System with lack of historical data

[Dhalsim]

Can someone advise on the following system i have created for Russel (TF) on 1min chart data and RTH only.

Limited to one trade per day. Averages 20 trades or more per month.

I don't know whether to take this strategy live because with Russel on ICE exchange i only have data from 2009 on-wards which is usable. Prior to 2009 market was too thin on this symbol and as my system is price action system it cannot work or be tested on thin market data.

My dilemma is that the system performs well and is long only, and we all know since 2009 the market has been up continuously in one massive bull market.

One half of me thinks this should not make too much difference because strategy is very low exposure and short term based on intraday 1min charts and i have checked performance during down periods in data from 2009 on-wards and it seemed to perform through most down-periods.

The profit factor is not where i want it to be but because it trades very often and average trade is 60 bars then i guess i can't expect massive PF. The Sharpe ratio is great and draw-down is great.

Anyway please advise guys. Ideally i want to test this on more data for Russel prior to 2009, i know Russel moved from CME to ICE - do you recommend testing on CME data?

Please see attached file:

 

[lindq]

Quote from Dhalsim:

Anyway please advise guys. Ideally i want to test this on more data for Russel prior to 2009, i know Russel moved from CME to ICE - do you recommend testing on CME data?

I used to trade the contract when it was RT. It was a great instrument at that time, and really disappointed many traders when it was sold.

You are correct to be somewhat suspect of any system trading only the long side of Russell 2000 derivatives since 2009.

Even though you are talking short time frames with your system, a strong bull market like we've experienced can put a safety net under any long system.

Your obvious choices with TF are to attempt to locate RT futures data from previous years, and/or to continue forward testing and/or trading in real time with small lots while you build confidence. When we hit a correction, just stay small with good system stops in place and be prepared to take your wacks with the rest of us.

My only other advice would be for you to subscribe to TF real-time data, if you haven't already, and watch the volume and spreads. I don't track it, preferring TNA, but have heard some reports that it can get thin, which could be a danger in a bad market. And if you have the cash to handle it, TNA is a nice instrument to follow the Russell. You might want to put TNA data through your system. (The drawback being that it too has only existed since late 2008.)

Good luck!

 

[d08]

Quote from lindq:

I used to trade the contract when it was RT. It was a great instrument at that time, and really disappointed many traders when it was sold.

Do you mean ER2? Used to trade it back then too, it wasn't the same after it started trading as TF.

 

[jeb9999]

Quote from Dhalsim:

Can someone advise on the following system i have created for Russel (TF) on 1min chart data and RTH only.

Anyway please advise guys. Ideally i want to test this on more data for Russel prior to 2009, i know Russel moved from CME to ICE - do you recommend testing on CME data?

Almost any system you come up with will stop working at some point in the future. The market changes and the system dies.

Just trade your system until it stops working.

You may find out that with real dollars it doesn't work at all.

 

[lindq]

Quote from d08:

Do you mean ER2? Used to trade it back then too, it wasn't the same after it started trading as TF.

Yes, my mistake. It was ER prior to being sold.

 

[Sergio77]

De-trend the data and see if it works. Then you could also try the system on E-mini data. I think if it does not work well there it may be fitted but not necessarily. Did you include sufficient slippage? See what happens to intraday E-mini trading according to this randomization study: 100% ruin of all systems.