Automated Trading Championship 2010

[Andy Nguyen]

Automated Trading Championship 2010

MetaQuotes Software Corp., Interbank FX LLC, MIG Bank, FXCM (Forex Capital Markets LLC), the TRADERS' magazine and Dow Jones hold the fourth annual Automated Trading Championship 2010. The best three developers of automated trading strategies will be revealed within three months. Assessment criterion will be the absolute profit. Winners will share the Automated Trading Championship 2010 Prize of totally 80,000 US dollars!

The championship will begin on October 4th and will end on December 24th 2010

 

[Eugene Krel]

mmm absolute profit.

Something along the lines of:

if(flip_coin()) buy
else sell

Should work just fine

 

[D Serg]

Does somebody know what is the difference between automatic and algo trading?

 

[Sarang]

Isn't the criteria 'absolute profit' a bit too hard? I mean, algorithmic trading helps you in locating arbitrage opportunities and off-priced securities, and the profit typically will be small.

I have an algorithm which, if deployed on a correct server farm, would generate returns of like 1% a month, which could be meager, but it's risk free! And it's still a lot more than what a savings account might pay.

Shouldn't risk be a consideration for a robust algo strategy?

 

[mercury]

According to my experience the concept of total profit alone often select a strategies that has a high risk profile (with the most likely "next event after the test" being losses) and I was wondering if that is also your experience.

 

[alain]

According to my experience ...

and what is your experience?

I probably agree with your but I'd like to know your experience in this field to weight your opinion.

 

[dhride]

The notion of absolute profit is very familiar to me as I only get a percentage of my profits as retribution for my work as a prop trader. If the profits are not there then bye bye New York city luxury apartment.
In my world the ability to manage risk and constantly generate profits are two concepts that cannot be separated and ultimately they will be judge only on your performance. A poor performance might mean that you have poor risk management or that your strategies are crap, good performance means that your strategies can really leverage themselves while controlling the risk. On that note, I think giving the first prize to whoever makes the most money is the best way to judge the participants.

I run 3(boxes) intraday strategies, a MM stat arb strategy, a pairs strategy, and a momentum strategy. If I have a number of bad trades, or many consecutive bad trading days, I tend to leverage myself more as I have a greater probability of turning a profit the upcoming trades/days. That's why I agree with Mercury, most likely the winner will have a high risk strategy that towards the end of the championship has a greater probability to run into losses.However, if losses are expected you can always minimize their effect.

---------- Post added 08-17-2010 at 12:14 AM ---------- Previous post was 08-16-2010 at 11:56 PM ----------

On another note, I read the rules to the contest and I found that it lacks depth to be a robust an interesting contest, lacks definition on what constitute and automated trading system
No scalping allowed!, that takes care of High frequency traders.
"Automated trading systems find their purest form in high-frequency strategies where speed of calculation and execution are the competitive advantage" -Benjamin Van Vliet - Building automated trading systems-

Reminds me of the GRE, it is probably easier to find a way to beat the test than to learn the material, find how the demo/sim machine executes/marks the trades and game it!

Many years ago, a [not automated] trading tournament sponsored by a university had the sim server give you the last price when you sent a market order. I wish the Direct Edge and BATS did the same....ha.

 

[mercury]

If we agree that total profit is just an easy way to select a winner (and that it is relatively easy to put together something that generates scads of money in a vacuum - meant regadless of the drawdowns) then should not we also point out that the real challenge is to put together a strategy that makes a lot compared to the drawdowns.

And just to tell the entire story, that really captures the essence of the markets, by having similar performances in different markets (since none can recognize a 10 min. silver from a weekly oil or daily google, etc, provided that you strip away price and time).

However, it's kind of interesting how life compesate for some "human errors". Although I wish the winner every success, I would not be surprised if - in the long run - the one who will REALLY make the most money is some guy that went unnoticed, since he did not scored high enough... (profit wise...) and sad to say - but true to best of my knowledge - the winning strategy, if "unattended", is never going to make that guy rich (but most likely put him out of business...)

It may sound crazy, but u do not need to take my words for it. It's in the Bible: " 1st will be last..."

 

[euroazn]

Isn't the criteria 'absolute profit' a bit too hard? I mean, algorithmic trading helps you in locating arbitrage opportunities and off-priced securities, and the profit typically will be small.

I have an algorithm which, if deployed on a correct server farm, would generate returns of like 1% a month, which could be meager, but it's risk free! And it's still a lot more than what a savings account might pay.

Shouldn't risk be a consideration for a robust algo strategy?

If your one percent a month is really risk free, then you should be able to make an infinite profit by shorting treasuries and going long your strategy.
Just saying.

 

[Sarang]

Infinite profit?

Theoretically, yes.

But it is constrained by:

1. Volume/liquidity of the trading underlying. I'm not dealing with liquid products. The strategy configuration I have ignores impact cost of my trades, which would play a major role if I up the volume. I'd hence be restricted to small-ticket trades.

2. It's kind of like arbitrage - the moment you fill up an opportunity, the opportunity is lost and you have to wait for a new one.

In essence, it would work only on small amounts and you'll have to wait for the market to open up small gaps for you to fill. I agree it's not suitable for large banks or hedge funds, but perfect for managing my household investment, as it gives me a greater-than-checking-account returns.

In either case, I don't think the competition allows you to play with options, so that's a bummer!