Trade ES with Tradesation

[feng2088 10]

Hello guys,

 

Do you guys know if there is any way I can trade ES with different strategies in one acct? Is that true Tradestation won't let you trade the same vehicle with different strategies? I am talking about automated trading. Thanks

[UrmaBlume 10]

TradeStation will "allow" you trade any number of strategies on any number of symbols in any number of accounts.

 

The issue is that if you are trading the same symbol with more than one strategy in the same account - then entries from one strategy can offset positions from another.

 

The answer is that in TradeStation it is very easy to clone your account and then link the different clones to the same platform and then you can have as many "independent" positions as you have clones.

 

I have six linked accounts at TS so I can have 6 different strategies all trading the ES and each is independent of the other. I run all 6 accounts from the same TS platform running on the one workstation.

 

Did you really think that you could be both long and short in the same instrument in the same account?

 

 

UrmaBlume

[feng2088 10]

Thanks UrmaBlume. So you are saying I don't physically need to open 3 accts if I wanna trade 3 strategies. All I need is to clone 2 accts and link them with the real acct? How do I do that? If I only have $6k in my acct, is it still doable? Thanks again

[UrmaBlume 10]

Thanks UrmaBlume. So you are saying I don't physically need to open 3 accts if I wanna trade 3 strategies. All I need is to clone 2 accts and link them with the real acct? How do I do that? If I only have $6k in my acct, is it still doable? Thanks again

 

If you only have 6k in your account and you are asking such questions - you probably should be trading on the simulator.

[feng2088 10]

I can always put more money in. I am just asking if it's doable.

[UrmaBlume 10]

I can always put more money in. I am just asking if it's doable.

 

I believe there is a 5k minimum for futures accounts at TS.

 

Clones are actually new accounts with different account number but with the same CustomerID.

[feng2088 10]

Thanks again. What do you think the 5-year performance of my 3 strategies? You think I can go live yet? I started coding my strategies 4 months ago. I am still relatively new to automated trading. So I really wanted to know how other people do.. Thanks!

Edited August 8, 2011 by feng2088

[RangeExpansion 10]

Thanks again. What do you think the 5-year performance of my 3 strategies? You think I can go live yet? I started coding my strategies 4 months ago. I am still relatively new to automated trading. So I really wanted to know how other people do.. Thanks!

 

Numbers look fantastic. But we can all make reports look fantastic with heavy optimization that doesn't hold up the real world trading.

 

What portion of your results represents trades made on out of sample data? Did you walk forward, and backwards? It's a mistake a lot of newcomers overlook. You're not going to get away with zero slippage in real market conditions either. I would put in a tick of slippage to add some pressure. "Look inside" also needs to be checked for more accurate results.

 

Best,

Gabriel

[feng2088 10]

Thanks Gabriel . I use at least one limit order for each trade. So I do have at least 1 tick slippage. Also I believe some of my orders may not get filled in real life. My overall profit should be lower than what I posted.

[RangeExpansion 10]

Thanks Gabriel . I use at least one limit order for each trade. So I do have at least 1 tick slippage. Also I believe some of my orders may not get filled in real life. My overall profit should be lower than what I posted.

 

Limit orders can be tricky, especially in the ES where price typically is required to trade through your limit price before a fill can be achieved. Good to see you're paying attention to the performance disparity that live conditions can produce.

 

Did you perform any walk-forward or walk-backwards testing on your trading system?

 

Meeting certain performance criteria on "out of sample" data should be one of your primary considerations before going live.

 

It's not easy that's for sure.

 

Best,

Gabriel

[feng2088 10]

Did you perform any walk-forward or walk-backwards testing on your trading system?

 

Meeting certain performance criteria on "out of sample" data should be one of your primary considerations before going live.

 

It's not easy that's for sure.

 

Best,

Gabriel

 

Do you mean to test the strategies by randomly picking a historical time period or using real live data? If yes, then I did. I can tell you that my strategies actually perform better in the past 5 yrs than in the past 10 yrs. I also observed how they traded with the real live data.

 

When you refered to "out of sample" data, did you mean how well it can handle the contingency on crazy days like today ? My system did not generate any trades today. I know it's sad. It would be a nice ride. But since it gapped down "so much" at open, I think it's not worth to take the risk. Yup..some might think I am stupid.

 

P.S.--I backtested them with @ES.D only and I incorporated a stoploss for every of my strategies.

 

Thanks!

[mojoe 10]

In your strategy report, you need to look at account size required. This is very important, because you have to assume that this is how much draw down you will experience. i Have seen wonderful performance results. But the account siize required is like in $10k range. Well, that is too risky because when you loose $10k, it is hard to figure out if your program is still running or it has failed.

 

Yes, try changing all your order entry to market order and then see what is the result.

 

OR choose this option in tradestation setting:

Format strategy:backtesting:

 

use this setting: Fill entire order when trade price xceeds limit price.

 

This will take care of fill issuse. This mode will make sure there is 1 tick better for all entry and exit.

 

You need to do at least 10 years because last 5 years are not good enough.

[Chippy 10]

nothing works until you trade it live...trust me lol

[eqsys 10]

I have two critical questions about the strategy performance reports:

 

Where are the MOST important stats: the drawdown calculations?

How many variables were used to generate these results?

 

The max intraday drawdown WILL absolutely happen to you in real time trading, so you have to have enough cash in the account so that when this happens it won't wipe you out.

 

If you have too many variables in the system then you can be sure that the results were due to curve-fitting and it's unlikely you can achieve them in real time trading.

 

If these results were generated using non-standard chart types, such as range or momentum bars and you used minute intervals, backtesting will NOT work in real time trading because the way these non-standard charts are drawn historically is not how they work in real time.

 

Finally, with any system like this you absolutely should trade simulated before actually committing real money to the system. You might find that the system doesn't work in real time, for one reason or another. If you can successfully sim trade for a while, then you can set the system to live auto trading. Better to find out with a sim account than to lose real money.

[procol33 10]

ninja trader professional version will let you trade multiple strategies on the same instrument in one account. Saw a webinar yesterday and that specific feature was mentioned ....

Lg