LONG vs SHORT Trades (1 System or 2)

[Chorlton]

Hello All,

I wasn't too sure where to post this thread but hopefully its in the right location.

I've been developing a mechanical system and to date have only focused on the LONG Trades but have now started looking at developing the Short side of the system.

However, what is beginning to become very clear is the fact that one can't simply "invert" (for a better word) the Buy Signals to produce adequate Sell conditions.

When I backtest the "short" side only, the results are extremely disappointing. I do however, have other ideas which I believe will offer a better performance once backtested for trading the Short side.

However, my question is this:

As the "Short" signal conditions will be totally different to those for generating the "Long" Signals, should I still try to develop the ONE system with both Long & Short signals or should I simply just design another system to trade the Short side only??

What advantages / disadvantages are there will both approaches? I can think of a few but would welcome comments from others......

Many Thanks in advance,

Chorlton

 

[Trader333]

However, what is beginning to become very clear is the fact that one can't simply "invert" (for a better word) the Buy Signals to produce adequate Sell conditions.

I had a similar issue a few years back and what I found was, (as a rule of thumb) that markets go down 3 times faster than they go up with the exception of forex.


Paul

 

[Chorlton]

I had a similar issue a few years back and what I found was, (as a rule of thumb) that markets go down 3 times faster than they go up with the exception of forex.


Paul

Hi Paul,

Good Point !!!

My "BUY" conditions are based purely on Price / Volume relationships but what is interesting to note so far, is that these specific relationships are not so reliable when looking to go Short, which has really surprised me.

One thing this has taught me is the importance of backtesting ones ideas...

Regards,

Chorlton

 

[barjon]

chorlton

do you have a set-up with associated conditions?

if so, is it the set-up that's failing or the conditions that are causing failure? you should be able to test that.

i find a difference in the significance of my associated conditions long side to short side.

cheers

jon

 

[Chorlton]

chorlton

do you have a set-up with associated conditions?

if so, is it the set-up that's failing or the conditions that are causing failure? you should be able to test that.

i find a difference in the significance of my associated conditions long side to short side.

cheers

jon

Hi barjon,

Could you elaborate on what you mean by both "set-up" and "conditions" as I would have thought they were part of the same thing.

Sorry for the probably stupid question

 

[barjon]

chorlton

I meant the primary pattern set-up (eg: long on break of first hour's high) which would trigger a trade and then the associated conditions which would determine whether or not the trade was to be taken (eg: volume must be high on break out bar, ma must be rising, must be no nearby resistance etc, etc).

hope that clarifies.

cheers

jon

 

[mr_cassandra]

from my own experience

I work in excel using visual basic code and I have one program which issues both long and short signals. I think you are spot on that you can't invert a long signal conditions and create a short.

This only means you need to take a different route to creat shorts. I work with the sp400 and found that only when its 17ema crossed down thru its 25ema, was it possible to crosscheck with other indicatorss and get a short signals

bear in mind i work with multi-day-week time frames.

regards, Steve

Hello All,

I wasn't too sure where to post this thread but hopefully its in the right location.

I've been developing a mechanical system and to date have only focused on the LONG Trades but have now started looking at developing the Short side of the system.

However, what is beginning to become very clear is the fact that one can't simply "invert" (for a better word) the Buy Signals to produce adequate Sell conditions.

When I backtest the "short" side only, the results are extremely disappointing. I do however, have other ideas which I believe will offer a better performance once backtested for trading the Short side.

However, my question is this:

As the "Short" signal conditions will be totally different to those for generating the "Long" Signals, should I still try to develop the ONE system with both Long & Short signals or should I simply just design another system to trade the Short side only??

What advantages / disadvantages are there will both approaches? I can think of a few but would welcome comments from others......

Many Thanks in advance,

Chorlton

 

[Chorlton]

chorlton

I meant the primary pattern set-up (eg: long on break of first hour's high) which would trigger a trade and then the associated conditions which would determine whether or not the trade was to be taken (eg: volume must be high on break out bar, ma must be rising, must be no nearby resistance etc, etc).

hope that clarifies.

cheers

jon

Jon,

Ah... that does make more sense. To be honest, I would have to say that in my circumstances the entry conditions / pattern set-up are one of the same thing, as I don't think I can distinguish between them.

Basically, I have 4 conditions which all have to be met for me to take the trade. These conditions are given the same amount of weight. If they are all met, the trade is taken without question. No other confirmation is needed. If on the otherhand, any one of them is not satisfied, the trade is not taken.

Hope this clarifies....

All the best,

Chorlton

 

[Chorlton]

I work in excel using visual basic code and I have one program which issues both long and short signals. I think you are spot on that you can't invert a long signal conditions and create a short.

This only means you need to take a different route to creat shorts. I work with the sp400 and found that only when its 17ema crossed down thru its 25ema, was it possible to crosscheck with other indicatorss and get a short signals

bear in mind i work with multi-day-week time frames.

regards, Steve

Steve,

Thanks for your observations / comments. Yes, I tend to agree that I need to look at other ideas to generate the Short Signals!!! That should keep me busy for a while...

 

[barjon]

ok, chorlton

each of your four conditions may have the same weight on the way up but are they equally significant on the way down. can you isolate them to find out?

cheers

jon

 

[Chorlton]

each of your four conditions may have the same weight on the way up but are they equally significant on the way down. can you isolate them to find out?

cheers

jon

Hi Jon,

Thats what I'm going to be doing over the next few weeks. To actually isolate them and backtest won't take very long.

To be honest, I don't believe that isolating them will really help but without backtesting I can't say for certain.

Out of interest, I posted the current results of my LONG System on the thread below. If you get a moment, I'd appreciate your views on it.

http://www.trade2win.com/boards/mechanical-systems-trading/28661-calling-all-system-developers.html

Thanks in advance,

Chorlton

 

[nobel_1101]

Hi Chorlton

I'm currently developing a system and I also have separated long and short trades, for all the reasons mentioned above. I treat the 2 systems in complete isolation because its easy to miss or gloss over when a particular rule/parameter for long trades works when short trades don't, particularly when backtesting results are aggregated/summarised. Downside is it takes a while longer to do.

Regards
Craig