Forex MegaDroid Hits a Losing Streak
- 5 Comment
Over at Trading-Gurus.com we’ve been running tests on Forex MegaDroid since it launched with the assistance of an army of Clickbank affiliates and a vast quantity of hype at the very end of March.
I don’t know if you remember, but this Droid (actually a MetaTrader 4 expert advisor) can apparently “see into the future” with the aid of some new “artificial intelligence technology” confusingly entitled “Reverse Correlated Time And Price Analysis”. For comparison purposes we’ve been testing it alongside a robot of our own, which doesn’t use any intelligence whatsover. In fact it’s so dumb it does everything wrong! It uses random entries, then it stays in the market 100% of the time and finally it uses Martingale style position sizing. In addition to this long list of “faults”, you can also download our “Robot” from our website completely free of charge, solely for educational purposes.
Here’s how the “Droid” has been doing throughout April and May:
By way of comparison, here’s how our “Robot” has been getting along over the same time period:
To summarize those equity curves:
- Both accounts started with $500.00 on April 1st
- By the end of April our Robot had reached $742.52. By the end of May that had increased to $961.00
- By the end of April the Forex MegaDroid had reached $573.76. By the end of May that had DECREASED to only $564.82
Which would you rather be using? Before you answer that loaded question consider these points:
- The Droid was used with the default input settings, which amongst other things means a fixed fractional position size.
- Other money management systems are apparently available, but the Droid’s manual doesn’t give you any guidance on how to use them successfully. We stuck with the default.
- Maybe a bigger position size or different money management would have assisted the Droid somewhat?
- Our Robot uses the much derided Martingale style position sizing. It hasn’t blown our account yet. Maybe it will one day soon?
- These tests have been performed on demo accounts, not on live accounts using real money.
Our Robot comes complete with source code and a Wealth Warning. It should NEVER be used on a live account under any circumstances whatsoever. That blowout might be just around the corner. Based on these results maybe the Forex MegaDroid should be accompanied by a similar warning? It’s clairvoyance does seem to have been rather short lived, doesn’t it?
Jim Hunt
Trading-Gurus.com
5 Comments on this post
Trackbacks
-
Adam Shwartz said:
I always thought that these ‘commercial’ robots are a compelte fraud, another smart Ponzi scheme model. This might prove my point.
Thank you,
AdamJune 3rd, 2009 at 4:22 am -
Jim Hunt said:
Hi Adam,
Thanks for your comment, and it’s my pleasure! As the old saying puts it “One swallow doesn’t make a summer”, and one bad month doesn’t necessarily make a bad system. In fact for me, it’s highly suspicious if a system never seems to have a drawdown. For the MegaDroid, however, it’s also suspicious that the “live” results on their site don’t show the drawdown evident in our own tests.
Jim
June 3rd, 2009 at 5:43 am -
doesntmatter said:
you are using wrong gmtoffset,
anyway it just shows your a complete novice you recommend a system with drawdown of 80% instead of a system which many traders use on their live accounts. your bot will wipe out the account in no time.if i backtest may i get profit on alpari uk with gmtoffset “1″ as recommended in manual for backtesting.
you just suck in what you are doing.June 19th, 2009 at 6:54 am -
Jim Hunt said:
Hi Doesn’tMatter,
Thanks for your comment, and for the one on my blog. Our backtest used the GMT offset settings recommended for our broker in version 1.11 of the Forex Megadroid manual. What do you think we should have used instead?
I’d love to take a look at some more impressive results for the Droid in May. Can you please provide a link?
How do you come to the conclusion that our “bot” has a drawdown of 80%? How do you come to the conclusion that I recommend trading that system? It does after all come complete with the warning that “This expert is for educational purposes only. It should NEVER be used on a live account.”
What I actually recommend is that you should D.Y.O.R. (and lots of it) Keep up the good work!
Jim
June 19th, 2009 at 7:47 am -
AllianceForexProvisions said:
Amazing the things you see pop up on blogs and websites about supposed theories on market forecasting.
A) There is no known form of true Random in nature.
B) The markets are anything but random, but due to the manner by which market makers correlate trade operations to drive the markets, they are a continuum of repeat moves, altered by influences of migrating momentum which “Skews” the “appearance” of the market vertically and horizontally, making it “seem” random, to the eye.
C) Conventional statistical process analysis cannot work in trading as the amount of range between the least significant values, (ticks), makes reliable child sampling unable to predict accurate parent projection.
D) Megadroid claims it does not rely on prior data in it’s forecasting element. Sorry, there is nothing but prior data to use for forecasting. That statement proudly displayed in their offering is truly humorous, especially from any true AI perspective.
E) Reacting to and acting upon changes in market conditions, is a valid pursuit but in “most” conventional cases the degree of human analysis required, versus what the MT4 can offer reliably… Not any day soon, sorry.
F) Market forecasting is the same as any other “Attribute” related sampling / forecasting, but it does not tell you, nor can you ever know, “the next tick”. You can only measure and act upon derived “probabilities” which is a very successful attempt, until one fails to account for the “extraordinary” events in the market, such as news extremes and other even less timely events.
G) There is truly only ONE order type in Forex, which is the “Limit” order. All other forms of order types are “contrived inventions” operating downstream, subject to both slippage and manipulation.
H) Given G above, any system which ever relies on order types other than limit orders… can never be deemed reliable or “in control”. That does not mean they cannot make a profit quite reliably, only that the reliability is subject to devastating exposures and given probability those exposures WILL occur.Conclusion, among the empirical truths above, one now has all the necessary information to design a system which can capitalize on the methods available and avoid the pitfalls that can destroy success.
Do the work. Good luck.
AFP
February 7th, 2010 at 4:20 pm
