Sid the Superstitious Robot

[Jim]

Today will go down in the history books. Today is the day that Ray the Random Robot meets a new member of his extended family. You too can now meet Sid the Superstitious Robot. Here he is:

SidMT4.zip

For the full story of how Sid came to be born please see this post on the Trading Gurus blog.

On the design side of things, the concept for Sid and his superstitious algorithm were created by Shing Tat Chung at the Royal College of Art.

On the technical side of things, the MQL4 programming was done by Jim Hunt at his laptop. Please also note that, amongst several other things, Sid the Superstitious Robot demonstrates how to read from and write to a (potentially remote) MySQL database using MQL4.

On the legal side of things please note that unlike Ray and the other robots available for download from this fourm, Sid is released under the GPL 3 open source licence, and comes with no warranty whatsoever.

[Dr_T]

Hi Jim,

Thanks for posting this, very exciting times! I just downloaded SID, everything seems to run with no errors, however I also don't get any trades on the MT4 backtester. Any thoughts? I've unchecked the UseMySQL option and not copied the dll for now.

When I find a little more time, perhaps I will revive the RAI fork of Ray...

Cheers,

Dr T

[Jim]

Hi Dr. T,

Thanks for posting this, very exciting times!

Quite so. Ray and I are so excited by what Sid's getting up to in town that we can hardly sleep!

I just downloaded SID, everything seems to run with no errors, however I also don't get any trades on the MT4 backtester. Any thoughts? I've unchecked the UseMySQL option and not copied the dll for now.

It may be a lingering "feature", in that nobody's tested that particular configuration for quite some time now, as far as I'm aware. Could you possibly zip up Sid's tester log and upload it here?

When I find a little more time, perhaps I will revive the RAI fork of Ray...

Sounds exciting also!

Thanks,

Jim

[Dr_T]

Hi Jim,
Where am I likely to find the tester log? I found a couple of logs with not very much in them.
Cheers,
Dr T

[Jim]

Hi Dr. T,

It is to be found somewhere along the lines of C:\Program Files\MetaTrader 4\tester\logs

Zip up the most recent log in there, however big the file is.

Thanks,

Jim

[Jim]

Sid is officially world famous!

David Wolman (author of "The End of Money") has mentioned him in a long article for the BBC. Unfortunately (if you live in the UK at least) it's on their "international" website.

Sid also got a mention on ABC News recently. Shing and his "superstitious fund" were featured on CNBC a couple of weeks ago, on Friday 13th, although sadly they didn't ask him about Sid.

What I find interesting is that whilst some commentators can evidently click a link or two, and even give Sid a mention along with the likes of Gann and Bayes, others say "There isn't much information on how trades are made". Maybe the final commentator isn't very comfortable interpreting "C like" source code?

[matka]

Sid in Keiser Report http://www.youtube.com/watch?v=UPW8zJmKlJI
Well done Jim!

[Jim]

Thanks for the link Matka, and for your kind words. I'm not a regular follower of Max, so I doubt I'd have spotted it unaided.

Sid can also now add Jamie Dimon, Paul Krugman and Denis Diderot to his list of well known associates!

Unfortunately Max seems to be labouring under the misapprehension that Sid was created on Wall Street. I guess he isn't very comfortable with clicking links or interpreting "C like" source code either?

[Dr_T]

Well done SID... and Jim of course!

How did SID perform in backtesting when he was developed? Based on this, do you have any quantitative hopes or predictions for how he 'might' perform? I'm not asking you to commit to anything, it might be interesting for gauging how well the fund is performing relative to past performance.

Cheers,
Dr T

[Jim]

Thanks Dr. T, though I think we all have Shing to thank for the mass media publicity.

I can't claim to have done a vast amount of backtesting, just enough to try and squish any obvious bugs. Since Sid generates his own "lucky numbers", every test will give different results depending on the start date of the test and/or what's already in Sid's database. In practice I'd expect Sid to lose money on average, since he's not consistently trying to exploit any "market inefficiency" that makes any sense to me. A bit like Ray the Random Robot but without the Martingale position sizing?

However given that his performance on his live account is but a single sample out of a myriad possibilities, maybe Sid will get lucky over his first year and prove me wrong? Alternatively perhaps my ideas about what "makes sense" are sadly lacking where "superstitions" are concerned!

Cheers,

Jim