Multiple systems at the same time

[elit]

Hi, do you combine many differerent strategies in your trading?
In other words do you trade many different strategies at the same time?

Like you have a couple of strategies that, on their own would not give signals frequently enough, so you when you trade some other strategies at the same time, you are then getting signal as often as you are comfortable with.

Or do you find it better to only concentrate on one single strategy at the time.

 

[yoohoo]

I have shown how I use different strategies to scalp the futures in the Interesting Charts thread.

You might find some ideas from myself or Goinglite that help.

http://www.elitetrader.com/vb/showthread.php?s=&threadid=76893

 

[elit]

Thanks, I'll check it out!

 

[horribilicus]

Quote from elit:
... you have a couple of strategies that, on their own would not give signals frequently enough, so you when you trade some other strategies at the same time, you are then getting signal as often as you are comfortable with.

Another rationale for trading multiple systems is: diversification. With multiple systems you can have diversification of

• Trading frequency (avg hold time + avg flat time between trades)
• System philosophy (pullback buying, trend following, chart patterns, oversold fading, support+resistance, seasonal trading, sector rotation)
• Entry and/or exit style (mkt orders, stop orders, limit orders)

However some people think diversification is stupid; they call it diWORSEification and they (correctly) point out that the surest way to make a lot of money in investing is to hold an extremely concentrated portfolio and be right. Just buy Microsoft at its IPO and don't ever sell; who needs diversification?

 

[aus_SPIder]

trading multiple uncorrelated strategies in different time frames and over 2 or 3 markets will dramatically reduce volatility of your equity curve. i trade 5 different strategies at teh same time, 2 Position trading strategies avg hold time 2mths. 1 intermediate strategy: hold time 1mth. 1 swing strategy: 1-2 weeks and 1 countertrend strategy that works in all 3 time frames.

Most are lightly correlated between -0.2 and 0.5 in correlation. Thats due to the different time frames as what constitutes a "whipsaw" in a 2mth hold period usually turns out to be a profitable trend using a swing strat and a countertrend strat.

 

[elit]

Ah ok, yes I didn't even think of the diversification. That's a good point.

Do you incoroprate some sort of ranking of the strategies; how do you choose what strategy to trade if two or more of your strategies gives you signals at the same time? Do you ever close a position only to open a position based on a signal from another strategy?

Thanks for your replies

 

[aus_SPIder]

each strategy has its own account. I end up paying more brokerage but it avoids the problem of conflicting signals. If its conflicting then ill just be long in one account and short in the other, hedged position.

 

[elit]

Do you use different brokers or multiple account at one broker?

 

[man]

Quote from elit:

Hi, do you combine many differerent strategies in your trading?
In other words do you trade many different strategies at the same time?

the more uncorrelated strategies the higher the
overall return/risk figures. period.

 

[icash55]

Yes, I do, and I keep adding new systems to my portfolio as they are developed. One system may be long and other short at the same time for the same instrument (if 1st system is long then the other system would reduce/reverse the position by selling X units, which would be later bought back to close the trade). More promising systems are generally given more % of capital (leveraged) to trade, but that also depends on avg profit/loss. All systems use the same account and if one system is doing well, the other systems will also increase the trade size (using "shared" compounding) as the capital increases or the opposite.

I think trading multiple, not too highly correlated, systems is better than trading just 1 system and exploiting 1 "edge", which later might turn out to be not an edge at all or soon stops working. + other benefits, already mentioned.