Ok, I’m ready to buy at the Marketplace, please help me

Forums ProRealTime English forum General trading discussions Ok, I’m ready to buy at the Marketplace, please help me

  • This topic has 97 replies, 11 voices, and was last updated 1 year ago by avatarGraHal.
Viewing 15 posts - 46 through 60 (of 98 total)
  • #180266

    @JS, yes correct, the word “unchanged” is certainly key here. I probably should have made that more clear in my posts.

    I do understand that markets and conditions change, but that’s not what this topic is about. The framing here is a profitable strategy that has been running live for some time (well, a minimum of 6 months) and I ask for live results. Backtests are not at all interesting (again, in this topic).

    But I have understood that this probably does not exist, since most people seem to update and change their parameters very often (in my view). I am very, very surprised over this.

    And one more thing, I do not think you have to make a strategy better and better if it continues to perform well. Draw-downs will always happen, and chasing the market conditions with optimized parameters instead of keeping it running – that’s not my thing. With that said, if a strategy is not working, of course I’ll turn it off (which I have done several times).

    Thanks.

    #180267

    since most people seem to update and change their parameters very often (in my view). I am very, very surprised over this.

    I’ve got loads that have been running for well over 12 months, some several years and 2 or 3 since 2016 … without parameter change and they are profitable.

    I am not (yet! 🙂 ) a Seller in the MarketPlace and I thought this Topic was with a view to you wanting to buy in the MarketPlace?

     

     

     

     

    #180269

    Further to above … the loads I refer have been running in my Demo Account.

    In my Live Account, I stop and start if I think an Algo is going the wrong way … it’s that ‘manual trading bias’ again I guess.

    In my Live Account, I regularly change / optimise variables to suit market conditions. I sometimes wonder if I’d make more overall by leaving well alone, but we are what we are! 🙂

    #180270

    When I speak about live results, I refer to live trading with real money, not demo account.

    And yes, you are correct, the main purpose of this topic was to find a strategy to buy/rent. So my intention was that this was obvious, sorry if not.

    #180271

    Just to give my brief comment … I have a trend following strategy that has been running on the DAX in the M15 since 2015 without optimization. The same strategy optimized in 2020 and a version optimized in early 2021. Guess which one works best? The 2015 version does the best. Followed by the 2020 version. The 2021 version is so far a looser. Just as a side note on optimizing strategies. 😉

    #180278

    But I have to say that the original version is from Doctrading. To be fair.

    #180279

    this was obvious, sorry if not.

    Yeah the title and op explains it well, so no need to apologise.  It has made for lively conversation, we maintained Forum standards and respected each other, so all is good?

    Live running on Demo should give very similar results (to Real Live) except IG don’t take overnight fees? Anybody know any other differences?

    #180282
    smp

    Hi Guys,

    The only other thing that we can’t account for (Demo v real) would be the spread of changes that vary in time and width!

    #180283
    smp

    Guys, sorry I pressed submit by accident.  Just in case we have a caption obvious remark.  We can limit the time the algo trades to ensure stable a spread (Given the broker does not change/widen due to volatility and unforeseen events!)

    #180286

    we can’t account for (Demo v real) would be the spread of changes that vary in time and width!

    Surely on Demo Live running / Demo ProOrder Autotrading (not backtesting) we would get / be subject to  the same spread as on Real Live?

    #180288
    smp

    Spread variation is something we can’t control, especially out of instrument normal trading hours.  Also, IG does treat Demo accounts differently to live accounts.  You will often find particularly on Mondays that IG is slow to adjust the demo to fall in line with live accounts so they have the same spread etc.  I have seen this many times, so much so I have a standard email I have to send IG to remind them to adjust the servers.  IG do treat demo accounts differently over weekends!

    #180294

    IG is slow to adjust the demo to fall in line with live accounts so they have the same spread etc.  I have seen this many times, so much so I have a standard email I have to send IG to remind them to adjust the servers.

    There was me thinking it was all automatic.  What would be in it for IG not to use same settings for Demo spread as for Live spread?

    In your observations, does the delay in spread change etc on Demo result in us making more money on Demo (than same Algo on Real Live and so our Demo results are enhanced  compared to same Algo on Real Live)?

    #180295
    JS

    I think everyone knows that markets change over time and, as @smp wrote, this change is related to volatility.

    Therefore, you will have to include volatility in your trading system (bot).

    For me, the standard deviation is the most important part of my trading system.

     

    Furthermore, it is about trust in your trading system, if you do not really trust your trading system you will be tempted every time to stop and / or change your system. I myself am perhaps the best example of this, for years I was tossed back and forth between the well-known emotions “fear and greed” with all the consequences that entails.

     

    But how do you gain confidence in your system?

    I think the only way is through backtesting, when you backtest over longer term and the results are good AND robust why not have confidence in it?

     

    The best thing about automatic trading for me is that I no longer have to sit behind that screen all day and more importantly that my emotions no longer play a role (I was my worst enemy). For me, it’s perfect.

    #180298
    smp

    @GraHal  the issues I mentioned happened just this morning which stopped a bot/trade from entering the market.  I was running a BOT on a demo account and got one of those irritating messages informing me that the FTSE has a minimum stop distance of 200 pips!!!! really!

    I have attached the error message for your view.  I have a pro account and a premium client help so sent off my usual message and got the normal, “sorry we forget to revert from the weekend setting” I am fortunate that I can get this fixed within 10 mins but it’s very irritating.

    Regarding spreads, they always are larger and go against you!

    #180304
    smp

    @JS  I run a trading community and you would be surprised how “everyone” simply does not understand but I do agree with the comment.

     

    I found that backtesting over a long period of time is good for establishing a good strategy for entries but not always for the best limits and stops.   Having said that by setting average stops and limits with a good consistent entry will always reward.  Walk forward testing is key over extending periods to provide confidence but not always.  The DAX would be a case in point, last month consisted of 30 companies and now its 40 companies (DAX40) —just how good is backtesting a strategy on the DAX40 against DAX30 data?

    Through backtesting and establishing long term (Monthly) issues you can then choose to exclude these months walking forward, you can also ghost the forward test with the suspect issues included and review results…Just a thought!

Viewing 15 posts - 46 through 60 (of 98 total)

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