a rather different story. Lets take a look at the two live logs:
The first thing that stands out is that GKFX offer 100 to 1 leverage, already higher than most other spread bet brokers, whereas Alpari offer an eye watering 500 to 1! Hence the margin requirement of £159.01 for a £1 "bet" at GKFX as opposed to a mere £31.79 at Alpari. Also take a close look at the GKFX log at 15:48:57. When one of the bracket orders is filled the EA attempts to cancel the other one, but in this instance it failed. Whilst Ray spotted the problem all he did was print an error message. It looks like a production "robot" would benefit from rather better error correction, doesn't it?
Ray's full results for all four accounts are available for inspection over on myfxbook. If you would like to continue to compare notes with us, you can download the source code for Ray Robot the second from the Trading Gurus community forum and try him out on a demo account at a MetaTrader 4 broker of your own choosing. If you do that please let us know how you get on, by adding a comment to the forum thread devoted to detailed reports and analysis of the results of Ray's experiments.
Now let's see what tomorrow brings!
Comments on Ray Robot II™ Spread Betting Live Test – Alpari UK versus GKFX
Jim @ 9:22 am
Tomorrow has brought much the same story. Another failed breakout on the short side, this time in the small hours of the morning. Another £1 loss across the board, although on this occasion it took 29 mins for the stop to get hit on Alpari, but 56 mins on GKFX.
See myfxbook for more details.