Search
× Search
Tuesday, November 19, 2024

Archived Discussions

Recent member discussions

The Algorithmic Traders' Association prides itself on providing a forum for the publication and dissemination of its members' white papers, research, reflections, works in progress, and other contributions. Please Note that archive searches and some of our members' publications are reserved for members only, so please log in or sign up to gain the most from our members' contributions.

Do you combine strategies which trade common assets?

photo

 Igor Bilyk, Open to remote collaboration as a data scientist

 Thursday, July 18, 2013

For example: arb strategy 1 trades EUR_USD, EUR_JPY, USD_JPY arb strategy 2 trades EUR_USD, EUR_GBP, GBP_USD both strategies trade EUR_USD. Would you treat them as separate strategies or combine in one aggregated strategy?


Print

5 comments on article "Do you combine strategies which trade common assets?"

photo

 Andreas Clenow, Chief Investment Officer, ACIES Asset Management

 Sunday, August 4, 2013



My preference is to treat them as separate trading models and let them work independently. You can analyze and aggregate them on a higher logical level of course.

I run multiple iterations of trend models as well as several types of counter models at the same time, for the same mandates, trading the same asset universe. At times they can disagree and take opposite sites of the same bet. That's not a problem. It's the net return you're after.

Whenever given a choice, I separate into different models. Instead of having a single model which increases and decreases sizes, I use different models. Net result is the same, but the logic is cleaner and easier to analyze.

I find that you end up with a logical mess if you have too complex rules. By separating it into different trading models you can easier evaluate each rule set independently, both pre and post trade.


photo

 Christopher Reeves, Programmer at A2X Capital

 Saturday, August 10, 2013



I would keep them seperate because then you have to determine which one takes precedence over the other, and if you are going to do that you might as well just run the best one.


photo

 Fabian Kostadinov, Researcher and Partner AVACO AG

 Friday, August 16, 2013



@Clenow:

Can you say something about how you aggregate the separate models? What do you exactly mean by "aggregating" them, or how do you do this?


photo

 Andreas Clenow, Chief Investment Officer, ACIES Asset Management

 Friday, August 16, 2013



By aggregate, I just mean that you net the positions.

So, your trend following model is short European Rapeseed with 100 contracts, and your counter trend model thinks the price move is over done in the short term. The counter model puts on a long of 50 contracts for a short term bounce trade. Your net, or aggregated position is now -50 contracts.

In the internal books you're short 100 on one model and long 50 on another, but your aggregated position is simply short 50.

The advantage of keeping them separate is that you can analyze the trend model and the counter trend model separately or analyze the net overall result.


photo

 Fabian Kostadinov, Researcher and Partner AVACO AG

 Friday, August 16, 2013



Ok, so this also means that if both models generate trading signals, you don't aggregate those signals, because this would not make much sense. Instead, you simply place orders according to both models and then in the end sum up your wins/losses.

Please login or register to post comments.

TRADING FUTURES AND OPTIONS INVOLVES SUBSTANTIAL RISK OF LOSS AND IS NOT SUITABLE FOR ALL INVESTORS
Terms Of UsePrivacy StatementCopyright 2018 Algorithmic Traders Association