Rounds Turns Per Million -- How Low Can You Go?

Discussions about the testing and simulation of mechanical trading systems using historical data and other methods. Trading Blox Customers should post Trading Blox specific questions in the Customer Support forum.
Post Reply
Algonquin
Roundtable Knight
Roundtable Knight
Posts: 158
Joined: Sun Jan 28, 2007 9:38 pm
Location: Greenwich, CT

Rounds Turns Per Million -- How Low Can You Go?

Post by Algonquin »

I have developed a few very long term trend following systems, with solid performance characteristics. I developed the systems using 50% slippage assumptions, and thus gravitated toward systems with extremely low trading frequencies. Most of the systems trade between 200-250 round turns per million. I have tested these systems over 30+ years, so I feel like I have a sufficient statistical data set (1400 trades for only 3-4 degrees of freedom). But I still have a gnawing feeling about such low RTPM figures, especially given that lowest publicly disclosed figures appear to be in the 400-600 range, with most being significantly higher than that.

I don't think this topic has ever been addressed in this forum, so I thought I would raise the issue with my fellow members. How low can you go with trading frequency? Does it really just distill down to the total number of historical trades in your data set viewed against the system's degrees of freedom. Or does the overwhelming use of much more frequent trading systems by professional CTAs suggest that something is amiss with that conclusion.

I ultimately believe the former, but would appreciate hearing any thoughts or comments.

Algonquin
Mathemagician
Full Member
Full Member
Posts: 20
Joined: Thu Jun 28, 2007 1:46 pm
Contact:

Post by Mathemagician »

Are you counting rollovers?

jj
Algonquin
Roundtable Knight
Roundtable Knight
Posts: 158
Joined: Sun Jan 28, 2007 9:38 pm
Location: Greenwich, CT

Post by Algonquin »

Yes, roll overs included.
AFJ Garner
Roundtable Knight
Roundtable Knight
Posts: 2071
Joined: Fri Apr 25, 2003 3:33 pm
Location: London
Contact:

Post by AFJ Garner »

I believe that you should follow the devices and desires of your own heart (and the results of your own tests) and ignore slavish convention. Actually, tangentially, there has been a lot of discussion on the topic; or at least on the subject of very long term systems.

Several traders have avowed their lack of concern with 1) the low trade frequencies involved and 2) the knock on question of the statistical validity of backtests of such systems.
Mathemagician
Full Member
Full Member
Posts: 20
Joined: Thu Jun 28, 2007 1:46 pm
Contact:

Post by Mathemagician »

Is your leverage lower than theirs?

jj
Algonquin
Roundtable Knight
Roundtable Knight
Posts: 158
Joined: Sun Jan 28, 2007 9:38 pm
Location: Greenwich, CT

Post by Algonquin »

Not really. If anything, geared slightly higher (CAGR=20%+).
Mathemagician
Full Member
Full Member
Posts: 20
Joined: Thu Jun 28, 2007 1:46 pm
Contact:

Post by Mathemagician »

Algonquin wrote:Not really. If anything, geared slightly higher (CAGR=20%+).
Be sure to remember that published figures include fees. They're likely leveraged more aggressively than you think.

jj
jankiraly
Roundtable Fellow
Roundtable Fellow
Posts: 52
Joined: Tue Feb 03, 2004 7:23 pm
Location: San Diego

Post by jankiraly »

You could put high-leverage and low-leverage trading plans on an equal footing by calculating (RTPM / Margin_to_Equity_Ratio) since the ME Ratio is usually published in professional manager's track records.

A word of caution, very low RTPM means your stops are very wide. This means your DollarRisk per contract is very large. You will need a very large number of dollars in the account, to be able to take all of the signalled trades. It sounds like you are a brand new futures trader, do you really want to allocate a quarter- or a half-million dollars to futures trading, as a rookie?
AFJ Garner
Roundtable Knight
Roundtable Knight
Posts: 2071
Joined: Fri Apr 25, 2003 3:33 pm
Location: London
Contact:

Post by AFJ Garner »

jankiraly wrote:A word of caution, very low RTPM means your stops are very wide.
Yes, although you can still get the benefits of trading very long term if you are willing to narrow your stops in a bit and accept a few more trades and a slightly different system profile. This will mean you can trade such a system with less cash.

Say you are testing a 300 day MA with the Bollinger Band breakout system: you may see pleasing and acceptable results using 1+ std dev bands rather than 2 std dev. Or see whether the use of an 80 or 100 day calculation for the std dev (rather than the 300 day length of the moving average of the close) helps. Or on a 300/50 day DMA system, experiment with a 5 or 8 multiple of the ATR stop for bet sizing rather than 10+ and so forth.

Unsoweiter.

Another avenue to explore is expanding the portfolio to monster size coupled with suitable group (and/or other) limits.

Even with very wide stops, you may find enough low volatility trades (even on the bigger contracts) to enable the use of a lower account size.
Post Reply