Advice on trading systems

[Fisherfisher]

so you did, I do apologise. you probably would have had 3 losses on the trot so a loss of 300 within that period which in itself isn't really that bad.
the only filter i can think that you would probably want is something to identfy the trend. you would i guess be looking at buying at 33 when the trend is up rather than at all times.
To be honest this should be quite easy to code and backtest. let me take a look and see how it may have performed historically and then see if a filter would improve the results if any

OK thank you, I was thinking of using a moving average as well.

 

[malaguti]

OK thank you, I was thinking of using a moving average as well.

here is just the base result, using 1 contract
buy at 33, sell at 66. immediately this raises an assumption that at 66 you now close the original and then short the same number until the next buy at 33 which is how the test was performed
so your total result..(this isn't annual) since 2005 is a little over £4100 without a stop loss and without being able to specify the exact number of contracts and without any additional filters
so you certainly do have something to work on..
as soon as I can code a little bit more into the equations I'll let you know

 

[kalott]

http://www.prorealcode.com/prorealtime-trading-strategies/rsi-2-strategy-larry-connors/


http://www.prorealcode.com/prorealtime-trading-strategies/internal-bar-strength-qqq-strategy/

 

[Fisherfisher]

here is just the base result, using 1 contract
buy at 33, sell at 66. immediately this raises an assumption that at 66 you now close the original and then short the same number until the next buy at 33 which is how the test was performed
so your total result..(this isn't annual) since 2005 is a little over £4100 without a stop loss and without being able to specify the exact number of contracts and without any additional filters
so you certainly do have something to work on..
as soon as I can code a little bit more into the equations I'll let you know

Thank you I really appreciate your time, what software to you use to test your systems?

 

[malaguti]

Thank you I really appreciate your time, what software to you use to test your systems?

I was using tradingview but their backtesting ability is very limited
I will do this in the software which i always use to backtest which is omnitrader. if you can wait a couple of days i'll see what i can do

 

[Fisherfisher]

I was using tradingview but their backtesting ability is very limited
I will do this in the software which i always use to backtest which is omnitrader. if you can wait a couple of days i'll see what i can do

Ok thank you

 

[darktrade]

I tried to backtest your system and it shows a very low win rate (~29%), a high drawdown (~40%) and a small annual return. If I am right of course, why do you want to trade something like that?

 

[malaguti]

Ok thank you

apologies for the delay
I've tested with your criteria using the 100pt stop
and the results are pretty low. I get a hit rate of just 32% which isn't good
I added a trend filter simply by only allowing a trade with a close over the 200sma
thats improved the results with a hit rate of 42%. Better, but still i'd be looking for more of an edge
Always good to backtest, and automating makes things easier.

good luck though

 

[Fisherfisher]

apologies for the delay
I've tested with your criteria using the 100pt stop
and the results are pretty low. I get a hit rate of just 32% which isn't good
I added a trend filter simply by only allowing a trade with a close over the 200sma
thats improved the results with a hit rate of 42%. Better, but still i'd be looking for more of an edge
Always good to backtest, and automating makes things easier.

good luck though

Thank you for testing the system, i have been trying to improve it and i think i am going to use the adx indicator aswell. Using rsi 33 or 66 for entry then rsi 66 or 33 to when the di+ and di- cross over against the position (- over + on a buy) for an exit.I think this will help cut loosing trades as the adx usually crosses on a bad trade before reaching the rsi exit point.

 

[malaguti]

Thank you for testing the system, i have been trying to improve it and i think i am going to use the adx indicator aswell. Using rsi 33 or 66 for entry then rsi 66 or 33 to when the di+ and di- cross over against the position (- over + on a buy) for an exit.I think this will help cut loosing trades as the adx usually crosses on a bad trade before reaching the rsi exit point.

well the back test results was quite clear in that your entry needed work. A 10% increase in hit rate by just adding the 200ma as a filter says a lot to me.
Not saying your exit also couldn't have improved, just an observation from the results
and 200ma is not exactly the best filter in the world

the other thing i noticed was not the filter, but the entry itself
your long entry is governed by prices falling..your entry gets hit as soon as it hits 33, ie price is going down. you wouldn't want to confirm that price is actually going back in your direction before buying...after its hit 33?
suppose for example your buy was triggered after its broken the previous days high or last two days high after the rsi had hit 33. that would probably give you a better entry criteria and at least suggests price is going back in your favour.
then add a trend filter given that increased your overall odds

once you've got your entry a little more refined, then start working on your exit

just a couple of points..not taking anything away from your idea, just trying to help. I've been there.

 

[Fisherfisher]

well the back test results was quite clear in that your entry needed work. A 10% increase in hit rate by just adding the 200ma as a filter says a lot to me.
Not saying your exit also couldn't have improved, just an observation from the results
and 200ma is not exactly the best filter in the world

the other thing i noticed was not the filter, but the entry itself
your long entry is governed by prices falling..your entry gets hit as soon as it hits 33, ie price is going down. you wouldn't want to confirm that price is actually going back in your direction before buying...after its hit 33?
suppose for example your buy was triggered after its broken the previous days high or last two days high after the rsi had hit 33. that would probably give you a better entry criteria and at least suggests price is going back in your favour.
then add a trend filter given that increased your overall odds

once you've got your entry a little more refined, then start working on your exit

just a couple of points..not taking anything away from your idea, just trying to help. I've been there.

I really appreciate your input, good advise is hard to find! ill have a look on my entries and find away to make sure the trend had reversed before i enter the trade. Along as i can make a consistent profit however much it may be id be happy :cheesy: I'm determined to make a success of this.