Algo stopped Position closed

Forums › ProRealTime English forum › General trading discussions › Algo stopped Position closed

Viewing 15 posts - 361 through 375 (of 419 total)
1 2 3 4 5 … 15 16 17 18 19 20 21 22 23 24 25 26 27 28
  • 09/29/2025 at 2:51 PM #251983
    jebus89

    to get back on topic (not spread related, is this maybe off topic now in this thread?? Lol)

    Has anyone talked directly to IG? Do they just blame PRT?

    Has anyone tried to go to IB, does the strategy look different (IB data vs IG data?)? Or can you really just press play there and expect same results as in IG?

    09/29/2025 at 2:56 PM #251984
    XORANDNOT

    This is for 1000 contracts (who can trade this ? I can’t) in Switzerland.

    When you look at the table of spreads, depending on order size, the spreads are higher in Switzerland than in the Netherlands, from 230 contracts upwards, because the margin is lower and IG’s risk is higher (I think).

    09/29/2025 at 3:01 PM #251986
    PeterSt
    jebus89 wrote:

    Or can you really just press play there and expect same results as in IG?

    Don’t skip too many posts while reading. 🙂
    That was all dealt with hours ago.

    Results are totally different because your algos need to be different.

    For this also keep in mind :
    IG reserves a huge bandwidth to correct things for herself (it contains their hedging). E.g. 24.880 for the NASDAQ Future while the price for US tech 100 (based on the underlaying NASDAQ) is 24.717. This is 160 points bandwidth (theoretically towards both sides thus 320 points) and in that bandwidth they can do what they want, which they really do (fishy as h*ll. But alas, this is how hey work and earn their money. Meanwhile the price bumps all over the place which does not occur at IB/reality. Also, IG bears a 10-fold of “ticks” which are their internal ticks/trades again.

    09/29/2025 at 3:02 PM #251987
    XORANDNOT
    jebus89 wrote:

    to get back on topic (not spread related, is this maybe off topic now in this thread?? Lol)

    Has anyone talked directly to IG? Do they just blame PRT?

    Has anyone tried to go to IB, does the strategy look different (IB data vs IG data?)? Or can you really just press play there and expect same results as in IG?

    I think they had a major overhaul of instruments at IG and adapted minimum order sizes and spreads, and additionally, they broke these down depending on the country. Now there are more instruments than before and PRT has some problems to incorporate them all.

    Every strategy will look different with IB data vs IG data. You can absolutely NOT expect the same results. When you optimized a strategy with IG data, it will in most cases be worse with IB data. Or the other way round : when you optimized it with IB, it will usually be worse at IG – because it has not been overoptimized with this data.

     

    It shows : when you think, DAX is DAX, this is wrong. The idea that you can get a functioning trading system by optimization of the past is in many cases simply wrong.

    09/29/2025 at 3:04 PM #251988
    PeterSt

    Side by side …

    09/29/2025 at 3:15 PM #251990
    XORANDNOT

    For Germany, exactly the same as in the Netherlands.

    09/29/2025 at 3:40 PM #251991
    PeterSt

    Scratched post. See later in new post.

     

    This may remain :

    The lot summarized : when you envision swings large enough to flatten details like more ticks on IG and averages working out differently, IG and IB can be quite the same. But don’t try Scalping, to mention an opposite.

     

    09/29/2025 at 3:52 PM #251993
    XORANDNOT

    I compared only intraday sytems on IG and IB. There were always huge differences. For a daily chart, it may work better, but I have not tried.

    09/29/2025 at 4:23 PM #251995
    PeterSt

    This is what I wrote in my previous post, but scratched :

    Anyway, for that 4 weeks worth of data both look sufficiently the same to me, so I won’t be optimizing it for IG (if I would run it for Live there at all).
    I nicely normalized both to the same investment (25 DAX contracts for IG), gave IG the spread of 1,5 we have just been talking about (0.75 ct per  side for IB (DXM 5 contracts)). Same start date/time, same end date/time (I may have missed a second or two).

    Mind you, this is an example of the System I mentioned a few hours ago : one means that explicitly should cover for both IB and IG. But no time to implement them because, well …

    I already know that Forex performs better on IG.  My explanation : more swinging within mentioned bandwidth (these Systems highly depend on swinging).

    So what I likED to do was present similar results for IB vs IG. But that miserably failed. The two pictures in the previous post seemed to do quite well, but on a longer term … no, not at all. But this is DAX and I never tried DAX on IG (the green pictures).

    US Tech 100 I did try (this was running Live when the lot crashed) and although it compares fairly well to the IB version (blue) I can not imagine that I would have let this run.

    So the systems can be the same ? maybe, and for that we also need to look broader. See the third attachment …
    The first two attachments comprised the backtest period (that’s right, only 2,5 months). The third attachment actually shows the forward test relative to that, because it was never backtested from of August 15. (BUT as said, runs real-Live on IB and was killed on IG).  So despite I wouldn’t have let go that System for IG, you can see it caught up with the IB version. People may attest : the data has changed ? Maybe, but I see the behavior being sufficiently the same, so I personally think “so what”.

    I hope I wasn’t too boring and I certainly hope I did not make mistakes. I deem it quite difficult to make these comparisons while taking all into account.

    09/29/2025 at 4:39 PM #252000
    PeterSt

    PS: For detailists, the green charts are in EUR and the blue in USD (different account currencies). This means that when we see 48.387 in the green chart, it is ~57.096  USD in the blue chart. This is quite close to the 61.717 shown, *if* that would be of importance at all (for me the curve is more important – to each his own).

    09/30/2025 at 8:15 AM #252014
    Nicolas

    Update on migration progress

    We have completed the migration of all chart objects.
    We are now actively migrating strategies and trading history.

    Regarding the spread issue
    Please note that the spread displayed in the platform is correct, as it is calculated using the Bid/Ask data sent directly by the broker.

    If you notice differences, please make sure you are comparing the exact same instrument/contract.
    If you still find a discrepancy, kindly provide us with the IG instrument URL (as explained in my previous message) so we can investigate further.

    Thank you for your patience and cooperation as we continue this work.

    2 users thanked author for this post.
    avatar robertogozzi, avatar justisan
    09/30/2025 at 8:52 AM #252017
    justisan

    I don’t find serious reasons why to horrify folks telling that trading cfd and futures will show “totally different” results. on other hand everybody needs to understand that moving an algo from cfd to futures of same instrument is definetely not “plug and play” activity. the reality is somewhere in between. there are few very simple reasons for that and few of them are more compex. one of the simple e.g., talking here specifically about dax: close price of the day for futures is at 22:00, for cfd currently it’s price at 01:00 next day (next day, Carl!). so if close price of the day is important for one’s system (directly or via some indicators based on it), results will deviate, maybe a bit, maybe a lot – depending on how sinsitive one’s particular system is reacting on this data. but it’s not that big deal to adjust the code to that.

    I assume the issues with the price data are getting bigger and bigger when one moves from daily data to the smaller and smaller granularity, down to seconds and ticks, and the reasons become more and more complex up to the level that one can possibly not adjust any more in simple way. cfd are not real/free market – counterparty/”broker” is to some extent  manipulating the prices, and to some extent this manipulation is huge, even if temporarily only. it’s a bit like in the bucket shops in U.S. 100-150 years ago, just with some high-tech behind. I assume U.S. were banning cfd for several very good reasons. but still – if one has generally sound concept for the system/trading logic, one probably will be able to “translate” it for free/real market environment of futures trading. I personally did not need to “scrap” any of my systems when moving them to futures. I needed to adjust all of them, yes, but these adjustments were mainly not some re-optimizations (not using any indicators, so there is only very little room for any kind of “optimization” anyway).

    also, even dax futures are not equal to dax futures: exact same algo on 25eur contract will produce different results on 5eur contract and again different on 1eur contract. all these contracts “move” not 100% in sync (especially the micro contract is a bit of desaster). so what? go back to buckt-shop? sure not.

    cheers

    justisan

    1 user thanked author for this post.
    avatar XORANDNOT
    09/30/2025 at 1:23 PM #252025
    XORANDNOT
    justisan wrote:

    I needed to adjust all of them, yes, but these adjustments were mainly not some re-optimizations (not using any indicators, so there is only very little room for any kind of “optimization” anyway).

    Maybe this is the main point ?

    09/30/2025 at 2:48 PM #252027
    justisan

    yes, probably it is, at least it’s one very heavy point. I was in fact amazed how similar (very, very far from being “totally different”) the results of many algos in backtest were, when I ~2 years ago moved first algos from cfd to futures. can you imagine, how reinforcing, trust-building it was? and the only explanation I had for myself was that it’s due to dumbass simplicity of the trading logic behind. another point is possibly that my stop loss orders are usually quite far away from latest price, so performance of all algos were not so much affected by internal manipulations of cfd prices / short term price spikes up or down, which one can observe on cfd while absent on futures prices.

    I think/assume somebody maneuvering in very tiny timeframes, correspondingly having entries and exits very close to the latest price, and possibly applying complex trading logic based on multiple indicators will possibly have totally different experience when moving from cfd to futures, options etc.

    09/30/2025 at 3:44 PM #252028
    phoentzs

    I’ve observed something similar… after this chaos, my strategies based on H4 candles are very stable. Even though these candles still end on the wrong hour. Why, exactly??!!! My scalpers in the TF M1 are also very stable. The basis of these scalpers, however, is again… the TF H4.

  • Author
    Posts
Viewing 15 posts - 361 through 375 (of 419 total)
1 2 3 4 5 … 15 16 17 18 19 20 21 22 23 24 25 26 27 28

Find exclusive trading pro-tools on ProRealCode Marketplace

Similar topics:
⚠️ missing closed trades in ig trade analytics – incident inc0927166
43 algos stopped – due to an error!
alarm from closed workspaces
algo alpha range filterd trend signals – conversion required
algo swap anyone?
attached orders not closing when order bar is closed
best indicators/strategies to exit from a long position
code to exit position after 30 bars if long still in profit
code to exit position at next bar
combine manual trading with algo

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