Running multiple variations of the same strategy

Forums › ProRealTime English forum › General trading discussions › Running multiple variations of the same strategy

  • This topic has 31 replies, 10 voices, and was last updated 6 years ago by avatarDerek.
Viewing 15 posts - 1 through 15 (of 32 total)
1 2 3
  • 12/11/2017 at 10:13 PM #55431
    Vonasi

    I’d be interested to know if anyone runs multiple variations of the same strategy?

    By this I mean for example multiple identical strategies but with one variable set at different levels. For example I have backtests that provide positive results with an averaging lookback period set from anywhere from 100 to 500 and with varying levels of profit in between. Rather than pick one averaging period and put £3 on that surely it would be better to diversify and run three versions of the same strategy with £1 on each of 200, 300 and 400 for example. That diversification should ensure a greater chance of a positive return in my mind.

    I’d be interested to know if this is something that works successfully for others or just interested in others opinions on this method of diversification.

    12/11/2017 at 10:23 PM #55432
    Henrik

    I do with different entry or exit. Depends on strategy.

    regards

    Henrik

    12/11/2017 at 10:50 PM #55435
    Vonasi

    Thanks for your reply Henrik.

    I am currently evaluating one strategy running multiple variations with varying exit levels and another different strategy with multiple variations of lookback for an average used for entry. 

    Do you feel that your variations of the same strategy with different entries/exits have benefited you?

    12/11/2017 at 11:11 PM #55436
    Henrik

    Yes i do. Its a way to get around the fact that you can’t partically close positions in one strategy and diversification for changing markets.

    12/11/2017 at 11:20 PM #55438
    Vonasi

    I am also testing out running multiple versions of the same strategy but with different exit points and with the 2nd 3rd and 4th strategies also moving stoploss to break even and to lock in profit as each strategy in turn hits its target. At the moment I am undecided as to whether the latter is better than just having four strategies with different fixed TP and SL levels. My heart says it likes the idea of having three bets running that can do no worse than break even but test results so far indicate that fixed TP and SL levels are more profitable.

    12/12/2017 at 9:22 AM #55453
    Nicolas

    I have read many times that partial closing have been studied at length and that they do not really increase the profitability of a strategy, nor even the management of risk. And it was always better to let his position run until his takeprofit or stoploss. I personally have a different opinion, I do not like to be categorical about any aspect of position management, I guess it’s always closely related to what is really expected of its strategy in terms of performance and his own affective with money and his sensitivity to risk.

    12/12/2017 at 11:49 AM #55473
    Vonasi

    Yes it seems obvious if you think about.

    If you open for example five equal size positions and close one at your target take profit and at that time move the stoploss for the other four to break even or above then these positions are now probability wise (due to the closer stoploss) much more likely to close out for no gain or little gain – so your risk to reward ratio is now very poor. You already risked the money when you opened the bet and now your return is more likely to be nothing or very little. If you’d just placed all 5 bets to close at the same time as the first position then you’d have won five times as much and had those funds now available to increase position size on your next bet.

    Quite often I find with trading that simpler is more often than not better.

    I am finding more and more that any stoploss (except for an emergency bank saving far away stop) is likely to harm a trading strategy more than it benefits it. It appears to be better to just accept that a bet is wrong and have the strategy close it for a loss and open one in the opposite direction. That is the direction my strategies are heading in at the moment anyway.

    Always interested in others thoughts…..

    12/12/2017 at 12:11 PM #55482
    Eric

    I run the same strategy with different levels of the trailing stop  in backtest they show almost the same profit over long term  but on short term it differs depending of what type of market and we never know what the market will do  

     

    (the drawdown  is easier to handle if not all positions are open at the same time)

    12/12/2017 at 5:00 PM #55512
    Vonasi

    I’ve never had much luck using trailing stops. They are either too close and stop strategies prematurely or too far away to be of any use for locking in profit in my (limited) experience. I’ve tried ATR based stops but without much luck as the market does not always want to do the same today as it did yesterday or even the same as it averaged over the last year.

    What do you base your trailing stops on Eric, fixed levels that just seem to work or off of an indicator?

    12/12/2017 at 6:12 PM #55521
    Eric

    I use “Nicolas” trailing stop from the forum and its based on points that i have backtested and optimized

    and it works live to (so far)

    but it depends on type of strategy i guess, mine is a breakout trendfollowing type

     

     

     

    12/13/2017 at 11:24 PM #55620
    jebus89

    i would think that with the same entry, only the exit should be different. As in you might want 1 exit at 50 pips while you let the other one run longer.   In my head that sounds logical. different entries based on the same system but with slighty different variables would most likely fail the same trades. meaning that if 1 trade fails, probably all X trades will fail. Say you get 6 multiple loss in a row, if you have say 4 systems failing at the same time its 6×4 = 24 failed trades in a row. That would definitly hurt my bank account. On the other side if you have 2 systems, with different exits, you might fail both at the same time, but the upside is that you get those long fat trends as a bonus while the one caps profit. Im not close to being semi-pro, just my thoughts.

    12/14/2017 at 12:02 AM #55621
    Nobody

    Correlated strategies increase risk  , ultimately id rather have totally uncorrelated systems running simultaneously , more likely to smooth the net curve than increase the bumps  . 

    12/14/2017 at 7:21 AM #55625
    mr blue

    Without having read all posts here I would like to suggest the following (well that’s what I do currently):

    I have a real trading strategy with different risk models implemented. to distribute the risk i have created multiple cfd accounts on the broker side. that allows me, to run them independently and not to risk that if one goes wrong, it will “burn” others account money.

    If this is not enough, there is the possibility to distribute variants of the strategy on different cfd contract types.

    For example DAX, you have on IG side the 1E, 5E and 25E contract available. That gives you at least 3 variants – combined with – for example 3 cfd accounts – it makes a total of 9 strategy variants. 

     

    For me, that’s more than enough.

     

    Hope this helps.

     

    12/14/2017 at 2:16 PM #55668
    Vonasi
    jebus89 wrote:

     different entries based on the same system but with slighty different variables would most likely fail the same trades. meaning that if 1 trade fails, probably all X trades will fail. 

    True but if you had £5 to put on one bet or five lots of £1 to put on five bets with one variable set differently then it is the same result if they all lose. The latter would have the benefit that maybe one would win and reduce your loss and if they all won then you would have increased your chances of having the variable in the sweet spot on at least one bet or at least have averaged your chances out of being somewhere in the zone.

    12/14/2017 at 2:25 PM #55669
    Vonasi
    Brisvegas wrote:

    Correlated strategies increase risk , ultimately id rather have totally uncorrelated systems running simultaneously , more likely to smooth the net curve than increase the bumps .

    Bumps are fine – as long as the up bumps out number the down bumps or are big enough to make the up return bigger than the down return!

    I’m not suggesting putting all our eggs in one basket but just running multiple versions of each strategy in an uncorrelated portfolio of strategies. Say five uncorrelated strategies with five versions of each running with one variable adjusted in each. Our normal position size divided evenly between each version so that our amount risked is no different.

  • Author
    Posts
Viewing 15 posts - 1 through 15 (of 32 total)
1 2 3

Find exclusive trading pro-tools on ProRealCode Marketplace

Similar topics:
running two instruments at the same time
adding input from custom indicator into strategy
andrew cardwell strategy on rsi
array/bubble sort in strategy doesn’t work, but works as an indicator
arrays & same candle programming
auto restart automated strategy
bollinger+fractal strategy idea
breakout box strategy (tick by tick not working for entries)
can you add compounding to a strategy?
central bank dealers range – reversal strategy help please!

Create your free account now and post your request to benefit from the help of the community
Register or Login